Reading Noise: A Flaw in Human Judgment, by Daniel Kahneman, Olivier Sibony, and Cass Sunstein, brought me back to an exercise I have used in my teaching: examining startup cemeteries, collections of postmortems describing ventures that closed, often after exhausting their seed capital.

I ask students to examine what happened. What did the founders believe? What evidence supported those beliefs? Which assumptions survived longer than they should have? What might someone have questioned before the money ran out?

These accounts are incomplete. They are retrospective explanations, sometimes written by people defending their decisions or trying to make sense of a painful outcome. We rarely get the full customer evidence, board discussions, or alternatives available at the time. Knowing how the story ends also makes earlier mistakes easier to see.

Still, a pattern repeatedly emerges in our discussions: a confident development path ends with some version of “we built something people did not want enough to buy.”

By then, the product exists and the runway does not.

Most of my students reach a similar conclusion. Early in an innovation journey, understanding what customers are trying to accomplish, where they struggle, and which underserved outcomes matter enough to motivate a purchase or a switch deserves serious attention.

That work can often begin before extensive development, hiring, and market entry. Interviews alone cannot establish demand, and some questions require a working product. But investigating customers’ circumstances and existing behavior can expose weak assumptions while changing direction is still relatively inexpensive.

The lesson seems accessible. The cases are public. The methods are familiar. People can explain why building on an untested demand assumption is dangerous.

So why does recognizing the mistake in someone else’s postmortem fail to prevent the next organization from funding it?

My interest in this is also personal. I have encountered these problems in corporate innovation decision boards and as a startup co-founder working with my own advisory board and board of directors (and it was also the topic of my master thesis). Across those settings, I experienced differences in judgment and shared assumptions that I now examine through the lens of noise and bias.

I was part of those judgments too. My proximity to a venture gave me information others lacked. It did not make my interpretation objective.

Noise sharpened a question that connects those experiences: how much uncertainty lies in the venture itself, and how much avoidable error enters through the judgments of the people evaluating it?

One venture, a hundred boards

A postmortem gives students an ending. A live investment proposal gives executives a team, a prototype, encouraging customer comments, commercial projections, and a request for resources.

The decision-makers bring experience. They recognize patterns, recall previous cases, and apply professional judgment. Their colleagues may find the interpretation convincing.

But would another equally qualified board see the same venture in the same way?

What about a tenth board? A hundredth?

Give them the same evidence, objectives, and resource constraints. Ask each to assess what has been demonstrated, what remains assumed, and what commitment is justified. Would their judgments cluster closely or scatter widely?

The board making the actual decision rarely sees that distribution. It sees its own interpretation, discussed among people who may share professional backgrounds, organizational norms, and years of exposure to the same internal information.

The shooting-target illustration in Noise makes an important distinction visible. Shots can scatter around the center, cluster away from it, or combine both problems. Noise concerns unwanted variability in judgment. Bias concerns systematic error.[1]

For a venture, however, the center is partly hidden. The commercial outcome has not happened yet. We can compare judgments before we can establish their accuracy.

Independent errors may partly offset when judgments are appropriately aggregated. But a board does not necessarily aggregate independent assessments. It may follow its most influential member. Its members may also accept the same misleading premise.

A tight cluster demonstrates agreement. It does not establish accuracy.

Nor is every difference a defect. One reviewer may recognize a purchasing barrier others have missed. Another may understand why enthusiasm for a prototype will not translate into switching behavior.

That disagreement is useful if it reveals information or assumptions that improve the decision. Other differences may reflect conflicting objectives, acceptable losses, or unexplained inconsistency. They require examination before being dismissed as poor alignment.

Follow me on LinkedIn

When contrary evidence changes nothing

In corporate innovation, I have presented primary market research that challenged claims made in venture proposals.

What had been described as demonstrated demand looked materially different in my findings. Expensive prototypes existed, but the evidence connecting a relevant customer struggle to willingness to pay or switch remained weak. Proposals to proceed depended on economics, acquisition approaches, or ecosystem arrangements that were still insufficiently established.

These distinctions matter. A useful product does not automatically establish a viable business. Paying customers do not automatically establish a repeatable acquisition model. Equally, a weak description of the customer’s job does not itself disprove demand.

My question was whether the evidence justified the next commitment.

In those cases, the discrepancies were tolerated by the decision board and the projects continued. What remained unresolved was the gap between what was claimed and what the research could substantiate.

Continuing was not necessarily the error. An uncertain venture can deserve another experiment. The problem was proceeding without adequately correcting the account of what had already been established.

“We are testing demand” and “we have demonstrated demand” describe different investment cases.

The first can justify buying evidence. The second can justify expenditure that depends on the evidence already being there.

Contradictory research does not automatically prove that a venture will fail, or that its supporters are biased. Samples, segments, prices, and conditions must be compared. But the discrepancy needs an answer.

If favorable findings repeatedly count as confirmation while unfavorable findings leave the claim untouched, the process is favoring one conclusion.

My experience as a co-founder also makes me cautious about assigning this problem exclusively to boards. Founders interpret the market before presenting it. We decide which observations matter, which objections seem temporary, and which previous cases appear relevant.

Even an honest presentation is a selection of evidence. The board then judges both the venture and the founder’s interpretation of it, sometimes without clearly distinguishing the two.

Experience without reliable feedback

Repeated decisions do not automatically produce better judgment. They can produce greater familiarity and confidence without a corresponding improvement in accuracy.

Kahneman and Klein argue that reliable intuitive expertise requires sufficiently predictable patterns and opportunities to learn through useful feedback. Subjective confidence alone does not establish expertise.[2]

Venture assessment can provide difficult learning conditions. Outcomes arrive late. Markets change. Funding and execution influence results. Rejected opportunities may never produce an observable outcome. Unless original forecasts are recorded, decision-makers may not discover which judgments were consistently too optimistic or too pessimistic.

An executive can become highly practiced at reviewing investment proposals without becoming equally accurate at assessing their prospects.

A board may accumulate experience faster than it accumulates evidence about the quality of its own judgment.

Shared experience can also make a particular interpretation feel obvious. Ross and Ward describe naïve realism: experiencing our own view as objective and expecting other informed, reasonable people to reach similar conclusions.[3]

Inside an organization, a contrary assessment may consequently appear to reveal the outsider’s lack of understanding before it prompts an examination of the prevailing assumptions.

An outsider is not automatically right. Their evidence and methods require scrutiny too. Their value lies in making reasoning visible that the organization may otherwise leave unquestioned.

A single independent question can expose a consequential gap: “What evidence establishes that these customers will switch, rather than merely express interest?”

The answer may reveal disagreement within the board. It may also reveal that everyone accepts the same unsupported inference.

Interested in working together?

Reach out

Agreement can conceal disagreement

Authority adds another complication. Research by See and colleagues found that greater power was associated with less advice-taking, with elevated confidence helping explain the relationship.[4] That does not diagnose any particular executive. It identifies a risk where confidence carries substantial influence.

Someone leading the discussion may sincerely believe they recognize the case. Someone else may see a weakness but judge the relationship cost of challenging it too high. The resulting silence can resemble confirmation.

Kahneman and Sibony describe how discussion can conceal differences when participants converge around a senior person’s opening assessment.[5]

A board can therefore align on an action without agreeing on why it is justified. One member believes demand is established. Another believes the next expenditure will test demand. A third assumes someone else has checked.

The vote is clear. The investment premise is not.

Genuine alignment allows disagreement to remain explicit. Two reviewers can estimate different chances of commercial success and still support the same bounded experiment because its results will inform the next decision.

The problem begins when agreement on an action is treated as agreement on the evidence, and agreement on the evidence is treated as proof that it is correct.

Meanwhile, commitments accumulate. An assumption about demand supports a revenue forecast. The forecast supports hiring, development, and market entry. Those expenditures increase the cost of being wrong without necessarily strengthening the original assumption.

Resources consumed here are unavailable elsewhere. Learning may remain if the venture closes, but its value depends on subsequent use. An archived presentation does not automatically repay the expenditure.

That is why the postmortem lesson is insufficient. Knowing that customer demand matters does not establish that founders, advisers, or boards can reliably judge the evidence of it.

Nor does a successful outcome retrospectively validate every decision that preceded it. A venture can succeed despite weak reasoning, just as a justified bet can fail.

Before funding the next stage, examine how the board knows what it believes it knows.

The paid section below provides a Venture Judgment Review worksheet and a worked example for exposing consequential differences before another commitment becomes difficult to reverse.

Thanks for reading INNOVATION&! This post is public so feel free to share it.

Share

Start paid learning and application: the Venture Judgment Review

Use this review before a consequential funding decision. It is a practical adaptation of independent assessment, not a validated diagnostic instrument.

Its purpose is to distinguish evidence, interpretation, and investment preference before discussion turns them into one apparently shared conclusion.

Read more

For a long time, I have returned to a question that has become harder to answer with every team and organization I have experienced: what makes a group exceptionally good—or exceptionally bad?

I have worked with many capable people, helped build businesses, and watched teams form, grow, reorganize, and disappear. Yet I have experienced that special, almost magical team feeling only once.

I mean the feeling that the group makes everybody in it better. The work is demanding and discussions can be intense, but you learn because somebody sees what you cannot see. You attempt things you might not have attempted alone. Differences in background and personality become useful rather than exhausting.

That experience happened more than twenty years ago, inside a large corporation responding to deregulation. The company established a new wholesale unit, recruiting many people externally. We came from different industries and parts of Switzerland, represented different generations, and brought very different professional experiences into the same group.

We had little shared history. What we shared was a challenge that none of us could address entirely alone.

Recently, I contacted former colleagues to ask what they remembered. One replied, loosely translated:

People from different industries came together and pursued one common goal. Different age groups, many different and quite special characters from across Switzerland. We learned openly from each other. And because you were there. 😁

I will treat the final point as friendly nostalgia rather than scientific evidence. What interested me was what remained after twenty years: difference, a common goal, and learning together.

Reading ’s The Collective Edge gave me a way to reconsider that experience. Perhaps what I remembered as magic was a reinforcing process we had never named.

Perhaps the magic was a feedback loop

Fisher describes groups as dynamic systems. Good work design, meaningful goals, coordination, effort, and performance can reinforce one another. When membership becomes valuable, people invest more, learn together, and develop capabilities that strengthen the group further. The reverse cycle can also occur: weak structures and disappointing work reduce participation, capable people leave, and the group loses some of its ability to improve [1].

That interpretation fits my experience, although our memories cannot establish what caused the team’s success.

My capabilities grew because I worked with people who knew things I did not know. I encountered unfamiliar industries, ways of negotiating, sales approaches, and interpretations of customer behavior. Over time, I could make judgments that would previously have required somebody else’s help. I also became better at recognizing when I still needed that help.

The work developed us while we were doing it.

A great team compounds human capability.

That is the proposition I now find more useful than discussions about an ideal team composition. Alongside what a team delivers, we should examine what happens to people while they belong to it. Does their judgment improve? Does their expertise become more accessible to colleagues? Can they tackle problems they would have struggled with alone?

Follow me on LinkedIn

Difference has to become useful

My former colleague remembered our different backgrounds. It would be tempting to conclude that diversity explained our performance, but that would skip the important part: what we did with those differences.

Van Knippenberg, De Dreu, and Homan developed their categorization-elaboration model to explain why research on diversity and performance had produced inconsistent findings. Different backgrounds can introduce relevant information and perspectives, but the benefits depend on whether the group examines and integrates them. Social divisions can prevent that information from becoming useful [2].

In our team, the changing market gave our different experiences a common purpose. Shared learning helped us develop enough common language to work together without requiring everybody to think alike.

We also gradually learned whose judgment to seek. A difficult negotiation, an unfamiliar customer situation, or a legal question did not require everyone to become an expert. It required knowing where relevant expertise existed and involving it at the right moment.

Research calls this a transactive memory system: the group’s understanding of who knows what and how to coordinate that knowledge. A meta-analysis by Bachrach and colleagues links these systems with team performance, while emphasizing the context in which they develop [5].

This is one reason I am skeptical of innovation-team recipes such as “hacker, hipster, hustler.” They remind us that complementary capabilities matter, but they can make team design look like a staffing puzzle. Filling the roles leaves a substantial question unanswered: can these people actually use one another’s knowledge?

A team can possess the required expertise and still make it difficult to access.

We noticed more than the work

Another quality became clearer as I reflected on the old team: people noticed what was happening to one another.

Sometimes the situation was straightforward. Someone struggled with a customer, encountered a legal complication, or faced a problem outside their experience. A colleague stepped in or connected them with someone who could help.

At other times, someone was going through a difficult period or behaving differently from usual. Another person noticed and approached them. Support could begin with a conversation rather than waiting for the difficulty to become a formal performance issue.

Looking back, I think that attentiveness contributed to how we worked together. We understood something about who knew what, but also about who might need help.

Woolley and colleagues’ research offers a relevant connection. In their group experiments, collective performance was associated with members’ average social sensitivity and more equal conversational participation [3]. Those findings do not retrospectively explain our team, but they provide a reason to look beyond individual intelligence when examining group capability.

For innovation work, the practical issue is whether relevant information receives attention. A colleague hesitates over an assumption. A salesperson hears a recurring objection. A developer raises a concern that does not yet fit the roadmap. Someone who usually contributes becomes quiet.

None of these signals has a single interpretation. Attentiveness means noticing and investigating, rather than assuming we already understand.

I would not claim that interpersonal sensitivity automatically produces better market sensing. But I would ask how effectively a team can investigate weak customer signals if it routinely overlooks the concerns of people sitting in its own meetings.

A system can lose access to talent before people leave

I have experienced the reverse dynamic too. The contrast mattered precisely because one of my best team experiences had occurred inside a large corporation.

In other environments, I experienced expectations becoming more ambitious while the strategy connecting them to reality became less clear. Work accumulated, but a growing share seemed disconnected from meaningful progress. Reporting and coordination consumed more attention, while the opportunity to use or develop expertise became harder to find.

You could invest your own time and money in education, return with new knowledge, and discover that little changed in the work you were trusted to perform or the decisions you could influence. Development was encouraged, but its practical use remained limited.

I watched capable people become more selective about their contribution. My interpretation was that some were protecting their energy after repeated efforts produced little influence, learning, or recognition. They continued doing their jobs while becoming less willing to invest everything they could offer.

That interpretation will not explain every instance of reduced participation. Workload, personal circumstances, unclear responsibilities, or a change in the task may also matter. But automatically treating the outcome as an individual engagement problem excludes the system from examination.

An organization can retain talented people while losing access to much of their talent.

This is where psychological safety becomes relevant. Edmondson’s research connects a shared sense of safety for interpersonal risk-taking with learning behaviors such as asking for help, seeking feedback, and discussing errors [6].

Innovation requires those behaviors because important information can threaten an attractive plan. Someone needs to explain that the customer problem may be less important than expected, that the technical approach has limitations, or that the pilot results do not justify the proposed investment.

The question is whether that information can be examined without turning the discussion into a test of the messenger’s loyalty or competence.

Interested in working together?

Reach out

When reasonable responses reinforce decline

The difficult part is that organizational decline can emerge from individually understandable responses.

People encounter work they cannot influence and become more protective of their effort. Managers see weaker results and request more reporting. Reporting leaves less time for investigation, customer contact, and development. Experienced people depart or narrow their contribution, increasing the coordination burden on those who remain.

This is a possible reinforcing cycle, not an explanation to impose on every struggling organization. Its value is that it directs attention to relationships between conditions that might otherwise be treated separately.

An engagement problem, a reporting problem, and a retention problem may be connected. Addressing each independently can leave the process producing them intact.

That is why I would evaluate organizations in both directions. We already ask what people contribute to the organization. We should also ask what participation in the organization is doing to them.

Are they developing better judgment? Is their expertise becoming more useful to others? Can they introduce uncomfortable evidence? Or are they primarily becoming better at navigating internal expectations while keeping their more valuable contributions to themselves?

The next uncertainty changes what the team needs

Innovation teams face an additional complication: the work changes what they need to know.

Early questions may concern customer struggles and switching behavior. Technical feasibility, purchasing, delivery economics, and retention can become more consequential as the venture develops. These uncertainties overlap; they do not disappear in a neat sequence.

A team that is excellent at understanding a problem may need additional expertise to establish a viable delivery model. A strong technical team may lack access to the people who control purchasing. The appropriate response could be development, temporary support, partnerships, or a change in responsibilities and membership.

The challenge is to recognize the gap accurately. Is the capability absent, or is existing expertise being overlooked? Would another specialist help, or would that person encounter the same constraints?

Continuity matters because trust, shared knowledge, and coordination take time to develop. Openness matters because an established team can become highly capable at addressing questions that are no longer the most important ones.

The task is to preserve what the group has learned while enabling it to become the team the next uncertainty requires.

What the team makes possible

I no longer think the team I remember was magical, although I still value how unusual the experience felt.

We had different starting points and learned to use them. We noticed when someone needed help or saw something others had missed. The work gave our contributions a purpose, and participating in it expanded what we could understand and do.

I cannot reconstruct every cause after twenty years. But the experience changed the standard against which I assess a team.

Individual talent matters. So do the conditions that allow it to influence the work and develop through other people.

Before asking whether you need better people, examine what your team makes possible for those already there. Which knowledge reaches decisions? Which contributions repeatedly disappear? What are people becoming better at through working together?

Those questions will not reveal a universal recipe. They can reveal whether the next intervention should concern expertise, the way work is organized, or the conditions under which people contribute.

The free section of INNOVATION& is offered freely, and you are welcome to keep reading it that way. If these articles help you make better decisions in your working life, please consider becoming a paid subscriber. Researching and writing them takes time, and paid subscriptions provide the income I need to continue. Your support also helps keep part of this work available to everyone.

Start paid learning and application section

Is your team missing capability—or making it difficult to contribute?

Before approving another hire, commissioning team training, or reorganizing responsibilities, examine what happens to the expertise already available.

The Collective Capability Review helps you investigate whether your team lacks a relevant capability, struggles to access it, or operates under conditions that make contributing it difficult.

In the lesson below, a worked example shows how similar symptoms can lead to different explanations and interventions. You can then use the review worksheet to examine your own team.

You will leave with a specific capability problem, competing explanations, a bounded change to test, and observable signs that would justify continuing or reconsidering it.

Read more

I deliberately try to leave my own field from time to time and listen to people who know far more than I do about something completely different.

I have written about this before in The Innovator’s Learning Pod. I strongly believe in cross-pollination because it exposes us to perspectives, questions, and mental models that are difficult to generate from inside our own field.

If new ideas emerge, at least partly, through new connections between things we already know, there is little reason to search for those connections only within the boundaries of our own expertise. Behavioral science has changed how I think about decisions. Poker has sharpened my view of uncertainty, outcomes, and luck. Urban planning has made me reconsider how environments shape behavior. More recently, while reading about Taoist thought, I came across an image that made me look at innovation differently.

That image was water.

I want to be careful with this because Taoism is not my field, and I am not trying to turn an ancient philosophical tradition into another innovation framework. I am simply borrowing one perspective and asking whether it helps me see something in my own domain more clearly.

Water appears prominently as a metaphor in the Dao De Jing. Scholarly treatments of Daoism describe water as an important model because of its capacity to yield, adapt to conditions, follow available paths, and avoid relying exclusively on direct force.[1]

What interests me is not the familiar interpretation of “go with the flow.” It is something more demanding.

Water does not respond to every obstacle in the same way. It can move around something, pass through an opening, accumulate behind a barrier, change direction, or gradually wear down what initially appears immovable. Its response depends on the terrain.

Innovation has a more complicated relationship with resistance. We often admire the person who pushes through it.

That admiration is understandable. If innovators changed direction every time an early customer disliked a prototype, an investor rejected a pitch, an incumbent dismissed a new approach, or adoption took longer than expected, many genuinely new things would never survive long enough to become useful. Innovation requires conviction, and conviction often means continuing in the presence of doubt.

I have explored that tension before in The Founder’s Forward Bias: the same ability to orient decisions toward a future that does not yet exist can create extraordinary conviction, but it can also make contradictory evidence easier to reinterpret than accept.

At the same time, resistance can contain information. The difficulty lies in knowing whether we are facing an obstacle that deserves persistence or a signal that should make us reconsider what we are doing.

That distinction has become increasingly important to me.

We just need to put another log on the fire

Years ago, I worked with a startup whose CEO had a phrase he repeatedly used when things were not developing as expected:

We just need to put another log on the fire.

I understood the reasoning behind it. New ventures need energy. Products rarely start out right. Markets take time to develop. Customers may need to experience something before they fully understand its value. Sales cycles can take longer than expected, and a viable business model rarely emerges fully formed from the first presentation.

There are moments when another log is exactly what the fire needs.

The problem was that, over time, almost every signal coming back from the market could be interpreted as another reason to add one. If customers did not respond, we needed more sales effort. If the proposition did not resonate, we needed better communication. If expected momentum failed to appear, we needed another iteration, another initiative, more time, or more money.

What gradually disappeared was the possibility that the resistance itself might be telling us something about one of our assumptions.

Eventually, this became the most expensive innovation failure I have personally witnessed in my career.

Looking back, I do not think the lesson is that somebody should simply have stopped earlier. That conclusion would be too easy, particularly with the benefit of hindsight.

Innovation happens under uncertainty. We make decisions with incomplete information, and the quality of a decision cannot be inferred cleanly from the outcome that follows. A good decision can produce a bad outcome, while a poor decision may occasionally benefit from timing, market conditions, or luck. I explored that distinction more deeply in Innovation Is Not Chess. It Is Poker, where the central argument is that innovation success depends on decision quality, learning, and adaptation without ever eliminating uncertainty.

Early evidence is often noisy, experiments are imperfect, and reality rarely provides an unambiguous signal telling us to continue or stop.

The more useful question is therefore not whether we should have shown more or less persistence in general.

It is this:

What evidence should tell us whether a particular form of resistance deserves more commitment?

Research on escalation of commitment makes that question more than a philosophical curiosity. Decades of work have examined why decision-makers sometimes continue investing in a course of action despite negative feedback, including the effects of previous investments, responsibility for earlier decisions, self-justification, and social or political pressure.[2]

That does not mean persistence after negative feedback is automatically irrational. Sometimes the underlying hypothesis remains sound and the obstacle lies elsewhere.

This is exactly why “persist or pivot?” is too crude a question.

Before deciding what to do, we first have to understand what exactly is resisting us.

Interested in working together?

Reach out

Resistance is not one thing

When an innovation does not develop as expected, we often compress very different situations into the same statement: it isn’t working.

But that tells us very little.

A customer may value the solution but face a painful switching cost. A user may struggle with an unfamiliar interface while still experiencing significant value after the learning curve. A buyer may want the product but be unable to get it through procurement. A promising technology may depend on infrastructure that is not yet ready. The person experiencing the value may not control the budget. The problem may be real while the chosen customer segment is wrong. The solution may work while the economics around it do not.

And there is another possibility: people simply do not care enough.

Those situations should not lead to the same decision.

I find it useful to distinguish several broad forms of resistance.

There can be discovery resistance, where we simply do not understand the problem or context well enough yet.

There can be adoption resistance, where customers must overcome habits, switching costs, uncertainty, learning, or perceived risk.

There can be value resistance, where the progress we offer is not important enough to justify changing behavior.

There can be system resistance, where regulation, procurement, infrastructure, channels, incentives, partners, or economics stand between potential value and actual adoption.

And there can be what I would call evidence resistance: reality repeatedly behaves differently from our hypothesis, while we become increasingly sophisticated at explaining why the evidence should not count.

That last form is particularly dangerous because it can look remarkably similar to conviction.

Follow me on LinkedIn

Be like water does not mean follow the easiest path

This is where I find the water metaphor especially useful.

“Be like water” does not mean choosing the easiest path, abandoning direction, or yielding at the first obstacle. Water can persist. It can stay. It can accumulate. It can eventually alter the terrain itself.

What it does not do is insist that every obstacle must be treated in exactly the same way.

For innovation, the equivalent is not to pivot whenever things become difficult. It is to understand the nature of the resistance before deciding whether to push through it, work around it, change the path, wait, or stop.

This is also where research on entrepreneurial experimentation becomes relevant. Camuffo and colleagues examined what happens when entrepreneurs are taught to articulate theories, formulate hypotheses, and test those hypotheses systematically rather than relying mainly on intuition.[3] The value of such an approach is not that it gives founders certainty. It makes their beliefs more explicit and therefore easier to revise when evidence changes.

A later large-scale replication across 759 firms found that this scientific approach influenced idea termination and produced more nuanced patterns of radical pivoting, reinforcing the value of disciplined belief updating rather than either unconditional persistence or repeated arbitrary change.[4]

Research on A/B testing among technology startups points in a related direction. Koning, Hasan, and Chatterji found that experimentation was associated with organizational learning and improvements in several startup performance measures, including increased product launches.[5]

None of this research tells an innovator exactly what to do with a specific venture tomorrow morning. What it does suggest is that the quality of persistence depends partly on whether we can turn our explanation for resistance into something testable.

“The market isn’t ready” may be true, but it is still a hypothesis. So are “customers need education,” “we need more awareness,” or “the product needs another six months.”

This is also the reasoning behind GPT Makes Ideas Cheap. Wrong Innovation Bets Remain Expensive.: a plausible explanation, sophisticated business case, or polished solution does not reduce uncertainty by itself. Before the next commitment, the more useful question is what would have to be true for the opportunity to work and what evidence could tell us whether those conditions actually hold.

Before adding another log, we should therefore be able to explain what we believe the additional commitment will change, why we believe that, and what evidence would eventually make us abandon that belief.

That leads me to four questions that I would carry through the entire innovation journey:

1. What did we expect to happen?

2. What actually happened?

3. What are the plausible explanations for the gap?

4. What evidence would distinguish those explanations before we increase our commitment?

If your team is deciding whether an innovation initiative deserves further funding, a pivot or a stop decision, my innovation strategy consulting provides an independent review of the assumptions, evidence and next commitment.

The free section of INNOVATION& is offered freely, and you are welcome to keep reading it that way. If these articles help you make better decisions in your working life, please consider becoming a paid subscriber. Researching and writing them takes time, and paid subscriptions provide the income I need to continue. Your support also helps keep part of this work available to everyone.

Subscribe now

Start paid learning and application section

Following resistance through the innovation journey

The nature of uncertainty changes as an innovation develops.

At the beginning, we may know very little about the problem or about who experiences it strongly enough to act. Later, uncertainty shifts toward the solution, behavior change, willingness to pay, channels, economics, dependencies, and eventually scalability.

The cost of being wrong changes as well. A mistaken assumption during an interview is relatively cheap. The same assumption embedded in a product, a sales force, infrastructure, or a multi-million-dollar investment is not.

The questions below are therefore not intended as another process. I see them as moments of interruption, especially when the instinct is to respond to weak evidence with more activity.

Read more

I am a big fan of Kitchen Impossible hosted by Tim Mälzer. Not because I love cooking, but because the program reveals what makes a chef’s expertise so impressive.

In each episode, accomplished chefs send one another to unfamiliar places, where the challenger receives a black box containing a local dish. They can taste, smell, inspect, and take it apart, but they are not given the recipe or a list of ingredients. Working in an unfamiliar kitchen and under severe time pressure, they must determine how the dish was prepared, reconstruct it, and serve their version to people who know the original.

What fascinates me most is how differently the chefs respond to this uncertainty.

Some commit quickly to their first interpretation and begin preparing the complete dish. Others treat the meal inside the box as a system that must first be understood. They separate its components, form hypotheses about ingredients and techniques, and conduct small experiments before committing to the final version.

Even under time pressure, some chefs prepare several variations of a sauce, dough, filling, or cooking method alongside one another. They alter the acidity, temperature, proportions, seasoning, or preparation technique and compare what happens. Each variation helps them understand what a particular variable contributes to the result.

They are not simply searching for something that tastes good.

They are trying to understand causality.

That process tells me something important about expertise. Expertise is not only the ability to recognize the correct answer immediately. It is also the ability to recognize what remains “known unknown”, turn uncertainty into testable hypotheses, and design small actions that produce useful evidence.

This is where innovation work frequently goes wrong.

Instead of preparing small variations and learning what causes what, organizations make one large bet. They commit the budget, build the complete solution, prepare the launch, and invite a large audience to an expensive dinner. Only when the meal is served do they discover that what they prepared does not work, nobody particularly likes it, and the evening has become difficult to recover.

The problem is not necessarily that the first interpretation was foolish. In genuinely uncertain situations, the first interpretation will always be incomplete.

The problem begins when the team treats an untested interpretation as though it were already a reliable recipe.

Tim Mälzer did not know the seasoning, so he created variations

One episode illustrates this experimental mindset particularly well.

In the second episode of the 2016 season, Meta Hiltebrand sent Tim Mälzer to the Zerda Café in Paris. He had to reconstruct a fish couscous and a pigeon pastilla from the Maghreb culinary tradition. He struggled to identify the ingredients and had little confidence that he understood how the dishes had been seasoned.

Instead of selecting one seasoning mixture and committing the entire dish to it, he conducted a small test with several spice combinations. He used the variations to move closer to the flavor he had experienced in the black box. The result was not a perfect reconstruction, but the jury still awarded the first challenge an average of 4.9 points. (rtl.de)

What matters to me is not the score. It is what he did when his expertise met a situation it could not solve through immediate recognition.

He did not stop being an expert because he did not know the answer. His expertise became visible in how he responded to not knowing.

He turned one large, unclear problem into several smaller comparisons. Instead of asking only, “Which seasoning is correct?”, he created alternatives that allowed him to ask, “What changes when I alter this combination?”

That is a very different question.

The first demands an answer. The second creates questions and information.

The distinction is central to innovation. A team can spend months asking whether its complete concept is right. A more useful approach is often to separate the concept into assumptions and ask which small variation would make one of those assumptions more observable.

A chef can compare several spice mixtures. An innovation team might compare different customer groups, problem descriptions, levels of service, pricing structures, onboarding methods, or commitments required from a potential buyer.

The purpose is not to create endless alternatives. It is to make differences informative.

Follow me on Linkedin

We once turned our entire apartment building into a black-box kitchen

I once initiated a more playful version of this idea in the apartment building where I live.

Each apartment received the same collection of ingredients, which had been chosen randomly by different people. Every household then had to create a dish from them. The apartments on the lower floors prepared the appetizers, those in the middle prepared the main courses, and the apartments on the upper floors were responsible for dessert.

The evening gradually moved upward through the building as we ate our way from floor to floor.

The ingredients were not selected to make the task easy. They included things such as cornflakes and other items that would not naturally appear together in an obvious recipe. Nobody could simply retrieve a familiar dish and follow the instructions.

Everyone began from the same raw material, but each apartment interpreted the possibilities differently. One household emphasized texture, another sweetness, another contrast, and another found a way to transform an ingredient so that it no longer played its familiar role.

It was not a controlled scientific experiment. We did not standardize quantities, equipment, cooking skill, or preparation time closely enough to make reliable causal claims.

Yet it revealed something that innovation teams regularly forget: the ingredients do not determine the outcome by themselves.

Interpretation, sequencing, technique, proportion, substitution, and combination all matter. Two teams can begin with apparently identical resources and produce completely different outcomes because they have formed different hypotheses about what those resources can become.

The evening also showed why constraints can stimulate exploration and its fun at the same time. Nobody could retreat comfortably into the recipe they already knew. We had to examine what was available, understand the properties of unfamiliar combinations, and improvise.

The strange ingredients were not an inconvenience added to the experiment.

They were what made the experiment useful.

Following a recipe and reconstructing a black box are different kinds of work

Cooking offers a useful way to distinguish three situations that organizations often manage as though they were the same.

In the first situation, you possess a reliable recipe, the specified ingredients, familiar equipment, and a controlled environment. The outcome is never perfectly guaranteed, but the relevant causal relationships are sufficiently understood. If you follow the process competently, you can expect a recognizable result.

This resembles delivery work. The customer, problem, solution, business model, and operating process are understood well enough that the central challenge is execution.

In the second situation, part of the recipe remains known, but you must substitute one ingredient or alter one condition. Perhaps the prescribed fat is unavailable, the flour behaves differently, or the oven produces less predictable heat. The uncertainty is limited, and you can isolate it reasonably well.

This resembles an improvement initiative or a relatively bounded product experiment. The overall system is understood, but one component needs to be tested.

The third situation is the Kitchen Impossible situation. You do not possess the recipe, some ingredients are unfamiliar, the equipment may behave differently, and you have to infer the preparation method from the result in front of you. The only thing you own is your long term adapted and gained expertise.

This is exploration.

The problem is that organizations frequently manage the third situation as though it were the first. They create roadmaps, deadlines, resource plans, and performance targets as if the recipe already existed and the team merely needed to execute it correctly.

When the result disappoints, the first conclusion is therefore often that execution was weak.

Perhaps the team did not cook hard enough.

Adaptive expertise begins when the familiar routine is no longer sufficient

Research on expertise distinguishes between performing efficiently in familiar situations and adapting effectively when the task, method, or desired result is novel.

A systematic review by Bohle Carbonell and colleagues describes adaptive expertise as the ability to maintain high performance when tasks or working methods change. Compared with routine expertise, adaptive expertise is associated with more flexible problem solving, stronger metacognitive abilities, continuous learning, and an understanding of why a method works and under which conditions it should be modified [1].

That distinction helps explain what makes the experimental chefs so impressive.

A routine expert may execute a familiar recipe with extraordinary precision. An adaptive expert can enter an unfamiliar kitchen, encounter a result they cannot immediately explain, and use what they already know to construct a path toward understanding.

The difference is not that the adaptive expert lacks routines. Deep domain knowledge remains essential. Tim Mälzer could not perform the black-box challenge without decades of experience with ingredients, heat, texture, timing, and taste.

However, he does not treat that experience as proof that his first answer must be right. He uses it to generate better hypotheses and more informative tests.

This matters for innovation because experience can play two opposing roles. It can expand the range of patterns a person recognizes, or it can make the first familiar explanation feel too convincing.

Adaptive expertise keeps experience available without making it untouchable.

Large bundled bets produce outcomes but little understanding

Many corporate pilots and startup launches are designed like complete dishes.

The team selects a customer segment, defines a problem, builds a solution, sets a price, chooses a channel, prepares a message, and creates a delivery process. It then presents the entire combination to the market and waits to see what happens.

When the result is positive, everyone celebrates. When the result is negative, the arguments begin.

Was the customer wrong, or was the problem poorly described? Was the proposition unattractive, or did the customer simply distrust the company? Was the price too high, or was the expected value too low? Did the channel fail, did the sales team fail, or was the purchasing process impossible from the beginning?

The team has generated an outcome, but the design of the activity makes it difficult to understand what caused it.

Experimentation research has long treated experimentation as a fundamental innovation and problem-solving activity. Stefan Thomke’s study of product development showed that different experimentation strategies can produce substantial differences in development time and cost. In the technical setting he examined, an approach enabling many prototype iterations outperformed a less flexible approach, with a significant portion of the difference attributable to the experimentation strategy itself [2].

Later work by Thomke and David Bell framed product-development tests as information-generating activities and showed that the appropriate frequency and fidelity of testing depend on the cost of each test, the cost of redesign, and the relationship among sequential tests [3].

The managerial implication is not that every organization should run as many tests as possible.

It is that the sequence and design of tests determine how economically the organization learns.

A large pilot that mixes several assumptions may look more realistic, but it can produce less useful information than a series of smaller tests designed around specific uncertainties.

Realism and learning value are not always the same.

Small experiments should reduce uncertainty, not merely create activity

The language of experimentation has become common in innovation. Teams speak about MVPs, prototypes, pilots, tests, and validation.

Not everything described as an experiment actually functions as one.

An activity becomes informative when the team can explain what it believes, which variable or assumption the activity examines, what it expects to observe, and how different results would change the next decision.

Without those elements, a pilot may simply be a small implementation.

The scientific approach to entrepreneurship examined by Camuffo and colleagues is relevant here. In a randomized controlled trial involving 116 startups, entrepreneurs taught to formulate explicit theories and test hypotheses systematically made more precise decisions and were more willing to change direction when the evidence warranted it [4].

A later large-scale replication across 759 firms found that the approach increased idea termination and encouraged a selective pattern of strategic change rather than either refusing to change or pivoting constantly. The authors argue that a scientific approach can develop what they call methodic doubt: an awareness that alternative explanations may fit the situation better than the one currently favored [5].

That is close to what happens when an expert chef places several spice mixtures next to one another.

The chef is not admitting that everything is unknowable. They are making competing explanations observable.

A useful innovation experiment does the same.

Do not confuse one-variable experiments with simplistic thinking

There is an appealing simplicity in changing one factor while keeping everything else constant.

When possible, this helps isolate the effect of the variable and makes interpretation easier. If one sauce contains more acid than another and the remaining conditions are sufficiently similar, the difference in flavor becomes more informative.

Business systems, however, often contain interactions.

A price that works with one value proposition may fail with another. A sales channel that performs well for one customer group may be inappropriate for a second. A new onboarding process may create value only when the service model changes with it.

Experiment-design research therefore warns that varying one factor at a time can be inefficient and may fail to identify interactions. Designed or factorial experiments can examine several factors systematically and reveal how combinations behave, although they also require greater design and analytical discipline [6].

The lesson for innovation teams is not that every business experiment needs to become a statistically sophisticated factorial trial.

It is that the experiment should match the question.

When the team needs to understand the influence of one uncertain component, a simple comparison may be enough. When several components are likely to interact, the team may need a small matrix of deliberately chosen combinations rather than a sequence of unrelated changes.

“Try something and see what happens” is not yet an experimental strategy.

Twenty variations may still fail to reveal the missing ingredient

Experimentation is not magic, and Kitchen Impossible also provides a useful warning about its limits.

In another 2016 episode, Tim Raue sent Tim Mälzer to Hong Kong, where he had to prepare fish maw and sea cucumber. The critical element was the sauce. Mälzer knew that something important was missing and reportedly prepared about twenty versions while searching for the absent note.

None solved the problem.

He eventually stopped. The missing ingredient turned out to be Chinese ham, which was outside the set of possibilities he had been testing. (rtl.de)

This is as relevant to innovation as the successful spice experiment in Paris.

Repeated iteration does not guarantee learning when the hypothesis space is wrong. A team can produce twenty versions of the same value proposition while overlooking that it has selected the wrong customer. It can optimize a product interface while the real barrier sits in procurement. It can test more marketing messages when the underlying problem is not important enough to justify switching.

The quality of an experiment depends partly on the quality of the alternatives it considers.

The Hong Kong episode also shows the importance of stopping. After many variations, sensory fatigue reduced Mälzer’s ability to distinguish the results. Continuing to produce more versions was no longer increasing understanding.

Organizations face the same risk. When a sequence of experiments stops generating new information, persistence can become activity without learning.

A stopping rule is therefore part of experimental discipline.

The purpose of small failures is to prevent one large failure

Small experiments are sometimes dismissed because they appear less serious than building the complete solution.

A rough prototype can look unimpressive. A test with a small number of carefully chosen participants does not create the excitement of a large pilot. Comparing several unfinished alternatives can feel indecisive when leaders expect one clear recommendation.

Yet the modest appearance is part of the economic logic.

Sim Sitkin’s work on learning through failure argues that small, contained losses can support organizational learning and adaptation when they provide information without threatening the survival or identity of the organization [7].

The goal is not to celebrate failure in general. Large, ambiguous, repeated, or politically threatening failures often produce defensiveness rather than learning.

The goal is to design failures that remain small enough to examine.

A chef can afford for three tablespoons of sauce to taste wrong. It is more expensive to discover the same error after serving one hundred plates.

An innovation team can afford for a landing-page proposition, mock sales conversation, concierge service, or pricing test to fail. It becomes harder to respond honestly after the organization has built a platform, hired a team, announced a launch, and promised the results to senior management.

The experiment protects the organization from the scale of its own confidence.

Experiments do more than improve the idea

The immediate purpose of an experiment is usually described as testing an assumption or improving a decision.

There is another important result.

Experiments build expertise.

Each well-designed comparison helps the team understand how the system behaves. It learns which variables matter, which effects interact, where customers behave differently from what they say, and which signals deserve more weight.

The team becomes better not only at developing the current venture, but at recognizing uncertainty in the next one.

This is why black-box experiments should not be treated as isolated project activities delegated to a research specialist. They are a way of developing adaptive expertise across the venture team.

Salespeople learn to separate interest from commitment. Product teams learn to distinguish feature requests from underlying struggles. Sponsors learn to ask what an activity will make more knowable before approving it. Founders learn to remain confident in their capacity to navigate uncertainty without becoming certain that the first recipe is correct.

The experiment improves the dish.

More importantly, it improves the cook.

Innovation is not the expensive dinner

The visible output of innovation is often the complete solution: the product, service, platform, pilot, or business model.

The early work should produce something else first.

It should produce a better understanding of the ingredients, their interactions, the preparation method, and the conditions under which the result becomes valuable.

That is why the natural posture of an innovation team is not to demonstrate how confidently it can execute its first interpretation.

It is to find the smallest useful way to discover where that interpretation may be wrong before inviting everyone to dinner.

Kitchen Impossible is compelling because accomplished chefs are willing to be seen not knowing. They open the black box, create a theory, test it, taste the result, and revise what they believe.

Their expertise does not eliminate uncertainty.

It allows them to work intelligently inside it.

Innovation should do the same.

Before you approve the next pilot

The diagnosis is straightforward: large bundled experiments make it difficult to know what caused the result.

The harder question is what to do with a real initiative asking for another USD 180,000, another four months, or another pilot.

For paid subscribers, the Black Box Experiment Board below turns that decision into a sequence of assumptions, experiments, evidence thresholds and commitment decisions.

Read more

I recently had the opportunity to read a draft paper by , Steven L. Winton, and Srikanth P. Mudigonda titled “AI-Assisted Decision Making as a Sequential Judgment Process”.

One distinction in the paper stayed with me. The authors separate perceived expertise from confidence in a current judgment. Perceived expertise concerns whether people believe they possess the knowledge and capability required to make an effective decision in a particular context. Confidence in a judgment concerns whether they believe that the particular conclusion they have reached is correct [1].

The two are related, but they are not the same. I can believe that I am generally capable of navigating a difficult situation while remaining uncertain about the answer in front of me. I can also possess limited relevant expertise while feeling extremely confident that my current answer is right.

The draft examines this distinction in the context of AI-assisted decision-making. People do not receive an AI recommendation with an empty mind. They have already formed an initial judgment, assessed their own capabilities, and developed some degree of confidence in their answer. These self-assessments influence whether they reconsider their position, engage critically with the recommendation, or dismiss it before examining what it might contribute [1].

Research supporting this wider argument shows why calibration matters. People who overestimate their own performance can under-rely on correct AI advice, while people who underestimate themselves may place too much weight on external recommendations. The objective is therefore not maximum trust in AI or maximum human independence, but appropriate reliance: knowing when one’s own judgment deserves to prevail and when it should be reconsidered [2].

While reading the draft, however, I was not thinking only about AI.

I was thinking about founders and corporate innovation leaders. I was thinking about the venture teams I have worked with, the projects I have analyzed, and the many arguments I have had after returning from the market with information that did not support what the founder or the team wanted to believe.

Ventures operate in a forest that rewards bold trees

A founder or corporate innovation leader needs confidence long before the evidence is complete.

Founders must persuade investors to provide capital for a business that does not yet exist in a reliable form. They must convince talented people to leave secure positions, persuade potential customers to spend time on an unfinished proposition, and repeatedly explain why an uncertain future deserves attention now. Corporate innovation leaders face a different version of the same challenge. They must compete for executive sponsorship, defend budgets, create internal legitimacy, and communicate enough direction that a steering committee will continue protecting the initiative.

There is not much room in these environments for leaders who appear “weak”.

A founder who repeatedly emphasizes everything that remains unknown may struggle to raise capital. An innovation leader who openly presents every uncertainty may be judged as someone who lacks a strategy. Investors expect conviction, employees want direction, sponsors need a story they can repeat, and the wider entrepreneurial culture tends to celebrate people who appear to have recognized the future before everyone else.

I understand why bold egos emerge in this environment. Confidence is not merely a personality characteristic. It can become an occupational requirement.

I am a long term Capoeirista and remembered a capoeira master participating in an event with many other bold masters once said something to me that I have never forgotten. Translated loosely, it was:

It is difficult to remain a humble, learning tree in a forest of bold trees competing to appear the strongest, especially when everyone expects a forest to look that way.

That metaphor captures the pressure well. When every tree is trying to appear certain, strong, and unmovable, humility can be interpreted as weakness. Changing direction can look like failure. Admitting that an assumption was wrong can appear to disqualify someone from leadership.

The answer, however, is not to become the least confident tree in the forest. It is to redefine what strength means.

Follow me on LinkedIn

Confidence becomes dangerous when it changes its object

The problem does not begin when a leader believes in their own capability. That belief may be essential.

It begins when confidence in the ability to navigate uncertainty quietly becomes confidence that the current interpretation of the opportunity must be correct.

The leader starts with:

I believe I am capable of finding a way through this uncertainty.

That is healthy confidence in capability.

Over time, it can become:

I believe the way I have already chosen is the right one.

That is confidence in a particular judgment.

The distinction in Grawitch, Winton, and Mudigonda’s draft becomes important here. A founder may have valid reasons to believe in their resilience, intelligence, experience, network, or capacity to assemble a strong team. None of these qualities proves that the current customer, problem, solution, price, channel, timing, or business model is right [1].

Confidence in the leader is not evidence for the venture hypothesis.

Studies of AI-assisted decisions show a related pattern. In research involving logical reasoning tasks, participants who overestimated their own competence tended to rely less on AI advice, which could prevent them from benefiting when the system was right [2]. Research involving chess players similarly found that greater expertise was associated with stronger self-confidence, lower trust in the AI system, and more self-reliance. That critical stance can protect people from incorrect recommendations, but it can also make them less receptive to correct advice [3].

The same tension can appear in venture leadership without AI being involved. Leaders who perceive themselves as capable may engage critically with new information, which is useful. They may also become so confident in their current interpretation that contradictory evidence no longer receives a fair hearing.

The issue is not confidence alone. It is whether confidence remains calibrated by what the market is teaching the team.

I went looking for resonance and often returned with conflict

I have experienced this more times than I can count.

In several startups and corporate innovation projects, part of my role was to go outside the building and look for resonance with the original belief behind the venture. I spoke with potential users, buyers, partners, industry participants, and other relevant stakeholders. I tried to understand whether the struggle described inside the venture was also visible outside it, whether it mattered enough to influence behavior, and whether the proposed business had commercial ground beyond the enthusiasm of the team.

Sometimes I returned with encouraging evidence. The problem was recognizable, people had already tried to solve it, and the current alternatives created meaningful cost or friction.

At other times, what I heard was less comfortable. The problem was not as important as we had assumed. The urgency was missing. The user experiencing the problem did not control the budget. The switching costs were greater than the value we could offer. The market liked the idea in principle but was unwilling to change behavior, allocate money, or accept implementation risk.

I brought those findings back because I understood that to be the work.

I also brought back questions, because field evidence is rarely a clean verdict. It needs to be interpreted, compared, and challenged. A few customer conversations do not prove a market, but they can reveal that the assumptions carrying a venture are weaker than the team believes.

I cannot remember how many internal arguments followed.

Instead of discussing the quality of the evidence, the participants we had selected, or the implications for the business model, the conversation often shifted toward my commitment to the idea. I was told, directly or indirectly, that I did not believe strongly enough in the venture.

My response was usually that this was not about whether I liked the idea. It was about whether facts outside the room resonated with the assumptions inside it.

Anyone can have ideas. Generative AI has made producing more of them easier than ever. What remains scarce is evidence that one particular idea deserves years of human life, investor capital, organizational credibility, and the opportunity cost of not pursuing something better.

Nevertheless, once an idea has become connected to the identity of the leader and the team, evidence against it can feel like disloyalty. The person returning from the market is no longer treated as someone helping the venture learn. They become the person who does not believe.

That is a critical moment because the social meaning of evidence has changed. A market signal is no longer simply information about an uncertain hypothesis. It has become a test of membership and commitment.

A team can argue constantly and still protect the important assumptions

Venture teams often describe themselves as open, direct, and willing to debate. Many of them are.

They may argue intensely about features, technology choices, brand positioning, campaigns, pricing presentations, sales processes, and delivery priorities. This can create a convincing appearance of intellectual openness.

The more fundamental assumptions may nevertheless remain protected.

Does the customer struggle matter enough to cause action? Are we working with the right segment? Does the person experiencing the problem have access to a budget? Is the current workaround genuinely inadequate, or merely inelegant? Will the customer accept the cost, risk, and politics of switching? Does this venture deserve to continue?

A team can debate everything downstream while treating these upstream assumptions as settled.

Research on psychological safety offers a useful lens. Amy Edmondson defined team psychological safety as a shared belief that a team is safe for interpersonal risk-taking. In her study of 51 work teams, psychological safety was associated with learning behavior such as discussing errors, asking for help, seeking feedback, and experimenting [5].

Psychological safety in a venture should therefore not be assessed only by asking whether colleagues are friendly or whether meetings feel inclusive. The more consequential test is whether someone can introduce evidence that threatens the strategy, the leader’s judgment, or the venture’s shared identity without paying a social price.

Can someone say, “The market is not supporting what we currently believe,” and remain a trusted member of the team?

When the answer is no, the venture may possess a bold culture, but it does not possess a strong learning culture.

The execution-defense loop

One pattern appeared repeatedly in the teams I observed.

When market results failed to support the original expectations, the explanation moved toward execution.

The product was right, but sales could not sell it. The market existed, but marketing had failed to communicate the value clearly. Customers needed the solution, but they were not sufficiently educated. The strategy was sound, but the organization was moving too slowly. The pilot would have succeeded if the business unit had supported it more strongly. Adoption would arrive after another feature, a sharper pitch, a different salesperson, or a more aggressive campaign.

Any of these explanations can be correct.

Innovative products can be sold badly. Marketing can obscure a valuable proposition. Customers may need help understanding a genuinely new category. Internal politics can destroy an otherwise promising corporate venture. Poor execution is real.

The warning sign appears when every contradictory market signal becomes an execution problem.

When positive results demonstrate the brilliance of the idea and negative results always demonstrate the weakness of sales, marketing, product, or operations, the original judgment has become impossible to falsify. Whatever happens, the idea remains right.

This is more than a cognitive bias. It becomes an organizational allocation mechanism.

If the leader’s interpretation is protected, the search for the cause moves downward. Salespeople receive more pressure, marketers are asked to improve the message, and product teams are told to build more. The venture becomes increasingly active while the assumption that most deserves investigation remains untouched.

People inside the organization learn from this. Bringing contradictory evidence creates conflict. Promising better execution attracts support.

The team begins optimizing for internal acceptance instead of external truth.

A learning tree is not a weak tree

The tree metaphor matters because the alternative to performative certainty is often misunderstood.

A learning tree is not passive, hesitant, or without direction. It remains highly sensitive to its environment. It responds to changes in the soil, redirects its growth when conditions change, and adapts before a less attentive tree recognizes what is happening.

Its strength comes from responsiveness rather than rigidity.

A strong venture leader should create the same capability. The leader does not have to appear certain about every answer. The leader must build confidence that the team can identify a better answer, recognize when the current one is failing, and change without treating learning as a collapse of identity.

The shift is not from confidence to doubt. It is from individual certainty to collective learning capability.

The leader moves from:

I believe I am capable of finding a way through this uncertainty.

to:

We are building the fastest-learning team in our ecosystem. We do not need to be right from the beginning. We need to understand what is true faster than anyone else.

Our advantage is not that we claim to be right first. Our advantage is that we discover what is true, what is false, and what remains uncertain before competitors, incumbents, or our own internal politics prevent us from responding.

A venture should be the fastest-learning entity in its ecosystem

An established company is primarily organized to execute what it already knows.

It has existing customers, products, channels, processes, capabilities, and sources of revenue. It still needs to learn and adapt, but much of its operating model is designed to repeat known activities reliably and efficiently.

A venture has a different purpose because the essential answers are not yet known. The customer may be wrong, the struggle may be misunderstood, the solution may not create enough value, the buying process may be inaccessible, and the economics may never support a viable business.

The venture should therefore not be understood primarily as a small company waiting to grow.

It is a temporary learning system created to discover whether a company deserves to exist.

Its job is to convert consequential unknowns into decision-grade knowledge before capital, credibility, and human energy become locked into the wrong direction. This does not mean removing all uncertainty. Some bets cannot be proven before they are made. It means identifying what can be learned, reducing the uncertainties that carry the greatest cost, and making the remaining bet explicit.

This changes the meaning of early productivity.

A prototype is useful when it helps the team examine an important assumption. An interview is useful when the team is willing to hear an answer that contradicts its expectations. A pilot is useful when it creates evidence that can change an investment decision. A roadmap is useful when it organizes learning rather than converting assumptions into promises.

The venture’s early outputs should therefore include more than products, decks, and demonstrations. They should include explicit assumptions, observations from outside the team, evidence that weakened the original view, unresolved unknowns, and thresholds for continuing, revising, deferring, or stopping.

These are not signs that the team lacks belief.

They show that the team has built a functioning learning system.

Fast learning does not mean reacting to every opinion

The fastest-learning team is not the team that changes direction after every conversation.

That would not be learning. It would be noise-driven movement.

A customer can misunderstand the proposition. A participant can be outside the intended segment. A potential buyer can reject a genuinely valuable idea because the timing is wrong. Evidence varies in strength, relevance, and representativeness, and leaders still need judgment.

Disciplined learning requires the team to make its assumptions explicit, design credible ways to examine them, evaluate the evidence carefully, and update its decisions proportionately.

Research on scientific approaches to entrepreneurship supports this distinction. In a randomized controlled trial involving 116 Italian startups, Camuffo and colleagues found that entrepreneurs trained to articulate theories and test hypotheses systematically performed better and were more likely to pivot when the evidence warranted it [6]. A later replication involving 759 firms across four randomized controlled trials found that the approach increased idea termination and encouraged a more selective pattern of strategic change rather than either never changing or pivoting continuously [7].

The implication is not that ventures should imitate laboratories or that every commercial question can be settled experimentally. It is that confidence becomes more useful when assumptions are visible, alternative explanations remain possible, and evidence is allowed to change the decision.

The fastest-learning venture is not the one that changes most often. It is the one that becomes less wrong without wasting unnecessary time and capital.

Confidence should belong to the learning system

A venture still needs confidence, but that confidence should not be concentrated in the leader’s current answer.

It should belong to the learning system.

The team should be confident that it can identify its most consequential assumptions, reach people who know more than those inside the room, distinguish polite interest from meaningful behavior, and respond when reality contradicts the strategy.

A strong leader can say:

This is our current direction. This is the evidence supporting it. This is what remains uncertain. This is what would change our mind, and this is how much we are prepared to invest before we know more.

That is not weak communication. It is a more demanding form of leadership than pretending the answer has already been earned.

It gives the team direction without protecting that direction from evidence. It gives sponsors a basis for deciding what the next budget should purchase. It also allows sales, marketing, product, and research teams to return with uncomfortable findings without being accused of lacking faith.

The team no longer has to believe in every assumption. It has to believe in its collective capacity to learn.

Sponsors create the confidence they reward

Investors, executives, and corporate venture sponsors also shape this behavior.

When they reward only certainty, teams learn to perform certainty. When they ask only for visible progress, teams present activity. When every change of direction is interpreted as evidence that the original team was wrong or weak, leaders delay admitting what they have learned until adaptation becomes expensive.

A more intelligent funding system asks what the venture believed at the beginning of the current funding period and what it understands differently now. It asks which assumptions became stronger, which weakened, and which remain unresolved. It examines what evidence arrived from outside the team and what decision will become possible after the next investment.

Budgets should be released in increments that match the maturity of the evidence.

At the beginning, a modest budget may be justified to investigate whether a sufficiently important and reachable customer struggle exists. The next increment may examine willingness to switch, budget ownership, pricing, and the buying process. Larger commitments (especially building solutions) become reasonable as the venture earns stronger evidence about demand, delivery, economics, retention, and scale.

This is not a demand for certainty before action. It is a requirement that each increase in commitment be connected to an increase in decision-grade understanding.

The venture is not funded to make the original leader look right.

It is funded to prevent the organization from remaining wrong for too long.

Our advantage is learning what is true faster

The draft that triggered this article concerns AI-assisted decision-making, not venture leadership. My extension from individual self-assessment to startup and corporate venture behavior is therefore an interpretation based on my own experience, not a conclusion the authors claim to have tested.

Yet the distinction between perceived expertise and confidence in a current judgment explains something I have repeatedly observed.

Venture leaders need confidence in their capability. They require enough resilience and ego to withstand rejection, mobilize other people, and move through environments in which the path is not obvious.

That confidence becomes dangerous when it is transferred to an untested interpretation of the market.

Once the leader’s current judgment becomes part of the team’s identity, market evidence can become a loyalty test. People who return with weak signals are accused of not believing. Sales and marketing are blamed for failing to execute a strategy that nobody is permitted to question. The venture becomes increasingly organized around defending the answer it was originally created to investigate.

The alternative is not timid leadership. It is a more ambitious form of confidence:

We are building the fastest-learning team in our ecosystem. We do not need to be right from the beginning. We want to understand what is true faster than anyone else.

The strongest venture leader is not the person who never doubts the idea. It is the person who makes it possible for the entire team to discover that the idea may be wrong without turning learning into betrayal.

We do not need to be the boldest-looking tree in the forest.

We need to understand the forest fastest.

Paid learning and application: The Fastest-Learning Venture Review

The argument becomes valuable when it changes a real investment decision. The Fastest-Learning Venture Review is designed for a startup or corporate venture requesting another meaningful commitment of money, time, people, or organizational credibility.

Its purpose is not to assess whether the leader appears confident enough. It is to determine whether confidence is attached to the team’s capacity to learn or to a current judgment that the evidence has not yet earned.

Read more

I am currently using GenAI to rethink the SEO and GEO strategy for INNOVATION&.

I would not describe myself as an expert in either field. I know enough to understand the basic mechanisms, follow an argument, question recommendations, compare alternatives, and recognise advice that is too generic to be useful. However, I do not possess the depth of experience that would allow me to judge every technical claim independently or immediately recognize every important omission.

This is precisely why GenAI is so valuable to me.

What would previously have required weeks of research, conversations with several specialists, or the hiring of external expertise can now begin with a dialogue. I can ask the model to assess my current strategy, explain unfamiliar concepts, compare different approaches, identify contradictions, propose priorities, and translate all of this into concrete changes. I can then challenge the result, add context, reject recommendations, request evidence, introduce another perspective, and revise the strategy again.

I rarely accept the first answer. The process is not a single prompt followed by passive acceptance. It is an extended conversation in which I repeatedly question what I receive. Sometimes we make clear progress. Sometimes a later response exposes a weakness in an earlier one. Sometimes we move backwards because a recommendation that initially seemed convincing no longer survives closer examination. Occasionally, we move in circles and return to an idea that had already been rejected.

The result of this process often looks impressive. At least, it looks impressive to me.

That qualification matters because it leads to a question that extends far beyond SEO or GEO.

Am I using GenAI to reason, reflect, judge, and decide more effectively, or am I gradually becoming biased towards its output because I lack enough expertise to recognize where the reasoning is incomplete?

The longer the conversation continues, the more difficult this distinction becomes. Each iteration makes the output more specific to my situation. The model adopts my terminology, incorporates my objections, and increasingly reflects the way I frame the problem. The result therefore feels less like a generic machine response and more like a jointly developed strategy.

Perhaps that means the strategy is improving.

Perhaps I am simply becoming more invested in it.

GenAI can raise the floor, raise the ceiling, or conceal the gap

Most discussions about GenAI at work begin with productivity. Can it help people write faster, analyze more information, produce more ideas, or perform tasks that previously required specialist support?

There is growing evidence that it can improve immediate task performance in several domains, although the magnitude and direction of the effect depend heavily on the task, the user, and the way performance is measured. The more important question for innovation work, however, is not simply whether the output improves. It is what happens to the expertise and judgement behind that output.

GenAI can influence expertise in at least three different ways.

It can raise the floor by helping people with limited experience produce work that is more complete, structured, or credible than what they could produce unaided. A non-specialist can create a reasonable first market analysis, a more coherent interview guide, a clearer opportunity statement, or a better-structured experiment.

It can also raise the ceiling by allowing experienced practitioners to explore more alternatives, examine more evidence, pressure-test assumptions, and make connections they may not have reached alone. In this mode, the model does not substitute for expertise. It expands the area over which expertise can operate.

There is also a third possibility. GenAI can raise the apparent quality of the work without raising the user’s ability to judge it by the same amount. The output becomes more sophisticated, while the person evaluating it remains dependent on the same system that produced it.

This third possibility is what I call B(AI)AS.

B(AI)AS is not an established scientific term. It is my name for a possible judgement gap: the distance between the apparent quality of an AI-assisted output and the user’s independent ability to evaluate whether that output is correct, complete, relevant, and appropriate for the decision at hand.

The risk is not limited to people who accept the first response without thinking. It can also emerge through long and apparently critical interactions. A user may challenge the model repeatedly and still become increasingly dependent on its framing, evidence selection, and interpretation.

The output improves. Confidence increases. Yet the underlying judgement capability may remain unchanged.

Follow me on LinkedIn

Why fluent reasoning feels like reliable reasoning

GenAI is persuasive in a way that previous decision-support systems were not. It does not merely produce a recommendation. It can explain the recommendation, defend it, criticize it, compare it with alternatives, acknowledge uncertainty, and reformulate the conclusion after receiving new information.

This creates the experience of reasoning.

The conversation may contain disagreement, revision, qualification, and reflection. These are normally signals we associate with thoughtful human deliberation. When the model responds to objections and changes its conclusion, it can appear as though the original proposition has survived a serious examination.

However, a conversation can become more coherent without becoming more accurate. The same system may generate the proposition, the counterarguments, the evaluation criteria, and the final conclusion. It can simulate opposition without introducing genuinely independent evidence.

The wording becomes sharper. Contradictions disappear. The argument becomes easier to follow. None of this proves that the underlying claims are correct.

Research on automation bias provides part of the scientific basis for this concern. Studies conducted long before the current generation of language models showed that people sometimes follow automated recommendations despite contradictory information, particularly when they perceive the system as competent or when independent verification requires additional effort [1]. More recent research has reframed the challenge as one of appropriate reliance: people need to know when AI advice should be accepted, rejected, or verified rather than simply learning to trust it more or less [2].

Generative AI complicates appropriate reliance because the recommendation is delivered through dialogue. The system appears responsive to the user’s situation and can provide an explanation for almost every answer. This responsiveness can increase the perceived legitimacy of the output even when the explanation does not provide new evidence.

Research also suggests that explanations do not automatically solve this problem. Depending on how they are designed and used, explanations can help people identify errors, but they can also increase reliance by making an incorrect recommendation appear more understandable [2]. An explanation should therefore not be confused with verification.

The capacity to explain an answer is one of GenAI’s most useful characteristics. It is also one of the reasons why an unsupported answer can become unusually convincing.

Iteration is better than accepting the first answer, but it is not independent validation

A common response to concerns about GenAI is that people need to learn better prompting. They should reject the first answer, request alternatives, ask the model to identify its assumptions, and instruct it to argue against its own conclusion.

This is sound practical advice. It is still incomplete.

Iteration can produce genuine improvement when the user contributes new information, corrects the model, introduces external evidence, changes the framing, and applies independent evaluation criteria. In that situation, the conversation is not merely generating more text. It is helping the user examine a larger space of possibilities.

Iteration can also produce what I would call iterative dependence. The user remains active, but the model continues to supply most of the hypotheses, objections, alternatives, and evaluation criteria. The interaction becomes more detailed and sophisticated, yet the epistemic loop remains largely closed.

This form of dependence is difficult to recognize because it does not feel lazy. The user may spend hours working with the model. Several alternatives may be compared, multiple objections considered, and numerous revisions made. The final output may contain assumptions, scenarios, risks, and a convincing decision matrix.

It looks as though the conclusion has been earned.

However, the amount of work invested in refining an argument is not evidence that the argument is better. Repeatedly asking the same underlying system to reconsider its answer does not create the same degree of independence as introducing customer observations, operational data, external research, domain expertise, or evidence that could contradict the conclusion.

Research on the metacognitive demands of GenAI helps explain why this matters. Metacognition refers broadly to the ability to monitor and regulate one’s own thinking. Tankelevitch and colleagues argue that GenAI places substantial metacognitive demands on users because they must define the task, monitor the quality of the interaction, evaluate outputs, and decide when further investigation is necessary [3].

The tool may reduce the cognitive effort required to generate material, while increasing the importance of monitoring and evaluation. This becomes particularly difficult when the user lacks the domain knowledge needed to evaluate the output. To identify an omission, one must usually know enough to understand that something is missing.

A study of 319 knowledge workers found that greater confidence in GenAI was associated with lower self-reported critical-thinking effort, while greater confidence in one’s own ability was associated with more critical evaluation [4]. The authors did not conclude that GenAI simply eliminates critical thinking. Their findings suggest that it changes where critical thinking occurs. Less effort may be required to produce a first output, while more effort is needed to verify, integrate, and govern what the system produces.

This is the central paradox behind B(AI)AS. The less expertise we possess, the more GenAI can raise our floor. Yet the less expertise we possess, the less capable we may be of recognizing where the output is wrong.

The problem becomes more consequential in innovation work

This judgement gap matters in many forms of knowledge work, but it becomes particularly expensive in innovation because the outputs are used to make decisions under uncertainty.

Innovation teams increasingly use GenAI to identify opportunity areas, synthesize interviews, formulate customer jobs, create personas, analyze markets, generate business models, design experiments, interpret evidence, and prepare investment cases. None of these activities is inherently inappropriate. Many can benefit substantially from AI assistance.

They do not, however, carry the same decision risk.

Using GenAI to improve the wording of a workshop invitation is different from using it to conclude that a customer problem is important. Generating an initial list of competitors is different from determining that a market is attractive. Drafting an experiment is different from deciding that a critical assumption has been validated.

The closer an AI-assisted output moves towards an irreversible or expensive commitment, the more important independent judgement becomes.

Consider a team evaluating a new opportunity. GenAI can identify relevant trends, propose customer struggles, formulate market segments, generate value propositions, describe adoption barriers, compare competitors, and recommend experiments. Within an afternoon, the team can produce a document that appears more comprehensive than what it might previously have created in several weeks.

The question is what the team has actually learned.

The model may have organized existing information, revealed connections, or introduced plausible hypotheses. It may have improved the team’s preparation for customer research and helped the team articulate what it needs to learn.

It has not observed a customer experiencing the problem.

It has not seen how people currently improvise around an inadequate solution. It has not watched a purchasing decision stall, identified who controls the budget, or established which constraint prevents switching. It has not demonstrated that the proposed problem is important enough and underserved enough to change behavior.

GenAI can make a hypothesis more articulate. It cannot turn that hypothesis into empirical evidence.

Yet articulation matters psychologically. A hypothesis expressed in professional language, supported by a coherent causal story, and embedded in a polished framework is easier to mistake for an informed conclusion.

This distinction is particularly important when AI is used to synthesize customer interviews. A model can identify themes across transcripts much faster than a human team. It can also flatten contradictions, remove contextual details, overstate weak patterns, or organize responses around categories introduced by the prompt. The resulting synthesis may appear clearer than the underlying evidence actually is.

The risk is not that GenAI always produces a poor synthesis. The risk is that the team loses visibility into the interpretive decisions through which the synthesis was produced.

Raising the floor does not necessarily build expertise

Experimental research provides credible evidence that GenAI can improve some forms of individual performance. Doshi and Hauser, for example, found that access to AI-generated ideas improved how short stories were evaluated. The stories were judged to be more creative, better written, and more enjoyable, with particularly strong benefits for participants who initially scored lower on creativity [5].

This is a clear example of the expertise floor being raised.

The same study also found that AI-assisted stories became more similar to one another [5]. Individual outputs improved, while the collective diversity of outputs declined. Other research into AI-supported ideation has reported related homogenization effects, suggesting that users can converge towards similar semantic territory when drawing on similar model-generated suggestions [6].

For innovation work, this creates an important distinction between the quality of an individual idea and the breadth of the search process.

A team may generate ideas that are more detailed, convincing, and professionally presented than before. At the same time, the overall set of ideas may become less varied. Each individual proposal appears stronger, while the organization explores a narrower opportunity space.

This pattern would be easy to overlook if each proposal were evaluated separately. The team sees that the ideas have improved. It does not see the alternatives that were never generated because multiple participants started from similar AI-produced anchors.

The technology may therefore raise the floor while narrowing the field.

That does not mean AI-generated ideas should be avoided. It means that teams need deliberate mechanisms for preserving variation, introducing independent starting points, and looking beyond the first plausible categories offered by the model.

The first answer does not have to be accepted to become influential. It can act as an anchor around which every later answer is organized.

Expertise changes the nature of the collaboration

A novice and an expert can use the same model, enter similar prompts, and receive outputs of comparable apparent quality. The value they derive from the interaction may nevertheless be fundamentally different.

A novice often uses GenAI to supply missing knowledge. An expert is more able to use it to interrogate existing knowledge.

The novice may ask what the innovation strategy should be. The expert is more likely to examine which assumptions the proposed strategy depends on, what evidence would contradict them, which alternatives have been excluded, and how the recommendation changes under different constraints.

This difference is often described as prompting skill, but prompting is only part of it. Experts possess reference points outside the conversation. They recognize patterns, but they also understand when a familiar pattern has been transferred into the wrong context. They can identify variables that have been omitted, causal claims that are unsupported, or recommendations that remain generic despite appearing highly specific.

Most importantly, they can reject a convincing answer for reasons that the model was never asked to consider.

The phrase “human in the loop” therefore provides little reassurance by itself. A human can be present while contributing almost no independent judgement. The relevant question is whether the person reviewing the output has the knowledge, incentives, time, and authority required to recognize when it should not be trusted.

Research on human–AI collaboration supports this more cautious view. A systematic review and meta-analysis by Vaccaro, Almaatouq, and Malone found that human–AI combinations did not, on average, outperform the better of the human or AI working alone. The combined performance was significantly worse than the best individual performer, with more favorable results in content-creation tasks and greater performance losses in decision tasks [7].

This finding does not show that human–AI collaboration is generally ineffective. It shows that complementarity should not be assumed.

For a combined system to outperform its individual parts, the human must understand when their own judgement is more reliable, when the model is more reliable, and how disagreement should be resolved. Without this calibration, the combination can inherit the weaknesses of both.

Innovation work makes such calibration especially difficult because feedback is often delayed and ambiguous. If AI misclassifies an image, the error may be detected quickly. If it helps a team misdiagnose why customers are not switching, the consequences may unfold over months.

The team may build experiments around the wrong assumption, interpret ambiguous evidence as support, and increase its commitment before reconsidering the original diagnosis. By the time the error becomes visible, capital, credibility, and organisational energy may already be locked in.

The cost of B(AI)AS is therefore not merely one incorrect answer. It is a sequence of reasonable-looking decisions built on an answer that was never independently established.

From first-answer dependence to expert expansion

The useful distinction is not between using GenAI and avoiding it. It is between different modes of reliance.

At one end is first-answer dependence, in which the user accepts the first plausible output with little challenge or verification.

A more sophisticated form is iterative dependence. Here, the user questions and refines the answer, but the model continues to provide most of the reasoning, alternatives, and criteria through which its own work is assessed.

The next mode is informed augmentation. The user brings external evidence, defines evaluation criteria independently, understands important limitations, and has enough knowledge to reject the model’s framing.

At the other end is expert expansion, in which GenAI allows a capable practitioner to examine more possibilities, connect more information, and test more interpretations than would otherwise be practical.

These modes cannot be distinguished by looking at the final output.

Two market analyses may appear equally sophisticated. One may be the result of extensive domain knowledge expanded through GenAI. The other may be the result of a non-expert spending several hours refining the model’s assumptions.

This is why the apparent quality of an output is becoming a weaker indicator of the capability behind it.

GenAI can improve the visible work faster than it improves the user’s ability to judge that work. It may make limited expertise look like deep expertise, both to the reader and to the person who produced the output.

The central question is therefore no longer simply whether GenAI improved the answer.

It is whether our ability to evaluate the answer improved at the same rate.

Paid learning and application section

Before you act on an AI-generated recommendation

Bring one market analysis, customer-research synthesis or investment recommendation you are currently using. In the paid lesson, you will work through the B(AI)AS Check to separate supported claims from assumptions, identify what requires independent verification, and decide whether to use, verify, test or stop the recommendation.

The aim is a clear next step for your own decision.

Read more

The idea for this article did not come from a startup pitch, a board meeting, or a strategy workshop. It came while I was standing in front of a supermarket shelf and noticed that one price label had been placed directly over another.

I became curious about what was underneath. Had the price increased or decreased? Had someone simply found it easier to cover the old label than remove it? Or had the retailer deliberately changed the price to observe what would happen?

I had no way of knowing. The explanation may have been entirely mundane. Yet the label brought back a discussion I had about two years earlier with an experienced senior consultant.

The discussion followed a post in which I had argued that every element of a startup’s business model begins as an assumption and should therefore be treated as an untested hypothesis to be tested against the market. The customer problem, target segment, solution, channel, revenue model, partnerships, cost structure, and pricing model may all appear reasonable, but none begins as established knowledge.

The consultant challenged that statement. Not every element of a business model could be tested before launch, he argued, and pricing was one example. Eventually, founders had to rely on experience, judgement, and their best available estimate.

We agreed to disagree.

I remember briefly questioning myself afterwards. Who was I to insist on my position when the person opposite me had spent many years advising companies?

Yet the argument stayed with me because it contradicted how I had worked throughout my career. In startups I co-founded, corporate ventures I worked for, and companies that later approached me for pricing advice, I had repeatedly experimented with economic assumptions rather than treating them as fixed decisions.

Seeing those two price labels brought the disagreement back to the surface. It also made me realize that my own use of the sentence “validated hypothesis” from Lean Startup needed clarification.

Validation is not certainty

When I said that every part of a business model should be validated in the past, I did not mean that founders can prove the correct pricing model, package, and price point before entering the market.

Nor do I believe that a series of early experiments can predict how an entire market will behave once a product is widely available.

What founders can do is make more of their unknowns visible, challenge what they currently believe, and reduce uncertainty before making commitments that are difficult and expensive to reverse.

This is also how business model experimentation is described in the research literature: as an iterative process through which organizations formulate and test assumptions, engage stakeholders, learn from evidence, and progressively narrow the range of viable business model choices.¹ ²

The purpose is not to know everything before acting. It is to avoid entering the market with critical assumptions that have never been made explicit, investigated, or exposed to contradictory evidence.

If validation required 100% certainty, almost nothing in innovation could ever be validated. A startup cannot conclusively prove product-market fit before the product is used more broadly. It cannot know its eventual retention, customer lifetime value, acquisition costs, support burden, competitive response, or contribution margin.

These variables continue to develop after launch. Research on business model innovation similarly shows that business models frequently emerge through trial-and-error learning rather than being designed correctly at the beginning and then executed without further revision.³

The same is true of pricing.

That does not make experimentation pointless. It makes experimentation necessary.

The objective is not to eliminate uncertainty. It is to reduce avoidable uncertainty before the size of the commitment increases and the consequences of being wrong become much larger.

Pricing is not a number

One reason pricing is often treated as difficult to test is that it is reduced to a single question:

Should the product cost 49, 99, or 149?

In reality, pricing is not one hypothesis. It is an interconnected system of economic hypotheses.

Who pays for the product may be different from who uses it. Customers may pay for access, usage, time saved, risk reduced, transactions completed, capacity provided, or outcomes achieved. They may pay once, monthly, annually, per user, per project, or according to consumption.

Payment may happen through a credit card, an invoice, a purchase order, a reseller, or an enterprise procurement process that lasts several months. The product may be sold directly, through partners, or through a combination of both.

Some customers may require onboarding, integration, training, implementation, and ongoing support before they can realize any value.

Each of these choices affects the economics of the business.

A subscription that looks attractive when customers are expected to buy through a self-service website may become unsustainable when the company discovers that every customer requires several sales conversations, a proof of concept, implementation support, and a dedicated account manager.

The price itself may not have changed, but the cost of producing the revenue has.

This distinction matters because financial viability depends not only on what customers pay but on the complete system required to acquire, serve, and retain them.

A company must create enough revenue and contribution margin, at a sufficient rate, to cover the resources required to operate the business. Without that, it will not reach break-even. It may also struggle to attract the additional capital, knowledge, employees, partners, credibility, and ecosystem support required to continue.

Pricing is therefore not a secondary implementation detail. It is one of the venture’s most consequential economic hypotheses.

The problem is often not the price

Across startups I have co-founded, corporate ventures I have worked for, and companies that have approached me for advice, I have seen several versions of the same mistake.

A pricing model was selected early, often by copying familiar industry conventions or looking at competitors. It was then treated as part of the solution rather than as an assumption requiring investigation.

In one case, the company expected indirect sales through partners to provide an efficient route to market. The assumption seemed reasonable until reality showed that customers required more education, demonstrations, commercial guidance, onboarding, and continued support than expected.

The company had to hire its own people to perform work that the original business model had assigned to partners. The subscription price remained largely unchanged, but the cost structure did not. What had looked scalable on a spreadsheet became a structural margin problem.

In another situation, a company expected customers to adopt the product with little human involvement. In practice, implementation required consulting, integration, and ongoing support before customers achieved the promised outcome.

Again, the problem was not merely that the price was too low. The company had designed its pricing around a delivery system that did not exist in reality.

These companies did not merely choose the wrong price.

They committed to an untested economic system.

By the time the assumptions became visible, the businesses had already committed to products, customers, contracts, teams, and market expectations. Changing the model was no longer a small experiment. It risked customer resistance, internal disruption, and damage to credibility.

This is what makes the bet so expensive.

Launching does not merely expose a company to feedback. It creates commitments. Customers expect continuity. Employees depend on the model. Investors evaluate performance against it. Partners organize their activities around it. Processes, incentives, and systems make the chosen model progressively harder to change.

When an untested economic assumption fails after launch, the company is not simply updating a number on a pricing page. It may have to redesign how it sells, delivers, supports, and finances the entire business.

What pricing experiments can reveal

None of this means founders must discover a perfect price in advance. It means they should investigate their economic assumptions before treating them as fixed.

I have used several approaches throughout my working life.

I have mentioned possible prices during customer conversations and observed whether people rejected the amount, questioned what was included, or immediately began discussing budgets, procurement, and payment terms.

I have created mocked physical price lists and presented different packages to potential customers. I have used mocked online pricing pages to see how people interpreted the offers and which comparisons they made.

I have examined the customer’s total cost of switching because the price of a new solution is only one part of the decision. A customer may also face integration costs, retraining, operational disruption, and the political risk of replacing an established supplier.

I have compared the proposed price with the total cost of doing nothing. In many situations, the real competitor is not another product but the customer’s willingness to continue living with the current problem.

I have also compared the offer with alternative ways of getting the same job done and explored what customers would pay for a proof of concept before committing to a full implementation.

Pricing and willingness-to-pay assumptions can be investigated through controlled experiments, field experiments, choice tasks, quotations, and real market offers. Such experiments do not produce perfect foresight, but they can reveal demand responses, price sensitivity, buying constraints, and differences between alternative offers.⁴ ⁵

The important limitation is that not all experiments produce evidence of equal strength.

A conversation can reveal objections, reference points, purchasing logic, budget ownership, and procurement constraints. A mocked pricing page can reveal how customers understand different packages. A request for a formal quotation provides stronger evidence of organizational intent. A paid proof of concept is stronger again. A purchase, renewal, or expansion provides more consequential behavioral evidence.

This distinction matters because research repeatedly finds a gap between hypothetical willingness to pay and behavior under real economic consequences. What customers say they would pay does not always correspond with what they pay when their own money, budget, or organizational credibility is involved.⁶

The earlier experiments are not worthless because they are incomplete. Their purpose is to reveal assumptions and help the company decide what deserves a more demanding test.

The mistake is not using weak evidence.

The mistake is treating weak evidence as if it were strong.

From validated learning to falsifiable learning

This suggests a useful refinement to one of the central ideas of the Lean Startup.

The Lean Startup made an essential contribution by shifting attention away from executing fixed plans and towards learning through experiments. Its concept of validated learning helped founders recognize that progress should not be measured only by features delivered, milestones completed, or money spent. Progress should also be measured by what the company has learned about whether its proposed business model can work.

Academic examinations of the Lean Startup framework similarly identify business model assumptions, customer development, minimum viable products, experimentation, and decisions to persevere or pivot as central elements of the approach.⁷

I still agree with that principle. However, I would now use the word validated more carefully.

Validation can easily become an invitation to search for evidence that confirms what founders already want to believe. A positive interview, a promising click-through rate, or an enthusiastic response to a price can then be interpreted as proof that the hypothesis is correct, while contradictory evidence is explained away.

This is not merely a startup problem. Confirmation bias is a well-established cognitive tendency through which people seek, select, interpret, or remember evidence in ways that support beliefs they already hold.⁸

A more scientifically disciplined approach would therefore be to speak about falsifiable learning.

Instead of asking:

How can we validate our pricing model?

founders should ask:

What evidence would convince us that our pricing model is wrong?

What would make us reject the assumption that customers will buy through partners?

What would show that a self-service model requires too much human support?

What would indicate that a subscription price cannot finance acquisition, onboarding, delivery, and retention?

What customer behavior would contradict our belief that the problem is recurring enough to justify recurring payment?

What would need to happen for us to abandon, revise, or qualify the assumption?

The underlying logic reflects the principle of falsifiability associated with Karl Popper: hypotheses should be formulated so that evidence could, in principle, contradict them. Repeated supportive observations may increase confidence, but they do not establish a general claim as permanently true.⁹

Of course, business experiments are not controlled scientific laboratory studies. Markets are open systems. Conditions change, customers differ, competitors react, and causes are difficult to isolate.

Yet the discipline remains valuable.

An experiment should give the assumption a realistic opportunity to fail.

This does not reject the Lean Startup. It strengthens it.

The important shift introduced by the Lean Startup remains valid: founders should learn before they scale. The improvement is to design experiments that expose assumptions to possible failure rather than merely collecting evidence that makes the team feel more confident.

The term validated learning describes the desired outcome. The term falsifiable learning sharpens the method used to reach it.

Progress is therefore not the accumulation of positive signals. It is the systematic elimination, revision, or qualification of assumptions that cannot survive contact with evidence.

Pricing as an experimental array

Pricing experiments are often too narrow. A company changes one price point and compares the result with the previous one. That may produce useful information, but pricing involves several interacting dimensions.

A more useful approach is to treat pricing as an experimental array.

The company might investigate several pricing models, such as subscription, one-off payment, or usage-based pricing. Within each model, it could explore different packages, price ranges, payment terms, sales paths, and service levels.

The objective is not to expose every customer to every possible combination. Different versions can be investigated with different customer groups through conversations, mock-ups, quotations, pilot offers, or real transactions.

This produces more useful evidence than testing one price in isolation because the company can begin to see which combinations work for particular customers and under which conditions.

It may discover that a subscription works for smaller customers, while enterprise clients prefer a license combined with implementation fees. It may learn that customers accept a higher price when onboarding is included, or that an apparently attractive low-cost plan generates so much support demand that it destroys the margin.

The preferred combination is not necessarily the one that produces the highest conversion.

High conversion is irrelevant if acquisition, onboarding, support, and retention costs make every additional customer economically destructive.

The relevant question is whether the configuration creates sufficient value for customers while allowing the company to acquire, serve, and retain them sustainably.

Launch is not the end of learning

Even the strongest pre-launch experiments remain incomplete. Once a product enters the wider market, the company receives behavioral data that were unavailable before.

This does not mean the earlier work was unnecessary. It means learning continues with stronger evidence and higher stakes.

Customers buy or decline. They renew or leave. They expand their usage or remain on the smallest plan. Sales cycles become visible. Support demand becomes measurable. Acquisition and delivery costs become clearer. Competitors respond, expectations change, and the value of the offering evolves.

The business model must continue learning alongside the product.

That is why I see pricing as part of Continuous Business Model Innovation rather than as a decision that is made once and then implemented.

Existing customers may not need to pay the same price as new customers. Early adopters accepted a less mature product and often contributed feedback, patience, credibility, and learning. New customers may receive a broader, more reliable, and more valuable offering.

Likewise, not every customer segment needs to see the same packages, payment terms, or pricing model. Different customers may have different jobs, procurement constraints, service requirements, alternatives, and cost-to-serve profiles.

Business model experimentation should therefore not stop once an initial model has been selected. Research into Lean Startup and business model experimentation similarly emphasizes iterative learning, continued stakeholder engagement, and the possibility that new business model configurations emerge through the experimentation process itself.¹ ²

The market will keep producing new evidence.

The responsibility of the company is to keep using it.

The real decision

Looking back, I would now phrase my original claim more carefully.

Not every element of a business model can be fully validated before launch. However, every critical element can be made explicit, investigated, exposed to possible falsification, and supported by progressively stronger evidence before the company commits at scale.

Pricing is not an exception to this principle. It may be one of its most consequential applications.

The economic model determines whether the company can create enough value for customers and retain enough of that value to become financially viable. It influences whether the venture can reach break-even, continue operating, and attract further capital, knowledge, employees, partners, and ecosystem support.

A mistake in this area is not a minor pricing error. It can become a bet on an entire system of sales, delivery, support, and financing that cannot sustain itself.

That bet is too expensive to place solely on confidence, convention, competitor comparisons, or somebody’s best guess.

The purpose of experimentation is not to prevent every mistake. It is to expose more of the unknowns, reduce avoidable uncertainty, and make the inevitable market learning less destructive.

Launch should not be the first time the economic logic encounters reality.

It should be the moment a partially tested business model begins its next, more demanding round of continuous falsification and learning.

Paid application section

Apply it to your decision

Understanding that pricing is an interconnected system of assumptions is useful. The more demanding task is applying that understanding to a real business model decision.

The following process helps you identify the economic assumption carrying the greatest risk, define what would falsify it, and decide what evidence you require before increasing the commitment.

Read more

That interpretation ends too early.

LEGO did not restore one business model and then leave it unchanged. It repeatedly adapted the elements around a stable System in Play.

Wooden toys established the play business. Plastic molding created scale. The System in Play made purchases cumulative. DUPLO extended the age range without breaking compatibility. The minifigure added stories and role play. LEGOLAND turned products into experiences. Education introduced schools as customers. Star Wars brought licensed stories into the system.

The 2004 crisis forced LEGO to reduce complexity, sell assets and reconsider which capabilities belonged inside the company. Branded retail and e-commerce deepened customer relationships. LEGO Ideas added fan co-creation. NINJAGO created a home-grown transmedia world. Adult products broadened the audience. LEGO Fortnite connected physical and digital play. The 2026 acquisition of the Discovery Centres selectively brought experiences back under direct ownership.

Each adaptation expanded what the LEGO system could do. Each also added new complexity that the company had to govern.

The familiar story ends too early

The familiar LEGO story centres on the financial crisis of 2003 and 2004. The company had expanded its portfolio, activities and asset base while facing declining demand for traditional toys, stronger price competition, retail consolidation and competition from consumer electronics.

LEGO’s 2004 Annual Report described a fundamental change in direction. The company would concentrate again on its classic core products, reduce its cost base and lower its exposure. The following year, it sold the LEGOLAND parks because they did not directly fit the core business as then defined.

Returning to the brick was consequential. But it did not complete the adaptation.

If recovery had meant only making fewer products around the original brick, the LEGO Group that followed would have been a smaller version of the company that entered the crisis. Instead, LEGO expanded again through licensed franchises, fan communities, owned story worlds, direct retail, adult audiences, loyalty programs, digital play and branded experiences.

The difference was not expansion versus focus. It was coherent expansion versus complexity that the business model could no longer support.

The brick remained important because it provided a stable foundation. The more consequential strategic asset was the System in Play: a commitment that elements bought at different times would continue to work together and increase one another’s usefulness.

LEGO did not win by returning to one product. It won by restoring coherence to the system around that product.

Four phases of adaptation

1. Turn separate toys into a cumulative system

LEGO began with wooden toys and other household products. By 1935, Ole Kirk Kristiansen had decided to concentrate on toys and establish a distinct brand around the idea of playing well.

The move into plastic during the late 1940s changed more than the material. Injection molding required new equipment, technical knowledge, tooling and capital. It also created a manufacturing model capable of scaling beyond the limits of woodcraft.

The decisive change came with the LEGO System in Play in 1955. Instead of treating every toy as a separate purchase, LEGO created a system in which elements acquired at different times remained compatible. The stud-and-tube principle patented in 1958 provided the stability needed to make that promise work.

Compatibility changed the economics of subsequent purchases. A new set did not replace the old one. It increased what the existing collection could become.

LEGO was no longer selling only individual toys. It was selling additions to an accumulating system.

2. Expand the system into new ages, jobs, stories and experiences

DUPLO extended the potential relationship to preschool children without creating a disconnected product architecture. The larger bricks remained compatible with standard LEGO elements. A new customer segment strengthened the existing system rather than fragmenting it.

The minifigure changed the job the product could perform. LEGO construction had concentrated largely on houses, vehicles and other physical structures. Characters introduced identity, narrative and role play while still fitting the same system.

LEGO Education then added schools and teachers as customers. The buyer, channel and expected outcome changed. LEGO added curriculum-linked materials, manuals and learning objectives, but the construction system remained recognizable.

LEGOLAND moved in another direction. It turned the creativity enabled by the product into a public, immersive experience. LEGO was no longer present only in the home, shop or classroom. The brand itself became a destination.

Licensed intellectual property extended the system again. LEGO Star Wars, launched in 1999, brought a recognized external story world into LEGO play. It gave LEGO access to existing demand while giving the franchise a new form of participation through construction.

These were substantially different adaptations. They changed customer segments, partners, channels, activities and value propositions. What connected them was that each could increase the reach or usefulness of the same underlying system.

3. Restore coherence and reconsider what belonged inside

The 2004 crisis forced LEGO to distinguish between extensions that reinforced the system and activities whose complexity, assets or economics it could no longer support.

The company refocused on the brick, reduced costs and risk, and sold the LEGOLAND parks. The experiences continued under Merlin Entertainments, allowing the LEGO brand to remain present without the LEGO Group operating the capital-intensive parks itself.

LEGO also tested the opposite boundary. Under its rescue plan, it outsourced a substantial share of production to Flextronics. The arrangement was intended to reduce costs. Instead, component shortages and coordination problems made it harder to respond to demand.

LEGO ended the partnership and brought production back inside. The company later described precision molding as a core competence it should not have handed over.

Selling the parks and insourcing production were not contradictory decisions. They reflected different constraints.

Park ownership increased capital exposure without being essential to the production of LEGO play. Manufacturing control affected quality, responsiveness and the integrity of the system itself.

The important question was therefore not whether LEGO should own more or own less. It was which capabilities and assets required control, and which could create value through partnership.

4. Deepen the relationship and connect physical with digital play

After restoring coherence, LEGO began expanding again without returning to the earlier logic of uncontrolled scope.

LEGO.com, branded stores and the LEGO Insiders program created more direct relationships alongside retail partners. LEGO gained greater control over presentation, access to shopper data, opportunities for exclusives and a continuing membership relationship.

LEGO Ideas extended that relationship into product development. Fans could propose concepts, attract public support and potentially see their designs become official products. LEGO retained selection and production control while using the community to supply ideas and reveal demand.

NINJAGO reduced part of the dependence on external franchises. It connected sets, television, games, films and merchandise around LEGO-owned characters and stories. What began as a product line with a planned end became an evergreen property because customers continued asking for it.

LEGO also broadened its definition of the customer. More complex, display-oriented and passion-led sets made adult fans an explicit growth segment rather than an incidental audience for a children’s product.

LEGO Fortnite moved the system into a continuously operated digital environment through the partnership with Epic Games. Physical Fortnite sets then translated characters and locations from the game back into construction products.

The movement now works in both directions. Physical play informs the digital experience. Digital participation creates demand for physical construction. Both can strengthen the wider LEGO system.

The 2026 acquisition of 29 LEGO Discovery Centres and LEGOLAND Discovery Centres changed the boundary again. LEGO brought smaller indoor experiences and their retail outlets under direct ownership, while Merlin continued focusing on the larger LEGOLAND resorts.

This was not simply a reversal of the 2005 park sale. LEGO selectively reclaimed the experiences that could deepen direct customer relationships while leaving capital-intensive destination parks with a specialist partner.

Three lessons from the case

1. The core can be an architecture rather than a product

Calling the brick LEGO’s core is understandable but incomplete. Many companies can manufacture compatible construction bricks. LEGO’s stronger core is the system around the brick: compatibility across generations, manufacturing precision, open-ended construction, recognizable design and the ability to combine building with stories and role play.

This distinction changes how leaders should evaluate adjacencies.

DUPLO changed the scale of the element. The minifigure introduced characters. Education changed the customer and outcome. Star Wars added an external story. NINJAGO created an owned story world. Fortnite added digital play.

Each moved away from the original product in some respect. Each could still reinforce the underlying architecture.

The relevant question is not simply whether a new offer resembles what the company already sells. It is whether the new offer increases the value, reach or defensibility of the system the company already possesses.

2. Expansion creates value when it compounds the system

An extension can add revenue without strengthening the business model. It may use the brand, customers or distribution while adding separate capabilities, costs and dependencies that do not reinforce what already exists.

LEGO’s strongest adaptations created cumulative effects. DUPLO extended the customer’s potential lifetime with the system. New themes increased the usefulness of existing bricks and minifigures. Fan-created sets added variety without replacing the core architecture. Digital worlds could create demand for physical sets. Direct channels strengthened customer relationships across the portfolio.

The 2004 crisis showed the other side. Growth can destroy value when the organization cannot explain how its extensions reinforce one another or support the economics needed to deliver them.

Business-model expansion should therefore be judged by more than the attractiveness of the new market. Leaders should ask what the extension makes more valuable elsewhere in the system — and which additional complexity the organization must absorb in return.

3. System value and system complexity can grow together

LEGO’s adaptations created more ways to enter and remain within the ecosystem: preschool play, role play, learning, licensed stories, fan-designed products, adult display, games, films, physical retail, memberships and branded experiences.

The same moves added products, partners, rights, platforms, data, channels, operating models and organizational dependencies.

This creates the strategic tension at the centre of the case:

Every new theme, license, audience, channel and physical-digital experience can strengthen the System in Play or recreate the complexity exposed by the 2004 crisis.

The problem is not whether LEGO should expand. It is how to distinguish extensions that reinforce the system from activities that merely borrow the brand.

The 2005 park sale, the outsourcing and insourcing of production, the Epic Games partnership and the 2026 acquisition of the Discovery Centres show that the answer can change over time.

What belongs inside the company should follow the constraint. When capital exposure threatens resilience, partnership may be valuable. When quality, learning or responsiveness depends on control, ownership may be necessary.

Continuous business-model innovation therefore includes expansion, contraction and reversals. The discipline lies in preserving coherence while the boundary of the business changes.

Future cases will examine other organizations that changed their business models as technologies, customer behavior and old assumptions became constraints. Subscribe for source-backed cases, strategic interpretation and practical decision tools.

3 Lego Continuous Business Model Innovation

78.5KB ∙ PDF file

Download

A two-page, source-backed chronology of fourteen business-model adaptations, including a Business Model Canvas View, Core Pattern and Strategic Tension.

Download

What would you challenge?

Which LEGO adaptation most fundamentally changed the company’s business model?

I am particularly interested in interpretations that challenge the chronology, the business-model elements I assigned or the tension between extending the System in Play and recreating organizational complexity.

Use this case with your team

I am developing Continuous Business Model Innovation as a format for executive briefings, strategy off-sites, workshops, bootcamps and teaching.

The purpose is not to copy LEGO or expand into another adjacency because it worked for someone else. It is to help a team determine what must remain invariant, which part of its business model has become a constraint, where complexity is accumulating and what should be tested before committing further resources.

If your organization is facing such a decision, send me the concrete situation you are working through.

This analysis reflects my personal interpretation of the LEGO Group’s business-model adaptations based on available public data and public sources. LEGO, the LEGO logo and related marks are trademarks of the LEGO Group. This independent analysis is not affiliated with or endorsed by the LEGO Group.

Subscribe to INNOVATION&

Evidence-based analysis, practical tools and innovation strategy consulting for leaders deciding what to test, fund, pivot, stop or scale.

Subscribe now

Anyone who knows me knows that I have been a passionate Capoeira practitioner for more than 26 years. It is an important part of my life. But getting older also means learning to manage the physical struggles that come with age, especially when you want to continue practicing something as demanding as Capoeira.

I was leaving my new physiotherapist after one of those sessions when I began to wonder whether getting the customer’s job done is really enough.

Most physiotherapy follows a familiar pattern: assess the problem, treat it over several sessions, and finish when the immediate condition has improved. The pain eases, movement improves, and I usually leave in a better state than when I entered.

By that measure, the job has been done.

But that is not really what I am paying for.

I do not want a collection of successful treatment episodes. I want something about the pattern itself to change. Ideally, the problem should occur less frequently, I should understand my body better, become more confident in how I move, and need less help managing the same situation when it returns.

That does not mean never seeing a physiotherapist again. Some conditions require ongoing support, and preventing deterioration can itself represent meaningful progress. But the value cannot be judged only by whether I feel better when I walk out of the treatment room.

Jobs to Be Done helps us understand the progress that pulls a solution into someone’s life. My experience made me interested in what happens on the other side of that moment:

The job may be done. But what was there an impact on the trajectory behind it?

Products Are Visible. Jobs Are Not.

Jobs to Be Done shifted attention away from products and customer categories toward the progress people seek in a particular circumstance.[1]

Someone buys a drill, but the drill itself was never the desired result. Someone enrolls in a course, although sitting through lessons is rarely the progress they ultimately want. Someone commissions a report, even though owning another document is seldom the true objective.

The product is visible. The job is not.

Understanding the job requires us to look beyond the transaction and reconstruct what was happening in the customer’s life when the solution became necessary. We need to understand what had changed, what had become uncomfortable, what the customer had already tried, and why they were finally willing to abandon a familiar routine and commit money, time, effort, access, or reputation to something new.

Different approaches within Jobs to Be Done have made this thinking increasingly actionable. Switching research examines the forces that push people away from their current situation and pull them toward an alternative. Solution-independent job mapping breaks progress into the steps customers must complete, regardless of the product they currently use. Outcome-oriented methods identify the measures customers use to judge whether the job has been performed successfully.[1][2]

These approaches help us understand why a solution is hired, where customers struggle, what prevents them from switching, and what better means from their perspective.

I want to extend that temporal lens in the other direction.

We reconstruct what happened before the hire to understand why a solution entered the customer’s life. But what happens after the hire?

Not merely whether the customer got the desired outcome, but whether using the solution changed the pattern from which the next struggling moment will emerge.

The Job Sits Inside a Trajectory

My physiotherapy session did not really begin when I entered the treatment room. It emerged from a much longer sequence of recurring discomfort, previous treatments, exercises I continued or abandoned, temporary improvements, relapses, growing frustration, and changing beliefs about what might help.

Switching physiotherapists and methods was one event in a broader history of attempted progress.

The same is true in most markets.

A company does not suddenly commission a strategy report. The purchase may be the result of months of uncertainty, internal disagreement, postponed decisions, earlier analyses, executive pressure, and perhaps an approaching board meeting that has made continued ambiguity increasingly difficult to defend.

A product team does not suddenly adopt a discovery platform. It arrives there through unreliable interviews, feature debates, weak evidence, pressure to demonstrate progress, and growing concern that the roadmap may not be addressing anything customers genuinely value.

A person does not suddenly subscribe to a course. The decision may follow a promotion they did not receive, repeated attempts to learn independently, anxiety about becoming obsolete, admiration for someone else’s career, or the growing realization that their current path is no longer leading where they want to go.

The job captures the progress sought in the struggling moment.

The trajectory captures the pattern that produced that moment and continues after it.

That distinction matters because two customers can hire the same product for what appears to be the same immediate job while moving along very different trajectories.

Two people may both hire a physiotherapist to reduce back pain. One may be recovering from an isolated injury and need temporary support before returning to normal activity. The other may be caught in a recurring pattern involving movement, work, stress, behavior, and repeated short-term interventions.

The immediate job may sound identical.

What needs to change afterwards may not be.

An Outcome Is Not Yet an Impact

An outcome tells us that something happened. Pain decreased. A report was delivered. A behavior changed. A prototype was launched. A task was completed faster. A customer returned. A user adopted a feature.

Impact asks a more demanding question:

What changed because the solution entered the system, compared with what would probably have happened without it?

In rigorous evaluation practice, impact is not simply the condition observed after an intervention. It is the difference between that condition and a credible estimate of what would likely have happened otherwise.[3]

That does not mean product teams need randomized controlled trials every time they change a service. It means we should be careful about attributing every positive outcome that follows an intervention to the intervention itself.

Imagine that my pain falls from seven to three after every physiotherapy session but returns to seven two weeks later. The immediate outcome is real, and the treatment may still be valuable. Nevertheless, the broader trajectory may be largely unchanged.

Now imagine that the pain improves more slowly, but I learn what triggers it, become more confident in managing it, and experience fewer serious episodes over the following year. The immediate result may appear less dramatic, while the impact on the trajectory is considerably greater.

The same pattern appears in business.

A consultant may deliver an excellent report and help an executive team make a decision before an important deadline. The immediate job has been completed.

What happens, however, when the next uncertain decision arrives?

Perhaps the organization has learned how to expose assumptions, evaluate evidence, and disagree productively. Perhaps it temporarily borrowed the consultant’s capability and returned to its previous state afterwards. Or perhaps it has become even more dependent on external interpretation because internal judgment was displaced rather than strengthened.

The visible deliverable can be identical in all three cases.

The trajectory is not.

Research on service value already challenges the assumption that value is embedded in a product and simply transferred at the point of purchase. Value emerges through use, context, relationships, and the customer’s integration of multiple resources.[4] Customers’ assessments of value can also evolve throughout a service relationship rather than remaining fixed at the moment of the transaction.[5]

The trajectory perspective extends this thinking by asking not only whether value emerged during use, but what direction the customer was left moving in afterwards.

Every Solution Changes More Than the Immediate Job

When a solution is hired, it does more than remove a problem. It can redistribute work, knowledge, judgment, responsibility, and capability between the customer and the provider.

A navigation system performs part of the orientation that was once handled by the driver. A consultant performs analysis that an organization might otherwise have developed internally. An AI system may research, interpret, compose, recommend, and increasingly act. A physiotherapist may provide relief, diagnose a pattern, teach self-management, or become the recurring mechanism through which the problem is controlled.

None of these arrangements is inherently good or bad.

People and organizations depend on others because no individual or company can, or should, develop every capability internally. Human agency includes acting directly, acting collectively, and acting through others.[6]

The more interesting question is what that delegation does to the trajectory.

Does the customer become more capable of recognizing and responding to the next situation? Does the solution provide legitimate ongoing support for something the customer cannot reasonably perform alone? Does it preserve the customer’s ability to question, switch, recover, or continue independently?

Or does each successful use make independent action more difficult?

Automation research has long shown that removing routine work does not necessarily remove the need for human expertise. In some cases, people remain responsible for rare and abnormal situations precisely after the routine practice that helped maintain their capability has disappeared.[7] Related work distinguishes appropriate automation use from misuse, disuse, and designs that disregard the consequences automation creates for the people relying on it.[8]

The immediate job can therefore be performed more efficiently while the wider trajectory becomes less desirable.

That is not only an automation problem.

It is an impact problem.

Repeated Use Does Not Automatically Prove a Better Trajectory

Companies naturally interpret repeated use as evidence of value. The customer returns, the subscription renews, another consulting engagement is commissioned, another therapy session is booked, or the software becomes more deeply embedded in the customer’s operations.

Sometimes this is exactly what it appears to be. The need legitimately recurs, the solution continues to create value, and the customer chooses to remain.

Recurrence, however, is ambiguous.

A patient returning to a physiotherapist may indicate trust and effective long-term care. It may also indicate that each session relieves the episode without changing the pattern that continues to produce it.

A company renewing software may demonstrate that the product has become essential. It may also reveal that switching costs have become prohibitively high.

A leadership team repeatedly hiring consultants may reflect sensible access to specialized expertise. It may equally suggest that internal capability has never transferred.

Retention can indicate value, but it can also indicate dependency, inertia, or an unresolved job that keeps returning in a form the provider is paid to manage.

Churn is equally ambiguous.

A customer may leave because the product failed. They may also leave because the solution succeeded so completely that it is no longer needed.

When a course helps someone become independently capable, cancellation may be part of the intended impact. When physiotherapy enables someone to manage a recurring problem with fewer interventions, declining appointment frequency may represent success rather than failure. When a consultant helps an organization make similar decisions without external support, lower repeat revenue may signal that capability has genuinely transferred.

This creates an uncomfortable business-model question:

What happens when the customer’s best trajectory is economically less attractive to the provider than repeatedly completing the immediate job?

When the Business Model Benefits From the Job Returning

Many business models are rewarded for solving the same job repeatedly. Revenue grows when customers return, consume more, renew, remain active, or become more deeply integrated into the solution.

That does not mean providers deliberately preserve the customer’s problem. It does mean that the commercial model can favor visible, recurring outcomes over impacts that eventually reduce future demand.

A consultancy may be rewarded for producing excellent decisions without transferring decision capability. A software company may become more valuable as the customer’s processes and knowledge grow increasingly inseparable from the platform. An AI service may benefit when people delegate more tasks, even if repeated delegation weakens their ability to judge the output.

Ongoing revenue and positive impact can certainly coexist. Continued support may help a customer improve, remain stable, or decline more slowly than they otherwise would.

The relationship should simply not be assumed.

A recurring business model should be able to answer:

Are customers returning because their trajectory is improving with continued support, or because it never changes enough for the support to become less necessary?

That is more difficult to assess than satisfaction, completion, usage, or retention. It may also tell us something those measures cannot.

Every Solution Becomes Part of What Happens Next

A trajectory is shaped by accumulated choices.

Every solution introduces new habits, knowledge, expectations, dependencies, routines, and switching costs. Small early decisions can become self-reinforcing, making some future paths easier while gradually making others harder to reach.[9]

A solution therefore does not merely compete to perform the current job. It becomes part of the conditions under which the next job will arise.

An AI assistant may make today’s report easier while changing how tomorrow’s analyst learns to reason. A platform may simplify today’s coordination while determining which data, processes, and relationships remain available during the next transformation. A consultant may resolve today’s strategic uncertainty while changing whether the organization can resolve the next one itself. Physiotherapy may relieve today’s pain while changing—or failing to change—the likelihood, severity, and manageability of the next episode.

This gives us three different questions:

The job asks: What progress is the customer trying to make?

The outcome asks: What happened when the solution was used?

Impact asks: What became different afterwards compared with what was likely to happen otherwise?

The job explains why the solution enters the customer’s life.

The outcome tells us whether it performed.

Impact asks whether its presence changed what was likely to happen next.

That is the trajectory.

And it leads to a harder standard for judging innovation:

Completing the customer’s job is not enough to know whether we made an impact. We also need to understand what direction the customer was left moving in afterwards.

The practical challenge is how to evaluate that trajectory without turning ordinary product and innovation work into a longitudinal research program.

That is what the Trajectory Impact Review is designed to do.

Paid application section

The Trajectory Impact Review

Understanding the distinction between a job, an outcome, and an impact is only useful if it changes how we evaluate a real product, service, intervention, or innovation decision.

The Trajectory Impact Review is designed to make that distinction practical.

Read more

When a start-up reaches Business Model Fit and demonstrates traction, it may become an acquisition candidate. For a large company, this can look like an attractive way to reduce innovation risk. Instead of financing years of uncertain discovery, it can acquire a product, paying customers, specialist talent, and a business model that has already demonstrated that it can create, deliver, and capture value.

That can be entirely rational. But it is easy to draw the wrong conclusion from the evidence the start-up has created.

Traction tells us that the business model worked under a particular set of conditions: the start-up’s ownership, incentives, team, decision speed, customer relationships, partner network, architecture, and cost structure. An acquisition changes many of those conditions. At the same time, the acquisition price may require the buyer to create considerably more value than the start-up has demonstrated on its own.

The uncertainty has therefore not disappeared. It has changed its object.

Before acquisition, the question was whether the venture could work. After acquisition, the questions become whether its value will survive new ownership, whether the buyer can create additional value, and whether that additional value is sufficient to justify the price paid.

This distinction becomes particularly important when the acquisition case depends heavily on synergies. The more value the buyer must create after closing for the investment case to work, the less reasonable it is to describe those synergies as upside.

They have become part of the acquisition’s burden of proof.

Business Model Fit Does Not Prove Acquisition Fit

Recent OECD research illustrates why corporations may rationally enter after external ventures have already created evidence. A 2026 study covering 240 corporate investors and more than 44,000 start-ups found that corporate venture investors disproportionately targeted technology-intensive ventures with large patent portfolios. Those ventures were also more likely to be acquired later, although not necessarily by the original corporate investor.[1]

There is a sensible logic behind this division of labor. Founders and investors finance much of the uncertain search for a relevant problem, a workable solution, and a viable business model. A corporation can commit later, after evidence has accumulated.

The problem begins when evidence about the venture is allowed to migrate into claims about the acquisition.

I find it useful to separate three investment theses.

The venture thesis asks whether the target has found a viable business model. The transferability thesis asks whether the value demonstrated by the target will survive the change of ownership. The synergy thesis asks whether this particular buyer can create additional value beyond what the target could achieve independently.

These theses are related, but evidence for one does not automatically prove the others.

A start-up may have excellent retention, strong unit economics, proprietary technology, loyal customers, and rapid growth. All of that strengthens the venture thesis. It does not tell us whether customers will behave the same way after the change of ownership, whether key employees will remain, whether partners will continue contributing, or whether the target will retain its learning speed inside a larger organization’s governance.

Nor does it prove that the buyer’s sales channel can sell the offer, that its technology organization can integrate it without damaging the product, or that the combined organization can create enough additional value to earn back the acquisition premium.

An acquisition case can therefore contain excellent evidence for the venture thesis while relying largely on assumptions for transferability and synergy.

That is the first place where proven traction can create false confidence.

The Price Creates a Second Investment Case

The target initially attracts the buyer because of what it has already demonstrated. Then the valuation arrives.

At some price, the target’s standalone economics may no longer clear the buyer’s investment hurdle. The premium needs to be defended, and the acquisition case begins to move from demonstrated value toward combined potential: cross-selling, international expansion, shared technology, data access, procurement savings, overhead reduction, or faster market entry.

None of these claims is inherently unreasonable. The problem is how they are treated.

If the target’s standalone value already supports the investment and synergies create additional returns, they genuinely are upside. But if the acquisition fails to meet the buyer’s required return unless those synergies materialize, they are no longer optional benefits. The board is now financing a second investment case: the buyer’s ability to create value after closing.

Research published in Management Science makes this burden measurable. Ellahie, Hshieh, and Zhang calculate the minimum improvement in a target’s return on equity that an acquirer would need to achieve to break even on the acquisition price. They call it implied return-on-equity improvement.[2]

Deals requiring greater improvement proved less attainable and predicted worse subsequent acquirer performance. Over the first three years, acquirer ROE growth was 11 percentage points lower for high-burden deals than for low-burden deals. High-burden deals were also associated with higher operating costs, tighter financial constraints, lower investment, and larger and more frequent goodwill impairments.

This does not mean that paying a high price makes an acquisition wrong. It means something more useful: the price determines how much improvement the buyer must subsequently produce.

Valuation and integration are therefore not independent parts of the transaction. The valuation defines the performance burden. The integration thesis explains how the buyer intends to carry it.

Synergy Dependence Matters More Than Synergy Size

Consider two acquisitions that each claim €100 million in synergies.

In the first, the standalone value of the target already supports most of the investment case. Even if only €20 million of the projected synergies materializes, the acquisition can still generate an acceptable return.

In the second, €80 million must materialize simply to justify the purchase price and integration cost.

Both presentations may contain exactly the same €100 million synergy number, but they describe very different decisions. The first acquisition can tolerate substantial estimation and execution error. The second cannot.

This is why I think synergy dependence matters more than synergy size.

When standalone economics support the deal, synergies can reasonably be treated as contingent upside. When the deal fails without them, they become load-bearing assumptions.

A recent NBER paper provides an interesting parallel. It shows that among proposed mergers, different sources of expected profitability can compensate for weakness elsewhere because the transaction must clear an overall profitability threshold.[3] The paper examines merger efficiencies and market power rather than corporate board decision-making, so it does not demonstrate that a particular synergy estimate has been exaggerated.

But it suggests an important governance implication. The part of an investment case doing the most work because another part is weak should not receive less scrutiny simply because the overall case appears attractive.

It should receive more.

The claim required to rescue the deal deserves the highest evidence standard.

The reason is simple: necessity raises the cost of being wrong.

More Detail Is Not More Evidence

Once a deal depends on synergies, there is a natural tendency to make those synergies increasingly concrete. The mechanism becomes more detailed. The spreadsheet gains more lines. Forecasts acquire precise dates, conversion rates, retention assumptions, integration milestones, and responsible executives.

That detail can be valuable because it makes assumptions inspectable. But it can also create an illusion of evidence. A forecast does not become better supported merely because an uncertain assumption has been expressed to two decimal places.

A 2026 working paper examining synergy guidance in 12,176 US acquisitions found that investors responded positively when buyers discussed synergies and more strongly when estimates were more intensive or numerically larger.[4] Longer-term results were less favorable. Synergy disclosers experienced weaker post-merger operating performance and more frequent and larger goodwill impairments. Larger numerical estimates also predicted downward revisions and lower long-run stock returns.

The paper has not completed peer review, so its estimates should not be treated as settled. But the distinction it raises is valuable.

A synergy statement can improve the acquisition story before it improves the business.

The deal team needs to make the acquisition understandable and approvable. The board has a different job: determining whether the material presented is not only coherent, but predictive.

The Buyer Must Become Part of the Diligence

This leads to an asymmetry in many acquisition processes that deserves more attention. Diligence examines the unfamiliar target in extraordinary detail while treating the buyer as if its capabilities were already known.

Once the acquisition depends materially on synergies, that assumption no longer holds. The buyer itself becomes part of the hypothesis.

Take cross-selling. A large installed customer base is often presented as evidence for a revenue synergy. But the customer base is not the synergy. It is a resource. The synergy exists only if relevant customers actually buy the acquired offer through the buyer’s channel at an acceptable acquisition and delivery cost.

The same applies to the sales organization. Having thousands of salespeople does not demonstrate that they can sell an unfamiliar product. They may lack specialist knowledge, appropriate incentives, the right customer relationships, implementation support, or simply the management attention required to prioritize another offer.

Technology, procurement, data, brand, and international presence should be treated in the same way.

Resources are inputs. Synergies are realized outcomes.

That creates an important distinction between inventory and capability evidence.

“We have the channel” tells us what the buyer possesses.

“We have repeatedly sold this type of offer through this channel to comparable customers under comparable conditions” tells us something about what the buyer can do.

If the buyer has never performed the intervention required by the acquisition case, then there is another venture hidden inside the acquisition: building the buyer capability required to justify the price.

That uncertainty belongs in both the valuation and the decision.

Ownership Can Damage the Evidence You Are Buying

There is another complication. The value of a start-up rarely resides only in its product, contracts, patents, and employees. It may also depend on less visible conditions: direct access to customers, rapid product decisions, trusted ecosystem relationships, a focused roadmap, founder and employee incentives, partner neutrality, or freedom from the buyer’s architecture, procurement, compliance, and reporting processes.

An acquisition does not simply transfer those conditions from one owner to another.

It changes them.

OECD research covering start-ups in 60 countries between 2001 and 2021 found that patenting declined after acquisition without a corresponding increase in the acquirer’s innovation activity.[5] Patents are an incomplete measure of innovation, and this does not establish that every acquisition suppresses useful output. It does show why purchasing an innovative company cannot be equated with preserving its innovation trajectory.

Research on acquisitions of firms involved in open-source projects provides another view of the mechanism. Collaboration intensified when acquirers sought to protect the target’s complementarities and continued investing in partner relationships. It declined when acquisitions were primarily oriented toward extracting technology or employees. Structural independence also supported collaboration in protection-oriented acquisitions.[6]

Customers, employees, and partners are not passive assets. Their behavior can change when ownership changes.

That means every material synergy claim should have a corresponding shadow question. If the acquisition asks what the buyer can add, it should also ask what ownership might remove.

The answer may not appear as a line in the integration budget. Lost value may instead surface as slower learning, weaker partner contributions, employee departures, customer churn, reduced differentiation, or a delayed roadmap.

Those are still acquisition costs.

Integration Must Follow the Value Mechanism

This becomes particularly important when an acquisition promises both cost and revenue synergies.

Cost synergies typically require some combination of standardization, consolidation, elimination, and control. Revenue synergies are more likely to depend on access, coordination, customer trust, local judgment, and continued differentiation.

The interventions are not necessarily compatible.

A study of 1,452 US acquisitions found that greater integration was associated with better performance in deals emphasizing cost synergies. For revenue-synergy deals, the relationship followed an inverted U: some integration improved performance, but additional integration eventually reduced it.[7]

A buyer can therefore make the cost case more visible while simultaneously weakening the conditions required for the revenue case. Standardizing technology may reduce operating expense while slowing the target’s roadmap. Consolidating sales may reduce duplication while weakening specialist customer relationships. Centralizing decisions may simplify governance while reducing the target’s learning speed.

This is why “How much should we integrate?” is the wrong question.

The more useful question is:

Which resources must be combined to create each source of value, and which conditions must remain protected for that value to exist?

Integration should follow the value mechanism rather than organizational preference.

Ownership Must Earn Its Irreversibility

There is one final assumption worth examining: that strategic value requires ownership.

Acquisition is only one way to obtain access to technology, customers, capabilities, data, talent, or markets. Licensing, distribution agreements, minority investments, joint ventures, earn-outs, staged ownership, and other structures can sometimes provide strategic access while allowing the buyer to purchase additional evidence before making a larger commitment.

None of these alternatives is automatically superior. In some situations, control genuinely is necessary.

But that is precisely what the acquisition case should demonstrate.

Which source of value actually requires ownership? Which could be obtained contractually? What uncertainty could be reduced before full commitment? What additional value justifies the acquisition premium, integration exposure, and loss of reversibility?

A seller may refuse an alternative structure. That can constrain the transaction. It does not demonstrate that full ownership creates value.

The relevant comparison is therefore not acquisition versus doing nothing.

It is acquisition versus the best credible alternative for obtaining the strategic value being sought.

What the Board Is Actually Deciding

Acquiring a start-up after Business Model Fit can be a rational part of an innovation strategy. It allows a corporation to commit after external actors have absorbed much of the uncertainty involved in discovering whether a viable business model exists.

But that evidence belongs primarily to the venture thesis.

The board still has to evaluate whether the value demonstrated by the venture will survive transfer and whether this particular buyer can create enough additional value to justify the price.

A decision-ready acquisition case therefore needs to distinguish the target’s demonstrated standalone value from the conditions that produced it, determine which of those conditions must survive new ownership, identify how much additional value the price requires, make the mechanism behind each material synergy explicit, and establish whether the buyer has actually demonstrated the capabilities required to execute those mechanisms.

It should also account for value the acquisition itself could destroy and compare ownership with less irreversible ways of obtaining the same strategic access.

A start-up with proven traction has already answered an important question:

Can this business model work?

An acquisition asks a different one:

Can it still work here, under us, at this price?

The higher the premium, the more value the buyer must create after closing. And the more the deal depends on that value, the less reasonable it is to call synergies upside.

They are load-bearing assumptions.

The assumptions carrying the commitment should carry the highest burden of proof.

Apply the Transferability and Synergy-Dependence Gate

The argument above identifies where uncertainty moves after a proven venture becomes an acquisition target. The practical problem is how to prevent that uncertainty from disappearing inside a valuation model, synergy spreadsheet, or integration plan.

The Transferability and Synergy-Dependence Gate is designed for the point before an investment committee or board approves an acquisition. It does not replace valuation, legal diligence, or integration planning. It connects them around a single decision:

Does the evidence justify paying today for value that must first survive transfer and then be increased by the buyer?

The result is a one-page Acquisition Burden Record that makes explicit what the buyer is committing to, which assumptions carry that commitment, and what evidence supports them.

Read more

Some Projects Should Have Stopped Before

I have inherited projects that should have been stopped months, and sometimes years, earlier. They had market research, customer personas, business models, five-year projections, partner shortlists, prototypes, and senior sponsors. Their steering decks looked more mature than many viable ventures.

What they did not have was decision-grade evidence.

The market analysis had been assembled internally. The customer need had been inferred but not observed. The commercial model had been calculated without testing whether a budget owner would pay. The customer shortlist contained organizations that had shown interest but made no commitment. Each artifact supported the next artifact, and the complete package satisfied governance because governance was checking whether the work existed, not whether a consequential uncertainty had been reduced.

My job was to return to the decision beneath the material: What would have to be true for the next commitment to be justified? Then we spoke to relevant customers, observed their current behavior, tested what they would change, and looked for commitments involving money, data, time, access, or workflow disruption. Only then did the board have something it could use—including evidence that supported stopping.

AI makes this distinction more important because the time and expertise required to produce the visible layer of innovation have collapsed.

AI Is Improving Innovation Work, Not Merely Accelerating It

The strongest version of the argument is not that AI produces shallow ideas. The latest research no longer supports that generalization.

A 2026 field experiment involving 791 Procter & Gamble professionals found that individuals using generative AI matched the solution quality of two-person teams without it. AI also helped commercial and R&D professionals produce more balanced proposals across their functional perspectives. Importantly, the study found that AI improved idea generation while human judgment retained value in selecting among ideas.[1]

Another 2026 study in Production and Operations Management found that LLM-generated product ideas achieved higher average purchase-intent scores and were seven times more likely than human-generated ideas to appear in the top 10 percent. The trade-off was lower novelty at the individual-idea level and lower diversity across the portfolio, although newer models and deliberate prompting strategies narrowed that gap.[2]

Field research on creative problem-solving has likewise shown that a strategically guided human–AI process can produce novel and valuable solutions at very low cost.[3] In a different domain, an AI system reported in Nature automated an end-to-end machine-learning research workflow that included idea generation, literature search, coding, experiments, analysis, writing, and peer review.[4]

Innovation automation is therefore not a future possibility. The production, combination, refinement, and preliminary evaluation of options are already being compressed.

That changes the sponsor’s job. When more ideas can be generated, evaluated, and presented at near-zero marginal cost, the scarce resource moves downstream. It becomes the capacity to determine which outputs deserve real-world testing and which results justify greater commitment.

AI Automates the Appearance of Innovation Work

Historically, effort provided a weak but useful signal. A detailed market analysis suggested that somebody had investigated the market. A working prototype indicated that scarce technical capacity had been allocated. A substantial research report implied that people with expertise had spent time collecting and interpreting information.

That signal is disappearing. AI can now create the artifact without the process that the artifact used to imply.

A team can produce a segmented market model without resolving who experiences the problem. It can generate persuasive personas without meeting a customer. It can summarize hundreds of documents without verifying the few claims that control the decision. It can create interview transcripts from synthetic respondents, turn those responses into themes, generate a value proposition, build a prototype, and use another model to evaluate the result.

The output may be useful. What it cannot do by itself is establish that an identifiable customer will change behavior under real conditions.

The distinction is provenance. An artifact tells the sponsor what was produced. Provenance tells the sponsor what the claim is based on.

This matters because fluency influences judgment. In the current revision of a field experiment with 228 evaluators screening 48 early-stage innovations, black-box AI recommendations improved decision quality relative to human-only screening. Adding explanatory narratives increased compliance with the AI but did not improve quality beyond the black box. The effect was strongest for rejection recommendations and risked increasing false negatives among high-potential ideas.[5]

AI can strengthen evaluation. It can also make an evaluation feel better justified than its underlying evidence warrants.

Productivity, Quality, Novelty, Evidence, and Decision Readiness Are Different Claims

Sponsors need to separate five claims that innovation teams increasingly collapse into one.

Productivity asks whether the team produced more output in less time. AI often improves this, although the effect is highly dependent on the worker, task, context, and performance measure. A large field study of customer-support agents found a 15 percent average productivity improvement, with greater benefits for less experienced workers.[6] By contrast, a 2025 randomized study of experienced open-source developers working in familiar repositories found that AI use increased completion time by 19 percent even though participants believed it had made them faster. The authors appropriately described the result as a snapshot of a narrow setting, not a universal estimate.[7]

Quality asks whether the output is better against a defined evaluation criterion. Recent innovation studies show that AI can improve average and top-tail idea quality.[1][2]

Novelty and diversity ask whether the team is searching a sufficiently different solution space. AI can improve individual output while compressing the variety of outputs produced across people. Doshi and Hauser found that generative AI support improved individual story quality but reduced collective content diversity.[8] The 2026 product-idea study found a similar diversity trade-off and showed that it can be mitigated, but not assumed away.[2]

Evidence asks whether the organization has learned something reliable about the external world. A polished concept, high evaluator score, or simulated response is not automatically evidence that a real customer will act.

Decision Readiness asks whether the available evidence is strong enough for the specific commitment under consideration. Even credible evidence can be insufficient if the next decision creates substantial financial, technical, reputational, or political lock-in.

AI can improve the first three claims while leaving the last two almost untouched. That is why faster innovation work does not automatically create better innovation decisions.

Synthetic Customers Are a Search Tool, Not Market Evidence

Synthetic respondents can be useful for rehearsing questions, identifying obvious objections, exploring segments, checking language, and exposing gaps in a research plan. Recent work also shows that LLMs can augment human datasets and, under carefully designed conditions, improve market-research estimates.[9]

The boundary is inferential. A synthetic customer has not experienced the problem, protected a budget, persuaded procurement, integrated a system, abandoned an alternative, or accepted the consequences of a decision.

Research on synthetic survey data illustrates the risk. Bisbee and colleagues found that LLM-generated responses could reproduce average public-opinion patterns reasonably well, yet showed less variation than real respondents, produced different statistical relationships, and changed materially with prompt wording and over time.[10] More recent methodological work argues that synthetic participants may support exploration, but confirmatory claims require explicit calibration with human data and assumptions that justify the inference.[11]

Sponsors should therefore allow synthetic evidence to improve the next test. They should not let it replace the test.

The Sponsor’s Conflict Has Not Been Automated Away

The final problem is not technological. Sponsors are rarely neutral observers of the initiatives they champion.

Research on new-product development found that managers who initiated a project were less likely to perceive that it was failing and more likely to continue funding it than managers who inherited the same project. The effect was stronger for more innovative projects, and better information alone did not eliminate the problem.[12]

AI can intensify this exposure. A sponsor who favors a technology, vendor, strategic theme, or internal team can now receive an almost unlimited supply of coherent support material. Every weak signal can be summarized positively. Every objection can generate a response. Every missing fact can be represented as an assumption inside a model that still produces a confident forecast.

The sponsor’s responsibility is therefore not to demand more material. It is to protect the decision from material that exceeds its evidence.

The free diagnosis is complete: AI has made innovation production more capable, cheaper, and more persuasive. That increases the need to separate output quality from evidence quality and evidence quality from commitment readiness. If governance does not make those distinctions, AI will automate both useful discovery and the appearance that discovery has already happened.

Apply an Automation-Adjusted Funding Gate

Consider an illustrative B2B industrial company exploring an AI-supported service that predicts equipment failures and recommends maintenance actions. After eight weeks, the team asks for €600,000 and six full-time roles to build a production platform and prepare a market launch.

The proposal contains an extensive market map, 40 synthetic customer interviews, six detailed personas, a clickable prototype, a five-year forecast, an implementation roadmap, and positive reactions from twelve demo participants. AI helped the team create and integrate this work quickly.

The sponsor is not deciding whether the work is impressive. The decision is whether the opportunity has earned a commitment that creates architecture, headcount, customer expectations, and political ownership.

An Automation-Adjusted Funding Gate begins with evidence provenance rather than deck completeness.

The Evidence Provenance Ladder

Read more

Capable Organizations Usually Do Not Fail Because They Forget How to Execute

Successful organizations become successful by learning how to make important outcomes repeatable. They improve quality, remove waste, standardize work, strengthen controls, and reduce operational risk. This is not bureaucracy in the dismissive sense. It is accumulated organizational intelligence. An airline needs procedures that constrain improvisation; a pharmaceutical manufacturer needs validation and quality control. Scale depends on deciding which variation is useful and which is dangerous.

Yet the systems that make an organization dependable can also make it strategically brittle. The organization does not suddenly become unintelligent. It becomes highly intelligent within a set of assumptions that the environment may no longer respect.

James March described the underlying tension as exploitation and exploration. Both are necessary, but exploitation tends to produce clearer, nearer, and more measurable returns.[1] Later work on organizational ambidexterity developed the same challenge: mature organizations must run established businesses efficiently while creating the conditions for discontinuous innovation.[2] This is often summarized as a need to “balance” the present and the future, but organizations actually face two different survival problems requiring two different ways of deciding.

One problem is operational: How do we execute what we know with increasing reliability?

The other is epistemic: How do we discover whether what we believe is true before we commit too much to it?

Treating both problems with the same governance logic is how rigorous organizations produce rigorously unsupported decisions.

The Optimization Immune System Protects What Has Earned the Right to Scale

Every mature organization develops what I call an Optimization Immune System. It consists of the mechanisms that protect a validated operating model: budgets, investment committees, performance targets, quality standards, operating procedures, compliance requirements, forecast reviews, procurement rules, and accountability structures.

Its questions are familiar because they are usually the right questions:

• Can the organization deliver this reliably?

• Does the investment clear the required return?

• Are costs, risks, responsibilities, and milestones understood?

• Is performance on plan?

• Can the process be standardized and scaled?

• Who is accountable when delivery falls short?

These questions create enormous value once customers, demand, the offer, delivery, and the economic model are reasonably well understood. At that point, uncertainty should become disciplined execution, variation can be expensive, and consistency becomes an advantage. The Optimization Immune System protects the organization from operational failure. Without it, organizations do not scale; they improvise.

The danger appears when success quietly changes the system’s purpose. What began as a way to protect validated knowledge becomes a way to protect the assumptions beneath the current business. Existing customers shape priorities. Current margins define attractiveness. Established channels determine which opportunities appear reachable. Resource-allocation processes favor proposals that resemble businesses the organization already knows how to evaluate. Christensen and Bower’s study of the disk-drive industry showed how rational attention to powerful customers and established resource-allocation patterns could contribute to leading firms missing strategically important shifts.[3]

The system is still functioning, but it may be defending a model of the world that is losing relevance.

The Validation Immune System Protects the Organization From Becoming Confidently Wrong

A Validation Immune System performs a different function. It does not lower standards, suspend accountability, or create a sanctuary in which innovation teams can avoid commercial questions. Its purpose is to stop the organization from treating assumptions as facts simply because a decision is important.

Its questions therefore sound different:

• Which assumption must be true for this opportunity to deserve further investment?

• What do we know, what do we merely believe, and what remains unknowable at the current stage?

• What observable evidence would increase or reduce our confidence?

• What is the least expensive credible way to obtain that evidence?

• What result would cause us to change the offer, the business model, or the strategic thesis?

• What must become repeatable before optimization governance should take over?

The Validation Immune System protects the organization from strategic error. It converts uncertainty into bounded commitments designed to improve a decision. Discovery-driven planning made this principle explicit: new ventures should surface and test assumptions rather than present projections as facts.[4] Saras Sarasvathy’s distinction between causal and effectual reasoning similarly showed why decisions under genuine uncertainty cannot always begin with prediction.[5] Validation is not the permissive counterpart to optimization. It is disciplined skepticism applied before scale.

An early-stage opportunity usually lacks exactly what optimization governance expects: a stable customer definition, repeatable sales process, proven willingness to pay, reliable delivery model, or credible economics. These are unresolved questions, not administrative omissions that a polished business case can repair.

When an investment committee asks for a five-year forecast before those questions have been answered, the team still produces one. Prices are assumed, adoption curves are drawn, and margins converge toward the corporate target. The spreadsheet is not necessarily dishonest; it is a translation device. The team has learned that uncertainty must be converted into the language of certainty before the organization will fund the work required to reduce it.

The Difference Is Not Creativity Versus Control

The two immune systems are often misrepresented as cultural opposites: imagination and autonomy on one side, discipline and control on the other. Both systems require imagination and discipline, but apply them to different objects. Optimization disciplines execution, reduces unwanted delivery variation, and measures performance against a plan. Validation disciplines belief, creates bounded variation to learn, and evaluates whether evidence should change the plan. Neither is superior; each becomes dangerous when it governs the wrong decision.

This table also exposes why installing an accelerator or innovation lab is insufficient. If the same budget cycle, return thresholds, performance metrics, and steering questions determine its fate, the organization has merely built an innovation room inside the first immune system.

The Hardest Problem Is Routing

The central capability is not choosing exploration instead of exploitation. It is correctly routing a decision according to its uncertainty.

A familiar product extension sold through a proven channel may need optimization governance even if it is called innovative. A modest service attached to an existing product may require validation governance if the buyer, revenue model, operational capability, and customer behavior are new. The amount of technology, executive enthusiasm, and strategic importance are not decisive.

The relevant question is: How much of the decision rests on validated knowledge, and how much rests on assumptions that have not yet survived contact with reality?

Organizations frequently answer by category: core work goes through normal governance and innovation work goes to the lab. That is too crude. A single initiative can contain both kinds of work. The organization may know how to build the product but not how to sell, support, regulate, or profit from it.

Correct routing therefore happens at the level of the consequential assumption, not the project label.

The immune system must also be able to switch. Validation is not a permanent exemption from performance. Once demand, delivery, and economics become repeatable, continued experimentation can become avoidance. Conversely, an established business may need to return temporarily to validation when customer behavior, technology, or regulation invalidates part of its model.

Organizational ambidexterity is not merely the coexistence of two systems. It is the ability to transfer authority between them without losing the evidence, capabilities, or people that made the transfer possible.

Speaking and Experimenting Still Do Not Guarantee Learning

The Validation Immune System depends on information that can challenge the current thesis. Psychological safety makes it more likely that people will raise mistakes, concerns, and dissenting observations.[6] Yet speaking up only creates a signal. Governance must still convert it into a changed assumption, test, or allocation decision.

Chris Argyris distinguished between correcting action within existing governing variables and questioning the governing variables themselves. Double-loop learning matters because an organization can respond energetically to disappointing results while leaving its core assumption untouched.[7] The campaign changes. The feature list grows. The sales team receives more training. The pilot is extended. Everything is revised except the belief that originally justified the investment.

That is a self-sealing system: evidence may change the activity but not the decision. A functioning Validation Immune System makes disconfirming evidence useful. Research on learning through small failures likewise suggests the value of bounded, informative losses that are limited enough to survive and meaningful enough to reveal something.[8] This is not a celebration of failure; it is a refusal to spend heavily to avoid learning cheaply.

The Diagnostic Is Complete Before the Solution Begins

An organization probably has a strong Optimization Immune System if it knows who owns delivery, which standards apply, how performance is measured, and how corrective action is enforced.

It has a functioning Validation Immune System only if it can answer an equally demanding set of questions:

• Who has authority to classify a decision as validation work?

• Which assumptions are important enough to test before more capital is committed?

• What counts as credible evidence, and who agrees to that standard in advance?

• How much is the organization prepared to lose in order to learn?

• Which result triggers scale, pivot, stop, or defer?

• When does validation end and operational accountability begin?

• What happens to the evidence, capabilities, and people if the initiative stops?

If these questions have no clear answers, the organization does not yet have two immune systems. It has one powerful system and a collection of innovation activities trying to negotiate exceptions from it.

The problem is not that the organization resists innovation. The problem is that it has no legitimate governance route for decisions that cannot yet be optimized.

Leave a comment

Apply the Two-System Test to One Consequential Decision

Consider an illustrative industrial manufacturer whose core business is selling and servicing high-value equipment. It has installed-base knowledge, customer relationships, field-service capability, and growing amounts of machine telemetry. A leadership team proposes a subscription service that would use operating data to predict failures and guarantee equipment uptime. The opportunity sounds adjacent to the core, which is precisely what makes it easy to misroute.

The board is asked to approve a three-year build program. Finance requests a five-year forecast, product specifies the platform, sales estimates installed-base conversion, and operations designs the service organization.

Every function is behaving responsibly, but the decision has entered the Optimization Immune System before the company has answered four questions:

Read more

I have changed industries several times as an innovator. I have worked in telecommunications and ICT, energy, transportation and logistics, and defense. What surprised me was not how different these industries were, but how often I encountered the same sentence:

“You are right. But this is not how we do things here.”

The first half acknowledges the logic. The second half delivers the verdict.

For a long time, I treated this as an argument that had not yet been won. I assumed that stronger reasoning, better evidence, or another experiment might change the decision. Eventually, I understood that the sentence was not primarily evaluating the proposition. It was marking the boundary of what the organization considered normal, legitimate, and safe.

The person presenting the idea had not necessarily made a logical error. They had moved outside the organization’s accepted way of thinking and were now asking others to follow. That step carried social, political, and professional risks that logic could not remove on its own.

Hired to challenge the business—until the challenge became real

In one of my last corporate innovation roles, my team and I were hired to create new business for an organization that generated around 80 percent of its revenue from one primary public-sector customer.

The existing business was substantial and stable. It rested on long-standing relationships, specialized capabilities, and multi-year service agreements. Protecting it was not irrational. It financed the organization and fulfilled real obligations.

We proposed exploring dual-use applications in private markets first. Our hypothesis was that private-sector customers would provide more accessible users, clearer budgets, and shorter decision cycles. This would allow us to investigate operational problems and mature potential applications before navigating slower public procurement processes.

We were not proposing that the organization abandon its primary customer. We were proposing a faster environment in which to learn.

The response was that serving the primary customer was the company’s mission. Our work was also described as consuming resources that operations could use and as giving our team privileges unavailable to the core business.

Some of those objections were legitimate. Exploration consumes resources without guaranteeing a return. The missing element was a governance mechanism for determining whether the proposed path could create future value.

Instead of evaluating the hypothesis, the organization gradually removed the conditions required to investigate it. Access to stakeholders became difficult. Budgets were withheld. Internal distance grew. We nevertheless ran several proofs of concept, but promising signals created another problem: leaders from organizations in adjacent industries objected that we were entering their territory.

As the political and reputational pressure increased, the activities were stopped. The whole team, including me, was dismissed.

I do not offer this as proof that our proposal would have succeeded. It might have failed for market, technical, economic, or organizational reasons. That uncertainty was precisely why it needed to be tested.

The problem was not that the hypothesis failed. The organization withdrew the conditions required to find out.

Its existing norm protected itself from being tested.

Norms are not the enemy of innovation

It would be easy to turn this into a story about conservative managers who did not understand innovation. That explanation would be emotionally satisfying and analytically weak.

Norms preserve accumulated knowledge. They coordinate work, protect customers, reduce unwanted variation, and make reliable delivery possible. In defense, transportation, energy, and other safety-sensitive industries, exceptions can have serious consequences.

Routines do not inevitably prevent adaptation either. Research on model changeovers at Toyota’s NUMMI plant showed how structured routines, participation, training, and trust could support both efficiency and flexibility.[1] Standardization can make some forms of change easier because people do not have to renegotiate the whole system every time something changes.

The problem begins when the organization cannot distinguish between an operational anomaly and an exploratory signal.

In operations, an anomaly is often a defect. It threatens reliability and should be corrected. In innovation, an anomaly may reveal changing customer behavior, an unmet need, a technological possibility, or the early breakdown of the existing business model.

Removing it too quickly can eliminate exactly the information the organization needs.

James March distinguished the exploitation of existing knowledge from the exploration of new possibilities.[2] Exploitation favors refinement, efficiency, implementation, and scale. Exploration requires search, variation, experimentation, and tolerance for negative findings.

When both compete under the same metrics, time horizons, and resource-allocation rules, exploration begins at a structural disadvantage. The current business can point to revenue, contractual obligations, established customers, and measurable performance. The future is asked to prove itself with evidence only the present can provide.

The three layers of norm defense

Across my experiences, I have seen organizations defend their existing norms through three layers.

The first is operational defense. The initiative is treated as a resource problem. It consumes money, attention, customer access, and expertise that could support the established business. Its different working conditions are described as privileges.

The second is identity defense. Once the initiative touches unfamiliar markets, customers, technologies, or business models, the objection changes. It is no longer merely expensive. It is said to conflict with the mission or with what the company fundamentally is.

The third is social defense. When the team continues challenging the first two layers, its members become the anomaly. Access disappears, colleagues create distance, and supporting the initiative begins to carry professional risk.

The innovation is first treated as inefficient, then illegitimate, and finally disloyal.

The sequence matters because the discussion progressively moves away from the evidence. What began as a hypothesis about a market becomes a judgment about identity and belonging.

Some organizations do not kill innovations when they fail. They kill them when they begin to matter.

The existing business is locally rational

Protecting an established customer is often a rational decision. Christensen and Bower showed how the demands of existing customers shape the allocation of resources inside incumbent firms. Opportunities that do not initially serve those customers can struggle to receive investment even when they may later become strategically important.[3]

Each decision may make sense in isolation. Current customers are real, current revenue is measurable, and current commitments must be fulfilled. The alternative is uncertain and often initially smaller.

The problem lies in the cumulative result. Every decision protects the present while nobody remains accountable for constructing the future.

“Our main customer generates 80 percent of our revenue” is therefore both a legitimate explanation and a warning. It explains why the relationship must be protected. It also explains why the governance of future opportunities cannot depend entirely on the system built around that relationship.

The core business was not necessarily wrong. Its logic was incomplete for the work the innovation team had been hired to perform.

A mandate is not a license to deviate

Organizations frequently give an innovation team a mandate without giving the wider organization a corresponding license to tolerate deviation.

A mandate tells the team that it may explore something new.

A license to deviate tells everyone else:

• Why the team may use different methods and performance measures

• Which customers, users, capabilities, and markets it may approach

• Which resources and access are protected

• Where the permission ends

• What evidence is required to continue

• Who resolves conflicts with the core business

• What happens when the exploration creates political discomfort

Without that second part, the team receives permission from above while remaining dependent on people whose incentives have not changed.

The result is conditional tolerance. The team may explore until its work competes for a meaningful resource, crosses an industry boundary, challenges an established customer relationship, or produces evidence that threatens the organization’s identity.

At that point, the formal mandate meets the informal veto.

This is why changing a norm cannot be delegated entirely to the people expected to violate it. The board and executive team control the resources, status, and consequences through which the organization learns what is legitimate.

If their behavior remains unchanged, the innovation mandate is an instruction to enter a conflict zone without protection.

Why reasonable people conform

People do not evaluate an innovation proposition in isolation. They evaluate it inside a network of dependencies involving colleagues, customers, authority, budgets, status, career prospects, and income.

An argument can be strategically convincing while remaining professionally dangerous to support.

Deutsch and Gerard distinguished informational influence—following a group because it appears correct—from normative influence, where people conform to remain accepted or avoid rejection.[4] This helps explain how colleagues can privately recognize an opportunity while refusing to defend it publicly.

Employee-voice research shows how unwritten rules reinforce that response. Detert and Edmondson found that employees hold implicit beliefs about when speaking up is risky or inappropriate.[5] Burris found that managers respond differently to voice that supports the current direction and voice that challenges it; challenging voice can affect how the employee is evaluated.[6]

The organization does not need an official rule against unconventional thinking. Delayed decisions, withheld introductions, inaccessible budgets, missing invitations, and visible career consequences can establish the same boundary.

Conformity is not always a lack of imagination. In a system where leadership retains the option to withdraw while employees carry the career risk, conformity can be a rational form of self-protection.

Presentations announce norms. Consequences establish them.

Social-norm research distinguishes between injunctive norms—what people are told they should do—and descriptive norms—what people observe others actually doing.[7]

This explains why transformation town halls often produce little transformation.

Leaders say that employees should challenge assumptions, act entrepreneurially, and pursue new opportunities. Daily decisions then demonstrate that existing revenue receives priority, current customers define strategic relevance, inconvenient experiments lose access, and political discomfort can overrule market evidence.

The presentation announces the desired norm. The consequences reveal the governing one.

Employees watch where the money goes, who receives protection, which objections can stop an experiment, who advances, and what happens to people who challenge the existing identity. When words and consequences conflict, consequences usually win.

Innovation becomes a credible norm only when people observe it surviving a difficult decision.

An experiment must be allowed to change the decision

An experiment is meaningful only when its possible results are allowed to change a decision.

If weak evidence can stop an initiative but strong evidence can also stop it because the result becomes politically uncomfortable, the organization has not established an evidence-led path. No possible result can earn the next commitment.

I call this self-sealing governance. The organization appears open to evidence, but every finding is interpreted in a way that preserves the existing course.

Teams conduct interviews, build proofs of concept, and present results, yet the evidence has no recognized authority to change resource allocation, customer strategy, or organizational identity. The experiment becomes a demonstration whose political consequences remain controlled.

Threat-rigidity research suggests that people and organizations may narrow information processing, centralize control, and return to familiar responses under perceived threat.[8] A recent review emphasizes that this response is conditional rather than universal, while confirming the importance of understanding when threat produces rigidity.[9]

Leaders may genuinely support innovation while it remains abstract. Their position becomes observable when the work threatens an important customer, political relationship, budget, or organizational identity.

The first serious conflict is not an interruption of the innovation process. It is the first valid test of the innovation governance.

The board kept the option. The team carried the risk.

Organizational identity directs attention and defines which opportunities appear legitimate. Tripsas showed how identity can filter how organizations notice and interpret technologies, while identity-challenging opportunities become difficult to capitalize on because identity is embedded in routines and relationships.[10]

That identity is also reinforced outside the organization. Conformity with other organizations can increase legitimacy, which helps explain why crossing an industry boundary may provoke resistance before it produces meaningful revenue.[11]

The board can commission exploration, observe what happens, and retreat when the political or reputational cost becomes uncomfortable. The innovator may lose income, status, relationships, and career continuity.

The organization retains the option. The individual absorbs the downside.

This does not mean innovation teams should receive permanent protection from evidence. Weak candidates should stop. But stopping a business candidate is different from dismantling the organization’s ability to explore.

Research on project termination shows that how organizations end work affects emotion, learning, and future career opportunities, and that redeployment and time for sense-making shape what can be retained.[12]

If evidence invalidates an opportunity, the team has generated useful knowledge. Its members can be redeployed with that learning. If leadership withdraws the mandate for political rather than evidential reasons and removes the entire team, it sends a different message: challenging the norm carries a downside the organization will not share.

The diagnosis is complete. The innovation team was not the underlying contradiction. It made the contradiction visible.

The organization asked people to create a different future while leaving the social, political, and professional consequences of deviation untouched.

Until those consequences change, most reasonable people will continue to protect the present.

The paid section provides the application: a License-to-Deviate Contract that makes the board’s authorization explicit before the first serious conflict arrives.

Leave a comment

Paid application: The License-to-Deviate Contract

Use this contract before launching an exploration that depends on access, customers, capabilities, or resources controlled by the core business. It is not a declaration of independence for the innovation team. It is a bounded governance agreement defining which deviations are legitimate, what accountability applies, and who carries the risk when conflict appears.

Apply it to one live initiative, not to “innovation” in general.

Read more

Innovation portfolios are often described as collections of bets.

The metaphor has value. It reminds executives that no amount of analysis can guarantee a successful outcome. Markets move, competitors respond, technologies change, and customers behave differently from what a business case predicted.

But the metaphor can also distort the work of innovation management.

A bet sounds like a decision made once. The organization chooses an opportunity, places capital behind it, and later discovers whether it won or lost. Governance then becomes a search for better bets: better ideas, better forecasts, better selection criteria, or leaders with better instincts.

Innovation does not unfold that way. An initiative is a sequence of decisions made as new information becomes available. The first commitment should create the conditions for a better second decision. The second should improve the third. Capital should increase only as the organization earns a stronger basis for committing it.

The central question is therefore not whether management picked a future winner at the beginning.

It is whether each investment made the next decision better than the previous one.

Outcomes are unreliable teachers

When an innovation succeeds, the original decision is usually remembered as insightful. When it fails, the decision is recast as misguided. Annie Duke calls this tendency to judge a decision by its outcome “resulting.” Outcomes reflect both decision quality and luck, so a good process can produce a bad result while a weak process occasionally succeeds.[1]

Formal decision analysis makes the same distinction. Decision quality depends on how well the problem was framed, which alternatives were considered, what information was available, how uncertainty was assessed, and whether the choice was consistent with the decision-maker’s objectives at the time.[2] None of this guarantees the desired result.

Innovation adds a deeper difficulty. Frank Knight distinguished measurable risk from uncertainty, where probabilities cannot be estimated reliably because the relevant knowledge does not yet exist.[3] Early innovation decisions often concern exactly these unknowns: whether the problem matters enough, whether behavior will change, who controls the purchase, whether the technology can perform in context, and whether the economics can work.

Detailed forecasts do not convert these uncertainties into facts. They often make untested assumptions look precise.

Once the outcome is known, hindsight makes the path appear more predictable than it was.[4] Outcome bias then leads people to evaluate otherwise identical decisions differently depending on whether they produced favorable or unfavorable results.[5] The successful initiative acquires an origin story of foresight. The unsuccessful one fills with warning signs that supposedly should have been obvious.

An organization that learns only from winners and losers therefore learns too late—and may learn the wrong lesson.

Project progress can hide decision stagnation

Because outcomes take sometime months or years to appear, organizations look for earlier signals of progress. They review whether the team completed interviews, produced a prototype, launched a pilot, met its milestones, or stayed within budget.

These are legitimate project-management questions. They are poor substitutes for innovation governance.

A prototype can be completed without testing the assumption most likely to destroy the opportunity. Fifty interviews can produce little more than favorable comments if the sample, questions, and interpretation protect the original idea. A pilot can demonstrate technical feasibility while revealing nothing about adoption, willingness to pay, or the operating model required to scale.

Activity has increased. The organization’s ability to decide may not have changed.

This confusion is reinforced by the different economics of exploration and exploitation. James March described exploitation as the refinement of what is already known and exploration as the search for new possibilities.[6] Established operations can be evaluated through delivery, efficiency, revenue, and variance from plan. Early innovation cannot, because the assumptions that would make those measures meaningful are still being investigated.

When both are governed through project progress, exploration begins to imitate execution. Teams produce plans and artifacts that make the initiative look increasingly real. Momentum grows faster than evidence, and stopping becomes harder precisely when the organization should still be preserving flexibility.

Stage gates should govern commitments, not presentations

Stage-Gate was not originally designed as a bureaucratic checklist. Cooper’s model divided new-product work into stages that generate information and gates that decide whether further resources should be committed.[7] Later versions emphasized adaptability, different pathways, spiral development, and “gates with teeth,” while warning against rigid and over-bureaucratic implementation.[8]

The problem is not necessarily Stage-Gate theory. It is what many gates reward in practice.

Teams arrive with polished slides, prototypes, roadmaps, financial projections, and evidence summaries. Gatekeepers assess whether required deliverables exist and whether the presentation supports continuation. The initiative receives another tranche because it appears to have progressed.

Yet the governance question is not whether the team has been busy or persuasive. It is whether the evidence generated since the previous gate justifies exposing more capital, credibility, and organizational energy.

Discovery-Driven Planning made assumptions explicit and treated plans as hypotheses rather than facts.[9] Real-options reasoning showed why small, staged commitments can preserve the right—but not the obligation—to invest more after uncertainty has been reduced.[10]

Both point toward the same principle:

Knowledge should increase before commitment increases.

That principle needs an operational condition at the gate. I call it Decision Readiness.

Decision Readiness is the degree to which decision-relevant uncertainties have been made explicit and investigated with signals sufficient to justify a specified next commitment.

It does not mean that uncertainty has disappeared. Nor does it promise a successful outcome. It asks whether the organization has reduced the avoidable ignorance relevant to the decision before it increases its exposure.

What Decision Readiness requires

An initiative is ready for its next decision when governance can see six things clearly:

1. The decision: What commitment is being requested now—not the eventual ambition, but the next release of money, people, access, or reputation.

2. The theory of value: Why this initiative is expected to create value, for whom, and through which change in behavior or economics.

3. The critical assumptions: What must be true for that theory and the next commitment to remain defensible.

4. The evidence change: What the organization knows now that it did not know at the previous gate, including contradictory observations.

5. The decision thresholds: Which findings support scaling, pivoting, stopping, or deferring.

6. The proportional commitment: Why the size and reversibility of the next tranche are appropriate to the evidence available.

If these conditions are absent, another deliverable will not repair the decision. The project may be ready for its next activity without being ready for its next investment.

This also changes how stopping should be interpreted. If a €40,000 discovery stage produces credible evidence that a planned €3 million program should not proceed, nothing has launched and no revenue has been created. Yet the stage may have generated substantial governance value by preventing a much larger, weakly supported commitment.

The return is not failure avoided with certainty; that claim would be impossible to prove. The return is a materially better capital decision made while the downside was still contained.

The question every gate should ask

The most revealing gate question is not:

Did the team deliver what it promised?

It is:

Knowing what we know today, would we make the same next commitment again?

If the answer is yes, governance should be able to show which evidence supports the commitment and why the next tranche is proportionate.

If the answer is no, the organization should not conceal that learning behind project momentum. It should pivot, stop, or redefine the commitment.

If the evidence is insufficient, it should defer the decision through a bounded investigation with a date, cost ceiling, and explicit learning objective—not extend the initiative by default.

Innovation governance cannot remove luck. It can reduce how much capital is exposed to assumptions that could have been investigated earlier.

That is why innovation management is not primarily about making better bets. It is about earning better decisions before each larger commitment is made.

If this diagnosis is useful, the paid section provides the application: a Decision Readiness Gate you can use on one live initiative within the next seven days.

Share

The Decision Readiness Gate

Use this gate before approving a new tranche for an initiative whose commercial outcome remains uncertain. It is not a project scorecard. Its purpose is to expose whether the proposed commitment is supported by decision-relevant evidence.

Start with one real decision. Do not apply it to the portfolio in general.

Read more

A salesperson reports that customers have stopped responding to the company’s value proposition. A support representative notices clients constructing the same workaround, again and again. An account manager sees purchasing authority shift to a stakeholder the product was not designed for. An operations employee watches a process become more expensive while delivering less.

They raise it.

Their managers listen. Nobody is punished. The employees are thanked for their candor. The meeting may even be remembered as evidence of a healthy culture.

Then it ends.

The targets do not move. The roadmap continues. The budget stays where it was. The next executive presentation describes the market using the same assumptions as before.

The information was accepted socially and rejected operationally.

This organization may be safe enough for people to speak. It is not yet capable of learning from what they say.

Psychological safety makes learning possible. It does not complete it.

Amy Edmondson defined team psychological safety as a shared belief that a team is safe for interpersonal risk-taking. Her research connected it to learning behaviors such as discussing errors, seeking feedback, asking for help, and experimenting.[1] Later reviews have reinforced its importance for voice, teamwork, and organizational learning.[2]

That contribution is difficult to overstate. In many organizations, admitting uncertainty or challenging authority can still carry reputational and career risk. If people expect embarrassment, punishment, or exclusion, important information will remain private.

Yet psychological safety solves a specific problem: whether a person can take an interpersonal risk. It does not, by itself, determine what the organization does with the information that emerges.

This distinction matters because voice and learning are not synonyms.

Voice is an act: someone communicates an idea, concern, observation, or challenge.

Learning is an update: the organization changes what it believes or what it does because new evidence has altered its understanding.

The first can happen without the second. A leadership team can encourage disagreement, run candid retrospectives, and score well on a safety survey while leaving every consequential commitment untouched. The conversation changes. The decision system does not.

Psychological safety is therefore an input to learning, not proof that learning occurred.

The hidden failure is safe futility

Research on employee voice and silence shows that people consider more than the danger of speaking. They also consider whether speaking is likely to make a difference.[3] The practical questions are simple:

Is it safe to raise this?

Will anything happen if I do?

Organizations have become more attentive to the first question. They train managers to listen without defensiveness. They create surveys, retrospectives, town halls, ethics channels, and escalation paths. These mechanisms can be valuable.

But the second question is easier to neglect.

An organization can reduce fear while preserving futility. Employees are invited to report what they see, but no process connects their observations to the assumptions behind a strategy, the evidence required to revise it, or the authority to change the allocation of resources.

The feedback mechanism then becomes a container. It absorbs disagreement without allowing disagreement to touch commitment.

The employee is heard. The assumption is not examined.

The concern is logged. The project remains protected.

The risk appears on a slide. The funding continues automatically.

The organization has built a pressure-release valve, not a learning system.

Morrison and Milliken described organizational silence as a collective pattern in which employees withhold information about potential problems, limiting an organization’s capacity for change and development.[4] Safe futility is a less visible route to the same destination. People may speak at first, but they learn from repeated inaction that the expected value of doing so is close to zero.

Eventually, candor becomes ceremonial. Employees continue attending the meetings while becoming more selective about what they genuinely challenge.

Silence returns, not because speaking is forbidden, but because experience has taught people that speaking is inconsequential.

A safe meeting can sit inside a defensive organization

Most leaders support employee voice in principle. The harder test comes when the message threatens a commitment they helped create.

Research shows that perceived managerial openness is strongly related to whether employees speak up.[5] It also shows that employees carry implicit rules about when voice is risky or inappropriate: do not bypass the boss, do not challenge authority in public, do not raise a problem without already having the solution, and do not speak outside your formal area of responsibility.[6]

The content of the message matters as well. Managers tend to respond more favorably to voice that supports the current direction than to voice that challenges it.[7]

“We should improve the sales script” asks for better execution.

“Customers no longer value the problem we built the product to solve” challenges the strategy.

“We need more training” protects the operating model.

“The operating model makes the behavior we want economically irrational” challenges it.

The first type of voice is easier to welcome because it leaves the governing assumptions intact. The second creates a threat: if the observation is correct, targets may need to change, forecasts may need to be revised, investment may need to move, and leaders may need to explain why the previous view no longer holds.

That is where an apparently open organization can become defensive.

Defensiveness rarely announces itself as a refusal to learn. It arrives as a plausible explanation. The signal is too early. The sample is too small. Customers do not know what they want. Sales is positioning the offer incorrectly. The market needs more education. The team needs another quarter.

Any of these explanations may be right. The warning sign is not their existence, but their asymmetric use. Evidence that supports the plan is treated as confirmation. Evidence that threatens it is treated as an exception requiring a much higher standard of proof.

Chris Argyris distinguished between single-loop learning, which corrects execution while preserving governing assumptions, and double-loop learning, which examines whether those assumptions, goals, or policies remain valid.[8] Defensive governance has a reliable way to avoid the second loop: it translates strategic contradictions into execution problems.

Declining demand becomes a sales-discipline problem. Repeated customer workarounds become a training problem. Weak adoption becomes a communication problem. Deteriorating unit economics become a scale problem.

The explanation keeps changing so the decision does not have to.

Information can disappear at five different handoffs

The path from speaking up to learning is not one event. It is a chain:

Voice → attention → interpretation → test → decision → allocation

Psychological safety mainly strengthens the first link. The remaining links are governed by attention, incentives, authority, process, and resource allocation.

1. Voice without attention

A concern is raised but never enters a forum with the authority to act on it. It stays inside a team retrospective, employee survey, customer-success report, or risk log.

The organization can honestly say the issue was surfaced while the relevant decision-makers never have to confront it.

2. Attention without interpretation

Leaders hear the signal but do not connect it to a specific belief. The discussion remains at the level of impressions: sales is concerned, support is frustrated, customers seem hesitant.

Without identifying the assumption under pressure, the conversation becomes a contest between opinions.

3. Interpretation without a test

The team agrees that an assumption may be wrong but does not design a way to distinguish among competing explanations.

The issue becomes an open question that can remain open indefinitely.

4. A test without a decision threshold

The organization gathers more data but has not agreed what different results would mean. When the evidence arrives, it is interpreted after the fact.

Every outcome can then be made compatible with continuing the current plan.

5. A decision without allocation

Leaders announce that they have learned, but the roadmap, staffing, targets, and capital remain unchanged.

The language updates while the commitment does not.

This last handoff is decisive. An organization’s real beliefs are visible less in what leaders say than in what they continue to fund.

The edge often sees change before the center can measure it

People close to customers, suppliers, operations, and service failures frequently encounter change before it becomes legible in an executive dashboard.

Salespeople hear new objections. Support teams see recurring workarounds. Field staff observe how a product is actually used. Procurement sees supplier behavior shift. Operations notices where formal processes are quietly bypassed.

Research on frontline sensing argues that these observations can provide early information for strategic decisions because frontline employees participate in the daily transactions through which changing conditions first become visible.[9]

This does not mean the edge is automatically right. Frontline employees have partial views, local incentives, and their own interpretive biases. Executives do too; their information is simply more aggregated, delayed, and filtered.

The governance advantage comes from treating frontline observations neither as truth nor as anecdote, but as signals that can be converted into questions:

• What exactly was observed?

• Which current assumption would be weakened if the observation were representative?

• What rival explanations could also produce it?

• What is the cheapest credible test that would separate those explanations?

• Which pending decision depends on the answer?

This shifts the conversation away from whether the employee can prove the entire strategic case. A person should not need a fully developed solution before the organization becomes willing to investigate a contradiction.

Requiring that level of proof gives the existing plan an unfair advantage: the plan receives resources, analysts, and executive sponsorship, while the challenge must arrive complete.

A weak signal is not a verdict. It is a reason to investigate.

Endorsement is not implementation

Managers can sincerely agree with employee input and still be unable to act. Research distinguishes between endorsing voice and implementing it; implementation depends partly on motivation, felt obligation, perceived control, and the surrounding network of relationships.[10]

That gap is easy to recognize in practice.

A line manager listens but cannot alter the product roadmap. A product leader accepts the evidence but remains locked into annual targets. An executive sees that the forecast rests on a weakening assumption but fears the consequences of revising guidance. A board asks for candor while rewarding predictability and treating deviation from plan as failure.

At that point, the problem cannot be solved by asking the direct manager to listen more skillfully. The information has reached the boundary of that person’s authority.

Voice becomes strategically useful only when it is attached to a decision system with four properties:

• A named decision that the signal could affect

• An owner with authority to investigate and act

• A date by which the evidence will be reviewed

• An explicit rule for what happens under different results

Without these, leaders can endorse the message while the organization rejects its implications.

The test of learning is observable change

Not every concern should change the strategy. Psychological safety does not mean that every challenge is correct, that all evidence is equally strong, or that leaders must follow every suggestion.

It does mean that material challenges deserve a fair route to resolution.

The strongest evidence of a learning organization is not the number of ideas submitted, the volume of discussion, or the score on a speaking-up survey. It is whether new information can produce an observable update.

That update should appear in at least one of four places:

1. Assumption: What the organization believes about customers, value, behavior, capabilities, economics, or the environment.

2. Test: What the organization will do next to reduce a consequential uncertainty.

3. Decision: Whether to scale, pivot, stop, or defer a commitment.

4. Allocation: Where time, attention, people, and capital go as a result.

If none of these changes, the organization may have communicated. It has not demonstrated learning.

The board-level question is not whether people feel heard

Feeling heard matters because it affects whether people will contribute again. But a board cannot infer adaptability from that feeling alone.

The sharper question is whether uncomfortable information can change what the organization is committed to.

A board or executive team should be able to answer:

• What new evidence has challenged a material assumption since the last review?

• Which signals from customers, sales, support, operations, or partners were investigated?

• What alternative explanations were tested?

• Which assumption, test, decision, or allocation changed?

• What evidence would cause us to scale, pivot, stop, or defer the initiatives currently under review?

• What happened to the last three material concerns raised inside the organization?

These questions expose the difference between a culture that permits voice and a governance system that converts voice into adaptation.

The diagnosis is complete: psychological safety makes uncomfortable information more available. Learning requires governance to make that information consequential.

Share

The Signal-to-Decision Protocol

The purpose of this protocol is not to reward every person who raises a concern by accepting their interpretation. It is to prevent a consequential signal from disappearing before the organization has learned what it means.

Use it when someone surfaces information that could change a material commitment: a product investment, market entry, transformation program, acquisition thesis, operating-model change, or strategic partnership.

The protocol begins with a real decision, not a general discussion.

Read more

Every struggling innovation project eventually becomes good at explaining why it should continue.

A missed milestone becomes a temporary resource issue. Weak customer interest becomes a positioning problem. Low willingness to pay becomes a pricing issue. A disappointing pilot is declared unrepresentative, while a new competitor is treated as proof that the opportunity must be real.

None of these explanations is necessarily false. Each may be reasonable on its own. The problem is their direction. They all lead to the same conclusion: give the project more time, more money or another chance.

I think of this as the press secretary in our heads—and, eventually, in our organizations. Its job is not to determine what caused the result. Its job is to produce an account that sounds coherent, defensible and compatible with what we already want to do.

The spokesperson is not the decision-maker

The metaphor comes from Robert Kurzban’s work on the modular mind. In Why Everyone (Else) Is a Hypocrite, Kurzban argues that the mind is not a single, unified decision-maker with complete access to its own motives. It contains specialized systems whose interests and conclusions do not always align. The conscious explanation can therefore resemble a press secretary: it presents a coherent public account without necessarily having access to the processes that produced the behavior.[1]

This is best treated as a useful model rather than a precise neurological description. Its value lies in the distinction between producing a decision and explaining one. The explanation may feel like an investigation even when it is closer to representation.

Ziva Kunda’s research on motivated reasoning supports the distinction from another direction. People seeking a preferred conclusion do not usually abandon reason. They access, construct and evaluate beliefs in ways that make the preferred conclusion easier to reach, while remaining constrained by the need to produce a justification that appears reasonable.[2]

Kevin Simler and Robin Hanson extend the argument to hidden motives. The reasons people publicly give for their behavior may coexist with less visible motives concerning status, signaling or self-presentation. We can sincerely believe the public account because introspection does not automatically reveal everything that shaped the decision.[3]

The press secretary therefore does not need to lie. It only needs enough ambiguity to select one plausible explanation from several.

Projects develop press secretaries too

Innovation projects reproduce the same pattern collectively.

Nobody appoints the spokesperson. It emerges through steering-committee decks, status reports, dashboards and the repeated need to explain why the expected results have not appeared. A weak signal enters the organization as evidence and leaves as a story about why the project remains promising.

This does not require manipulation. Product leaders, sponsors, finance teams and executives can all interpret the same evidence sincerely while protecting different interests. The sponsor protects the original decision. The team protects its work and identity. The executive protects a strategic narrative already communicated upward. Finance protects the credibility of earlier allocations.

The resulting story belongs to no single person, yet it has a consistent policy: continue.

This is why asking whether an explanation is true is often insufficient. Several explanations may be compatible with the available evidence. The more important question is whether the organization is testing among competing explanations or selecting the one that preserves the existing commitment.

Continuation protects more than the project

Recommending that a project stop creates an asymmetric personal risk.

If the project would later have succeeded, the person who argued against it may be remembered as the one who nearly killed the opportunity. Recommending another quarter is safer. It preserves relationships, avoids challenging the original sponsor and postpones the point at which someone must admit that the initial decision may have been wrong.

The dynamic becomes stronger when the people reviewing the project are the same people who authorized it. Barry Staw’s classic experiment on escalation of commitment found that decision-makers who felt personally responsible for an earlier choice allocated more resources to it after receiving negative feedback.[4] Continuing did not merely protect the project. It protected the meaning of the original decision.

Prospect Theory provides a complementary mechanism. Kahneman and Tversky showed that people evaluate choices relative to a reference point and may become more willing to accept risk when facing losses.[5] Applied cautiously to innovation investment, stopping can feel like realizing a loss, while continuing preserves a possibility—however uncertain—of recovering it.

The financial case for the project may be weakening at the same time as the psychological case for one more investment becomes stronger.

Organizations often call this commitment. Sometimes it is. Sometimes it is self-justification with a budget line, and the two are difficult to distinguish from inside the room.

Private doubt becomes public confidence

The press secretary does not only protect individual judgment. It also solves a coordination problem.

Several people may privately doubt the project while each assumes that everyone else still supports it. Speaking first carries personal risk. Silence is safer, particularly when the sponsor has more status or the group has already communicated confidence publicly.

Morrison and Milliken describe organizational silence as a collective phenomenon in which employees withhold information about potential problems because speaking appears risky or futile.[6] Irving Janis’s work on groupthink describes a related failure: groups committed to cohesion and consensus can suppress dissent and discount contradictory evidence.[7]

The result is an organization that appears more confident than its members actually are. People update their private beliefs while the public narrative remains unchanged.

What looks like alignment may therefore be a measurement error. The meeting records the story people are willing to defend, not necessarily the judgment they privately hold.

Evidence does not speak—and it does not decide

This is why calling an organization evidence-driven tells us very little.

Evidence does not interpret itself. A lower conversion rate may indicate a weak value proposition, the wrong segment, poor execution, a flawed channel, insufficient time or a measurement problem. The data alone cannot decide among these explanations.

The important capability is not collecting evidence but allowing it to change a decision. Many organizations reward visible confidence, commitment and delivery more reliably than they reward someone for disproving an expensive assumption early.

Eric Ries built the Lean Startup around experiments that reduce uncertainty before more resources are committed, with evidence informing whether to persevere or pivot.[8] The organizational press secretary reverses that logic. Evidence stops testing the project and starts supplying material for why it should continue.

A dashboard full of green metrics may mean that the underlying assumptions are holding. It may also mean that the team has become skilled at selecting metrics that remain green.

Psychological safety helps, but it does not solve this problem alone. Amy Edmondson’s research shows that psychological safety is associated with learning behavior because people can discuss errors, ask for help and take interpersonal risks.[9] That makes difficult evidence speakable. It does not automatically make the evidence consequential.

I have sat in rooms where doubts were expressed clearly and without punishment, yet the project continued because the meeting had no mechanism for translating a concern into a capital decision. Psychological safety had done its job. Governance had not.

The governance failure happens before the pilot

The press secretary gains influence when nobody defines in advance what the evidence must change.

Ask a steering committee which result would make it stop a project and the answer often becomes vague: the numbers will be reviewed next quarter, the market needs more time or the team will know when it sees the signal.

Without a pre-committed threshold, every result remains negotiable. Continuing is the default. Stopping requires an active decision that can be challenged, delayed and reframed.

The team does not need to distort the evidence. It only needs the decision rules to remain soft enough that almost any outcome can be interpreted as encouragement.

Organizations would not accept a capital-expenditure threshold invented after the money had been spent. Yet innovation projects routinely define their stop conditions after a pilot has disappointed. At that point, the threshold is no longer an independent decision rule. It is part of the negotiation over whether the existing commitment should be protected.

The fix is not more analysis. It is better sequencing.

A threshold agreed before the pilot is different from the same threshold proposed after the result, even when the number is identical. The first was set before people knew which conclusion it would support. The second is proposed by people who already know what it will do to their project.

The board-level question

For executives and boards, the practical question is not whether the story supporting a project sounds plausible. Most continuation stories do.

Ask instead who defined the conditions under which the project would scale, pivot, stop or defer—and when those conditions were set relative to the capital already committed.

Then ask the harder counterfactual: If this initiative did not already exist, would we fund its next tranche today, knowing what we now know?

The question is not a complete decision rule. Stopping may carry costs, and an existing project may contain option value that a new proposal would not. But the counterfactual exposes how much of the current case depends on future value and how much depends on defending the past.

Evidence is not decision-grade because it is quantitative, recent or displayed on a dashboard. It becomes decision-grade when it is tied to a prior assumption, compared with an explicit threshold and allowed to change the allocation of capital.

Without that structure, the project is not being governed by evidence. The evidence is being recruited by the press secretary.

Leave a comment

The Press Secretary Protocol

Read more

The exercise is commonly associated with Tom Wujec because his 2010 TED talk made it widely known. Its origin predates that talk, and Peter Skillman is generally credited with originating it. Skillman says he began running what he called the Design Challenge with IDEO clients, later used it in university lectures and helped introduce it into corporate Design Thinking training through the IDEO University curriculum. He also notes that the exercise evolved with IDEO colleagues Dennis Boyle and Christine Kurjan and that he no longer remembers exactly how its first version arose. He presented it at TED in 2006. What is clear from his account is its original purpose: it was a tool for teaching the value of iteration and revealing how status, ego, vulnerability and interaction affect team performance.[1]

Wujec popularized the exercise and extended its meaning. His materials describe it as a way to help teams build rapid prototype solutions and learn about collaboration, creativity and innovation. They even invite organizations to use it to consider what would dramatically increase innovation and describe the early testing of hidden project assumptions as the mechanism that produces effective innovation.[2]

This evolution is understandable. The exercise demonstrates several behaviors that innovation needs: teams experiment, expose a hidden assumption, learn through interaction and adapt before time runs out. In that sense, it is partly an exercise in innovative thinking.

The problem begins when it is treated as a model of innovation as a whole. It demonstrates how to explore solutions after the challenge has been framed. It does not demonstrate how to discover whether the brief, customer, outcome, metric or decision to build should exist in the first place.

Give a team twenty sticks of spaghetti, one meter of tape, one meter of string and one marshmallow. Ask them to build the tallest freestanding structure in eighteen minutes, with the marshmallow on top.[1]

The Marshmallow Challenge produces a reliable contrast. Many adult teams spend most of their time discussing the ideal structure, allocate tasks and begin building once they believe they have a plan. They test the marshmallow only near the end, when its weight reveals that their imagined structure and the physical one are not the same.

Children and architects tend to do better. They put the marshmallow on early, build small structures, watch them fail and adapt while they still have time. They do not eliminate uncertainty through discussion. They interact with it.

This is why I continue to use the exercise. It makes the difference between planning and learning physically visible. Explore before you commit is a better response to solution uncertainty than plan first and execute later.

But innovation does not start when a team begins building a solution. It starts earlier, when the team investigates whose situation should improve, which outcome matters, what may cause that outcome and whether a solution should be built at all. The Marshmallow Challenge starts after those decisions have already been made.

Before anyone touches the spaghetti, the problem has been defined, the desired output has been specified, the resources have been allocated and the success metric has been fixed. The facilitator has decided that height matters, that every team must work with its own materials and that the teams are competitors. Participants may question how to build the tower. They are not allowed to question the frame around it.

That makes the Marshmallow Challenge an excellent exercise in iterative execution. It does not make it a complete innovation exercise.

The tower is an output, not an outcome

The winning structure must be tall and remain standing long enough to be measured. Nobody has to use it. Nobody needs to make progress because it exists. There is no customer who values height more than stability, portability, surface area or cost. Height matters because the facilitator can measure it with a tape, not because anyone has shown that it is valuable.

That distinction is easy to miss because the activity looks like innovation. Teams build, experiment, collaborate and work under uncertainty. Yet the uncertainty is contained inside a task whose purpose and boundaries are treated as facts.

Several questions remain outside the room:

• Who is this structure for?

• What are they trying to achieve?

• Is height the outcome that matters, or merely the easiest output to measure?

• What would make them adopt, use or pay for the result?

• Which other actors must participate for the outcome to occur?

• Would combining resources with another team create more value than competing with it?

These are not secondary questions to ask after the tower stands. They determine whether the tower is worth building at all.

Innovation contains three different uncertainties

Organizations often speak about uncertainty as though it were one problem. It is more useful to separate at least three kinds.

Discovery uncertainty concerns the opportunity. Who experiences the problem, in which context, how important is it, what progress are they seeking and which assumptions support our belief that an opportunity exists?

Solution uncertainty concerns the response. If the opportunity is real, which intervention could create the required outcome, what will people adopt and how must it fit into their existing behavior and system?

Execution uncertainty concerns delivery. Once the problem and solution are sufficiently understood, can the organization build, operate and scale the solution at the required quality, speed and cost within a business model that creates enough value for all the involved parties?

Each uncertainty requires a different capability. Discovery needs observation, problem framing and evidence about what matters. Solution development needs design, comparison and behavioral tests. Execution needs engineering, coordination and reliable delivery.

The Marshmallow Challenge operates mostly in the last two categories. The opportunity, user, metric and rules are fixed. What remains uncertain is which structure will work and whether the team can build it in time.

Children and architects do not win because they have discovered a better customer or a more important outcome. They win because they explore possible solutions within the given frame and expose structural failure earlier than teams that rely on planning.

That is a valuable capability. It is simply not the whole innovation job.

Most innovation projects are set up the same way

The limitation matters because many corporate innovation programs and startups reproduce the same structure without noticing it.

A team receives a budget, a deadline and a mandate to build an app, introduce an AI assistant, create a platform or enter a market. The mandate already contains assumptions about the customer, the problem and the appropriate response. A roadmap then converts those assumptions into milestones, and progress is measured through delivery: prototypes completed, features released, users acquired or pilots launched.

The team may work in sprints and test early versions. It may be highly agile inside the assignment. But agility does not make the original frame true.

Perhaps the selected users do not experience the problem strongly enough to change their behavior. Perhaps the metric being optimized is easy to count but weakly related to customer value. Perhaps the apparent customer cannot decide alone because procurement, operations, regulators or channel partners can prevent adoption. Perhaps an adjacent team, supplier or external partner has capabilities that make a joint model more credible than an internal build.

If these possibilities were excluded before the project began, the team is not discovering an opportunity. It is improving its execution of an inherited belief.

The organization then judges the team on how well it builds a tower whose purpose, user and rules it was never allowed to question.

Agile cannot rescue the wrong frame

Steve Blank distinguished a startup searching for a repeatable business model from an established organization executing one it already understands.[3] Eric Ries translated this search logic into Build-Measure-Learn.[4] Both challenged the assumption that detailed planning can remove uncertainty before contact with reality.

The Marshmallow Challenge demonstrates this argument well. Early prototypes reveal something that analysis alone cannot: the marshmallow changes the structure.

But faster iteration is only useful when the experiment addresses the uncertainty that matters. A team can build, measure and learn repeatedly while leaving the customer, outcome and business-model assumptions untouched. It becomes faster at learning inside the wrong frame.

Research on scientific decision-making by entrepreneurs shows why the framing step matters. In a large-scale replication and extension, entrepreneurs trained to articulate theories, make assumptions explicit and design tests around them were more likely to terminate weak projects earlier, make more focused pivots and perform better.[5] The improvement did not come from generating more ideas or running more activity. It came from connecting experiments to explicit beliefs and decisions.

Related work on theory-driven strategic decisions makes a similar distinction. Under uncertainty, the first task is not to select a plan of action but to develop a theory of value: an explicit explanation of why a particular action should produce a valuable outcome. That theory can then be challenged with evidence.[6]

The Marshmallow Challenge begins with the theory already embedded in its rules: tall towers are valuable, isolated teams are the unit of action and the winner is whoever produces the highest output in the given time. Participants test structures, not that theory.

Commitment changes how evidence is treated

This omission becomes expensive once a project gathers momentum.

Budgets, teams, roadmaps, executive sponsors and public promises do more than support delivery. They create attachment to the original decision. As commitment grows, evidence is no longer interpreted neutrally. A negative signal threatens previous investments, internal credibility and sometimes professional identity.

Research on escalation of commitment has documented this pattern across organizational decisions.[7] When a project becomes difficult to abandon, teams often respond to disappointing evidence by defending, reinterpreting or extending the commitment rather than revisiting its underlying assumptions.

The longer a team has been building its tower, the harder it becomes to ask whether height was ever the point.

This is why discovery must precede large commitments. Its purpose is not to prove that an idea will succeed. It is to expose the assumptions that could make the commitment irrational while changing direction is still relatively cheap.

The timing matters because early research is not expensive, but late research is. Once a team has committed to a solution, research must also overcome attachment, sunk costs and internal expectations.

Keep the exercise, but change the conclusion

I will continue to use the Marshmallow Challenge. It remains one of the clearest ways to show why teams should test the load-bearing assumption early, why a prototype can reveal more than a planning discussion and why iteration beats a single late attempt.

I will no longer present it as a model of innovation as a whole.

It teaches teams to explore before committing to a solution, but only after someone else has committed them to a customer, an outcome, a metric, a resource boundary and a competitive structure. Those upstream decisions often determine the fate of an innovation before execution quality becomes decisive.

Real innovation begins one step earlier. Before asking how to build the tower, a team must make the frame around it visible and testable: who is it for, what outcome matters, which evidence would justify building and whether the organization should build alone.

The tower was never the hardest part, understanding why you should build beforehand is.

If your team is already building inside an inherited frame, the Start, Pivot, Stop or Scale workshop helps make the assumptions, available evidence and decision criteria explicit before further resources are committed.

Leave a comment

If your team needs to make upstream assumptions explicit before prototyping, my innovation training and workshops use live initiatives to connect discovery, evidence and the next decision.

The Upstream Decision Test

Read more

That interpretation begins too late.

Adobe did not replace one business model once. It repeatedly changed whichever element was limiting its ability to create, deliver or capture more value.

PostScript established an OEM-licensed printing standard. PDF extended Adobe into universal document exchange. Creative Suite integrated previously separate tools. Omniture added analytics and marketing. Creative Cloud created a recurring relationship. Behance added a creator community. Document Cloud expanded PDF into workflows. Magento connected content with commerce. Express opened creation to non-professionals. Firefly and GenStudio are now moving Adobe towards AI-enabled enterprise content production.

Each adaptation strengthened Adobe’s position in the customer workflow. Each also increased the customer’s dependence on Adobe.

The familiar story begins too late

The familiar Adobe story starts with Creative Cloud. In 2012, the company began moving customers from perpetual licenses and periodic upgrades towards recurring subscriptions and continuous software delivery.

It was a consequential and controversial decision. But it was already Adobe’s sixth major business-model adaptation in this chronology.

Adobe’s original commercial model was built around licensing PostScript to printer manufacturers. Acrobat and PDF then expanded the value proposition from reliable printed output to exchanging digital documents across different systems. The free Reader reduced adoption friction and helped PDF become more valuable to those paying to create and manage documents.

Creative Suite integrated specialist applications into a broader workflow. Macromedia expanded that workflow into web and interactive creation. Omniture moved Adobe beyond producing content and towards measuring and optimising what that content achieved.

Creative Cloud did not begin Adobe’s transformation. It changed the constraint again.

Periodic product releases and upgrade decisions were replaced by continuous delivery and recurring access. Revenue became more predictable, but Adobe also became responsible for maintaining an ongoing service and continually demonstrating value.

The move changed far more than the price.

Four phases of adaptation

1. Establish standards and reduce adoption friction

PostScript created value by connecting computers, software, fonts and printers. Adobe did not need to manufacture the entire system. It occupied an important position between its components and captured value through OEM licensing and royalties.

PDF expanded that position. Instead of focusing only on producing printed output, Adobe enabled documents to retain their appearance across systems.

The free Reader was central to the model. It reduced the friction of receiving and viewing a PDF, making the format more useful to those paying for document creation and professional capabilities.

Free access on one side supported paid value on the other.

2. Integrate and broaden the creative workflow

Creative Suite combined specialist applications into a broader offering. This increased the amount of the creative workflow Adobe could support and the value it could capture from each customer.

The Macromedia acquisition added Flash, Dreamweaver and other interactive-media capabilities. Adobe was no longer concentrated primarily on print, imaging and documents. It expanded with the changing forms of digital content.

Integration improved the customer workflow because tools could be used together. It also increased switching costs because the customer relationship no longer depended on one replaceable application.

Adobe was becoming more valuable partly because more of the customer’s work happened inside its ecosystem.

3. Turn products into recurring relationships and workflows

Omniture moved Adobe into analytics, marketing and hosted services. This introduced new buyers, including marketing departments and large enterprises, and required capabilities in data, cloud operations and enterprise sales.

Creative Cloud then changed payment, delivery and the customer relationship. Instead of selling a major release and later convincing customers to upgrade, Adobe could deliver improvements continuously and maintain a direct recurring relationship.

Behance extended that relationship beyond the use of software by adding portfolios, professional discovery and community. Document Cloud expanded PDF from a file format and desktop product into mobile access, storage, collaboration and electronic-signature workflows.

Adobe Spark, later relaunched as Adobe Express, created a freemium entry point for first-time creators, communicators, students and small businesses. Magento extended Adobe’s enterprise portfolio into commerce and transaction workflows.

These were different moves, but they followed the same direction: Adobe kept expanding from the individual tool into the surrounding work.

4. Move from proprietary tools towards governed AI workflows

Firefly embedded generative AI inside the applications Adobe customers were already using. Generative credits also introduced metered usage within subscription plans.

This was a business-model response as much as a product response. Generative AI creates variable computing costs that a conventional software subscription does not automatically accommodate. Credits allowed Adobe to connect usage with commercial limits without abandoning the subscription relationship.

GenStudio extended the proposition further into the enterprise content supply chain: planning, creation, brand governance, activation and measurement.

Adobe then began adding partner models from companies including OpenAI and Google to Firefly. This changed the platform logic again. Adobe no longer had to depend entirely on its proprietary models being superior for every task. It could compete as the trusted workflow and governance layer through which customers access different AI capabilities.

The individual model may become interchangeable. The surrounding workflow is harder to replace.

Three lessons from the case

1. Business-model innovation can be cumulative

Not every Adobe adaptation replaced the previous model. Many changed only one or two elements while preserving capabilities, customer relationships, standards or assets created earlier.

PostScript established Adobe inside the publishing ecosystem. PDF expanded the job. Creative Suite increased the share of the workflow. Creative Cloud changed delivery and revenue capture. Later moves added community, analytics, commerce, collaboration and AI.

Continuous business-model innovation is therefore not necessarily a sequence of dramatic pivots. It can be the repeated reconfiguration of selected elements before the existing configuration becomes a constraint.

Netflix’s business-model evolution provides a useful comparison. Both companies repeatedly changed parts of their models, but the constraints differed: Adobe expanded its role in customer workflows, while Netflix adapted distribution, content ownership and revenue capture.

2. A new revenue model requires a different operating model

Creative Cloud is often reduced to subscription pricing. But recurring billing alone would not have created recurring value.

The change also required continuous product delivery, cloud infrastructure, account management, direct customer relationships and a different development rhythm. Omniture, Document Cloud, Magento, Firefly and GenStudio required additional capabilities in analytics, enterprise sales, commerce, AI computing, governance and workflow orchestration.

Changing how a company captures value without changing how it creates and delivers that value produces a commercial mechanism, not a coherent business model.

This is why business model validation must test the whole system, including the connections between customer value, delivery, capabilities, revenue and economics.

3. Customer value and customer dependence can grow together

Adobe’s integrations can make work easier. Files move between applications, teams share assets, organisations manage permissions, and enterprises connect creation with approval, activation and measurement.

The same integrations make Adobe more difficult to leave.

Files, skills, subscriptions, workflows, shared assets, communities, governance rules and enterprise systems accumulate around the ecosystem. Exiting may require more than replacing a product. It may require changing how work is organised.

This creates the strategic tension at the centre of the case:

Each move expanded customer value and revenue predictability, but also increased customer dependence on Adobe. The model became stronger partly by making exit, substitution and ownership less simple.

The relevant distinction is not simply between integration and lock-in. It is between retention created by value customers repeatedly choose and retention created by the cost of leaving.

Both may produce similar commercial results. They indicate different qualities of business-model strength.

Future cases will examine other organizations that changed their business models as technologies, customer behavior and old assumptions became constraints. Subscribe for source-backed cases, strategic interpretation and practical decision tools.

2 Adobe Continuous Business Model Innovation Final

54.4KB ∙ PDF file

Download

A two-page, source-backed chronology of fourteen business-model adaptations, including a Business Model Canvas View, Core Pattern and Strategic Tension.

Download

What would you challenge?

Which Adobe adaptation most fundamentally changed the company’s business model?

I am particularly interested in interpretations that challenge the chronology, the business-model elements I assigned or the tension between increasing customer value and increasing customer dependence.

Use this case with your team

I am developing Continuous Business Model Innovation as a format for executive briefings, strategy off-sites, workshops, bootcamps and teaching.

The purpose is not to copy Adobe or introduce a subscription because it worked for someone else. It is to help a team determine which part of its own business model has become a constraint, which assumptions require evidence and what should be tested before committing further resources.

If your organization is facing such a decision, my innovation strategy consulting provides independent challenge and structured decision support before further resources are committed.

This analysis reflects my personal interpretation of Adobe’s business-model adaptations based on available public data and public sources. Adobe and related marks are trademarks of Adobe Inc. This independent analysis is not affiliated with or endorsed by Adobe.

Subscribe to INNOVATION&

Evidence-based analysis, practical tools and innovation strategy consulting for leaders deciding what to test, fund, pivot, stop or scale.

Subscribe now

A venture can have a team, a budget, a roadmap, a name, a pitch deck, and even a recognizable culture before it has a business.

That is not necessarily a problem. Ventures need enough structure to begin operating, and founders need enough confidence to act before the outcome is known. Without that confidence, nobody takes the first risk, makes the first call, interviews the first potential customer, builds the first prototype, or asks another person to join something that may never work.

Confidence is necessary because every venture begins before certainty exists.

Overconfidence starts somewhere else. It begins when the venture starts behaving as though the company already exists while the market has not yet confirmed that it should. The team gradually treats its roadmap, funding, hiring, internal support, or early enthusiasm as evidence that the business itself is working.

The distinction is not between founders who believe and founders who doubt. It is between belief that remains exposed to evidence and belief that begins protecting itself from evidence.

A calibrated team can believe strongly in its abilities and still distinguish among what it hopes, what it assumes, what it has observed, and what it has learned. It can explain which parts of the business appear credible, which remain uncertain, and what evidence would cause it to change direction.

An overconfident team starts treating these categories as though they were interchangeable. Its assumptions acquire the language of facts, plans become promises, and internal activity begins to serve as evidence of external progress.

This matters because a startup is not simply a smaller version of an established company. It is a temporary structure operating under uncertainty and trying to discover whether a repeatable, commercially viable business deserves to exist.

The same is true of a corporate venture. A project inside a large organization does not become a business because it has a sponsor, a budget, a steering committee, or a place in an innovation portfolio. It may already look organized, but organization is not validation.

A venture is not yet a company seeking success. It is a project trying to earn the right to become a company through contact with the market.

This connects to the broader question of how a startup can earn its right to exist through evidence. User-level desirability is necessary, but it does not by itself establish that the venture is viable, scalable, or worth continued investment.

Confidence needs calibration, not suppression

It would be too simple to conclude that founders are generally overconfident and should become more cautious.

Entrepreneurial action requires people to make decisions before all relevant information is available. Founders need enough confidence to pursue opportunities that other people do not yet see, tolerate repeated rejection, attract employees and investors, and maintain direction long enough to learn something meaningful.

The research nevertheless gives us good reasons to examine entrepreneurial overconfidence seriously.

Daniel Forbes investigated why some entrepreneurs appear more overconfident than others and found that overconfidence varied with individual and organizational conditions, including age, the comprehensiveness of decision-making, and the presence of external equity funding [1]. His findings suggest that overconfidence should not be understood only as a fixed personality trait. It can also be influenced by the environment in which entrepreneurial decisions are made.

Robert Singh’s review of the entrepreneurship literature argues that entrepreneurs may overestimate both their own abilities and the prospects of their ventures, leading them to pursue opportunities that deserve more skepticism [2]. Singh’s article is primarily a conceptual review rather than direct causal proof that overconfidence produces failure in every case. Its value lies in showing how consistently overconfidence appears as a concern across entrepreneurship research.

Gudmundsson and Lechner explored how cognitive biases interact with organizational characteristics and entrepreneurial firm survival [3]. Their study suggests that the relationship is more complicated than a simple claim that confidence is always harmful. Optimism and confidence can support action, while excessive confidence can weaken the quality of decisions and reduce a venture’s willingness to respond to unfavorable information.

The problem is therefore not confidence itself. It is confidence that has lost its calibration.

A calibrated team can say that it believes a particular customer has an important problem while acknowledging that it has not yet established how frequently the problem occurs, who controls the budget, whether the customer would switch, or whether the value of the solution could exceed the cost and risk of adopting it.

An overconfident team often speaks differently. It says that it already knows the customer, that the market only needs education, that users do not understand the solution yet, or that sales will arrive after the next feature is completed. A positive pilot becomes proof that the business has been validated, while contradictory feedback is rejected because it supposedly came from the wrong segment.

Any of these explanations might be correct in a particular situation. The problem is not the statement itself, but the job it performs.

In a learning system, the explanation remains a hypothesis that can be examined. In an overconfident system, it becomes protection against evidence that might threaten the existing strategy.

That is the moment useful entrepreneurial confidence starts becoming a liability.

Overconfidence becomes more dangerous when it becomes social

Overconfidence is often discussed as though it exists primarily inside an individual founder’s mind. In practice, the more consequential version may be social.

Venture teams are fast-forming social systems. People develop shared language, assumptions, rituals, hopes, and explanations for why the project matters. This cohesion is useful because uncertainty is difficult to tolerate alone. Teams need trust and a sense of common purpose if they are going to continue working through ambiguity, setbacks, and disagreement.

The same cohesion can cause the idea to become part of the group’s identity before the market has confirmed the business.

Baumeister and Leary’s review of the need to belong argues that forming and maintaining meaningful social bonds is a fundamental human motivation [6]. Tajfel and Turner’s social identity theory provides a related explanation for how people derive part of their identity from group membership and begin distinguishing between those who belong to the group and those who do not [7].

Neither source studies venture teams specifically, so the application to startups and corporate ventures is an interpretation rather than a direct empirical finding. The theories nevertheless provide a useful lens for understanding why a venture can become socially important beyond its commercial prospects.

For founders, the venture may carry hopes of independence, wealth, recognition, or legacy. For corporate venture teams, it may carry sponsor expectations, professional reputation, career opportunity, and evidence that the innovation program is producing results. For employees, it may provide a role, an identity, and a group to which they have chosen to belong.

The venture gradually becomes more than a commercial experiment. It becomes a shared belief system, and shared belief systems have reasons to protect themselves.

Once this happens, evidence from the market no longer arrives as neutral information. Customer indifference may feel unfair. Pricing resistance may be interpreted as ignorance. Internal skepticism may be labeled negativity. A colleague who brings contradictory evidence from the field may be treated as someone who no longer believes strongly enough in the team.

This resembles some of the dynamics Irving Janis described as groupthink, in which cohesive groups protect consensus, rationalize warning signs, and make disagreement socially costly [8].

The concept should be applied carefully. DiPierro and colleagues’ later scoping review found that research on groupthink in professional teams remains conceptually inconsistent and that much of the literature is still based on commentary and theory rather than strong empirical measurement [9]. Their review concerned health care teams rather than venture teams, so it cannot establish that startup teams experience groupthink in the same way.

The narrower warning is still relevant. A group can become deeply cohesive around an idea while remaining far from evidence that the problem, market, and business model deserve that level of confidence.

The venture may still be searching for problem-solution fit while its social system has already moved on to execution.

A project can begin behaving like a company too early

Many ventures cross this line without noticing it.

They hire employees, assign formal roles, develop a brand, establish rituals, write roadmaps, manage stakeholders, and start discussing scale while the core business assumptions remain unresolved. These activities make the venture look serious. They create momentum, improve morale, and provide investors or executives with visible signs of progress.

They also increase the cost of discovering that the original idea was wrong.

Before a venture begins behaving like a company, it should have credible evidence about more basic questions. Who experiences the problem, and how important is it? What are people doing today instead? Why are existing solutions or workarounds not good enough? Who controls the budget, who influences the decision, and who can prevent adoption? What would motivate people to switch, and what evidence indicates that the apparent interest extends beyond polite encouragement?

When most of these questions are still answered by the team’s belief, the venture remains largely an assumption rather than a business.

That does not mean that the team should avoid all structure, refuse to hire anyone, or stop developing the idea. It means that structure should support learning rather than create the impression that learning has already occurred.

A roadmap can organize a sequence of experiments, or it can convert untested assumptions into delivery promises. Hiring can expand the venture’s capacity to investigate the market, or it can create a cost base that now depends on the idea continuing to look viable. A brand can help the team test whether a proposition resonates, or it can make abandoning the original concept feel like destroying something the team has already built.

The same activity can support learning or premature commitment. The difference lies in whether the venture treats the activity as a way of reducing uncertainty or as evidence that uncertainty has already been reduced.

Early market exposure is founder work

I believe founders should investigate the initial user, problem, and commercial assumptions themselves before hiring a large team or committing substantial resources.

This is not because founders should continue doing everything forever. Research, product development, sales, operations, and delivery will eventually require dedicated people with deeper capabilities. The reason is that direct market exposure shapes the founder’s judgment at the stage when the business model remains most uncertain.

Founders need to hear how potential users describe the problem in their own language. They need to understand what people have already tried, where money currently flows, what creates urgency, which constraints prevent switching, and whether the supposed struggle is important enough to change behavior.

An initial set of serious conversations with users, buyers, partners, or other relevant stakeholders will not validate an entire business. There is no universal number of interviews that proves that a market exists. The purpose of the first research cycle is more modest and more useful: it should reveal whether the original customer and problem assumptions deserve the next investment.

This work is difficult to delegate because the founder is not merely collecting information. The founder is developing the judgment required to interpret what the market is saying.

Hiring too early can turn uncertainty into obligation. Salaries begin, roles are created, and roadmaps become promises to people who need work to perform. Investors, sponsors, and managers expect visible progress. The venture is no longer only examining an idea; it is also maintaining an organization that now benefits from the idea continuing to exist.

When founders are unwilling to test their core assumptions personally before asking other people to build around them, they are not simply delegating execution. They are outsourcing their own overconfidence.

Corporate ventures borrow the appearance of certainty

Corporate ventures make the same mistake through a different set of incentives.

The people involved may not have founder identity in the same form, but they operate inside a system with its own pressures. A sponsor wants visible progress, a business unit wants strategic relevance, an innovation function needs successful portfolio stories, and a steering committee expects the team to communicate confidence. Budgets and careers may depend on the venture not appearing weak too early.

The project therefore starts collecting internal signals that look like validation. It receives executive approval, funding, strategic-fit scores, a formal team, a roadmap, workshop enthusiasm, pilot interest, and a polished business case.

These signals may be useful, but they are not market evidence.

Corporate sponsorship is not customer demand. Internal alignment does not demonstrate willingness to pay. Funding an initiative does not transform it into a business.

The corporate setting can make overconfidence harder to recognize because the venture borrows credibility from the parent organization. It has titles, meeting rooms, templates, governance processes, logos, and communications support. These elements make the project appear more substantial than an independent startup with the same level of external evidence.

Uncertainty does not disappear because the idea belongs to a respected organization.

Edison and colleagues examined lean internal startups in two large software companies and identified organizational conditions that could enable or inhibit their work [5]. Senior management support and cross-functional collaboration could help internal ventures, while the way the venture was initiated, governed, and connected to the parent organization could create additional constraints.

The study was based on two cases and seven interviews, so it should not be treated as a universal model for all corporate ventures. It nevertheless supports an important point: internal ventures do not escape uncertainty. They encounter it inside a more complex organizational and political environment.

A corporate venture may have more resources than an independent startup, but those resources do not tell it whether the business deserves to exist.

The market still knows more than the room.

The business model is not hidden inside the pitch deck

No founder, consultant, accelerator, investor, executive sponsor, or innovation team knows the complete business model in advance.

The business model is not contained in the pitch deck, the workshop, the financial model, or the roadmap. It must be discovered through contact with users, buyers, non-buyers, budget holders, procurement processes, partners, switching costs, pricing reactions, actual usage, and the operational realities of delivering the proposed value.

This is the practical contribution of hypothesis-driven entrepreneurship.

Eisenmann, Ries, and Dillard describe an approach in which an entrepreneurial vision is translated into falsifiable business-model hypotheses that can be tested rather than treated as a plan that only needs to be executed [4]. Their Harvard Business School material is a teaching note rather than an empirical trial, so it should be understood as a structured method rather than proof that Lean Startup practices always improve venture outcomes.

Its underlying logic remains valuable. The venture begins as a set of assumptions, and early action should determine which assumptions deserve further investment.

That sounds obvious, but it is difficult in practice because evidence is not only analytical. It is also social.

When evidence supports what the team already believes, it creates energy and strengthens cohesion. When evidence contradicts the belief, the team must decide whether it is willing to learn or whether it will defend the story it has already built.

That is where overconfidence becomes visible.

What I look for in startup teams

When I work with startup or venture teams, I usually begin with questions that appear simple.

What was the original idea, and which hypotheses followed from it? Which assumptions have been examined, and which remain beliefs? What evidence came from people outside the team? What did potential customers actually say, do, reject, ignore, or question? What changed in the team’s thinking because of that evidence?

I then examine the social system surrounding the venture. Which assumptions can be challenged openly, and which appear protected? Who is allowed to bring bad news? What happens when the findings do not support the current narrative? Is disagreement treated as useful information, or does it become evidence that someone lacks ambition or loyalty?

When confidence is high and market evidence remains weak, the appropriate response is usually to return to the field. The purpose is not to prove the team wrong. It is to determine whether the proposed business has commercial ground outside the room.

This is where the difference between learning and defending becomes visible.

A learning team may question the quality of contradictory evidence, but it also examines what the evidence could mean. It asks whether the original segment was wrong, whether the problem is less important than expected, whether another stakeholder controls the decision, or whether the cost of switching makes the solution unattractive.

A defending team uses every limitation in the evidence to protect the original direction. Weak signals are reframed as messaging problems, timing problems, feature gaps, customer-education problems, or mistakes in participant selection.

Any one of these explanations might be correct. When every contradictory finding produces an explanation that preserves the existing strategy, however, the team is no longer examining the venture. It is protecting its shared belief.

The strongest venture teams are not those that never doubt themselves. They are confident learners who can maintain ambition while allowing evidence to change their direction. They do not collapse when the market contradicts them, and they do not obey every customer comment without judgment. They evaluate the quality of the signal, update their assumptions, and become more precise about what remains unknown.

Early gates should be learning gates

This is why early investment gates matter.

A gate should not be a ceremony through which a team receives permission to continue because it has completed another activity. It should protect the organization against increasing commitment faster than it reduces uncertainty.

In the earliest stage, the gate should concentrate on the user and problem assumptions. Before the venture builds too much, hires too much, or spends too much, it should know whether the struggle is sufficiently real, important, specific, and underserved to justify further investigation.

Later gates should address other assumptions, including willingness to switch, willingness and ability to pay, the buying process, access to customers, delivery requirements, acquisition economics, retention, and the conditions required for scale.

Each step should earn the next one.

This is not bureaucracy. It is a way of preventing ambition from turning into premature commitment.

The question at an early gate should not be whether the team has produced enough visible work. It should be whether the venture has learned enough from outside the team to justify the next increase in cost, organization, and confidence.

Premature company-building makes honesty expensive

The deeper a venture moves into development, the harder honest reassessment can become.

The runway becomes shorter, the team has more to lose, and the founder’s identity becomes more closely connected to the venture. In a corporate setting, the sponsor has accumulated more reputational exposure, the story has been repeated to more stakeholders, and the organization increasingly expects progress.

At this point, the venture often asks for the wrong form of support. It seeks go-to-market advice while the underlying customer problem remains uncertain. It requests solution validation when it still needs to understand the user’s struggle. It asks how to scale sales before establishing why enough customers would switch.

Once traction remains weak, the next problem is how founders and stakeholders interpret weak traction. The same signal can support several role-protecting narratives, making continuation appear reasonable long after learning has begun to stall.

The claim that late-stage persistence is sometimes overconfidence is a practical interpretation, not a proposition directly tested by the nine sources used here. The research on entrepreneurial overconfidence, social identity, group cohesion, and internal venture constraints makes the interpretation plausible, but it does not prove that every persistent venture is overconfident.

Sometimes persistence is exactly what the situation requires. Difficult markets, long procurement cycles, unfamiliar technologies, and changing customer behavior can all produce weak early signals even when an opportunity is real.

The test is whether the team remains willing to specify what it believes, confront contradictory evidence, and explain what would change its mind.

Persistence that remains exposed to evidence is confidence.

Persistence that explains away every possible contradiction is something else.

The antidote is disciplined humility

The opposite of overconfidence is not insecurity, cynicism, low ambition, or endless analysis.

It is disciplined humility.

A venture can have a clear North Star while remaining willing to change its route. It can believe strongly in a customer struggle while abandoning its first solution. It can remain ambitious while acknowledging that the market has not yet confirmed the business model. It can move quickly without scaling commitment faster than evidence.

A calibrated venture team does not claim that it already knows. It explains what it currently believes, which evidence supports that belief, what remains uncertain, and what would cause the team to reconsider.

That is not weakness. It is the operating discipline required to keep a venture exposed to reality.

Founders and corporate venture teams do not need less confidence. They need confidence that points them toward the market rather than away from it.

Every venture starts with assumptions. The danger begins when those assumptions acquire the social status of facts, become embedded in roles and roadmaps, and eventually form part of the venture’s culture.

A startup is not yet a company. A corporate venture is not yet a business. Both are projects operating under uncertainty and attempting to earn the right to become something more.

That right is not earned through funding, hiring, internal alignment, a roadmap, or the elegance of the story. It is earned when market evidence becomes strong enough to justify the next commitment.

The market does not require the team to be right from the beginning. It requires the team to recognize when it is wrong before remaining wrong becomes unaffordable.

Build the company after the evidence begins to justify it, not before.

If your venture is facing this threshold, my innovation strategy consulting helps leadership teams examine the evidence, expose the weakest assumptions and decide whether to continue, pivot, stop or scale.

Paid application: The Venture-to-Company Gate

Read more

A recent exchange about user research stayed with me.

The original conversation was about a tool designed to reduce the operational friction of research. Recruiting appropriate participants takes time, arranging interviews can become a calendar exercise, conducting the conversations requires preparation, and making sense of what people said is rarely as simple as summarizing a transcript. Anyone who has tried to organize customer research inside a busy organization will recognize that friction.

The tool was intended to make all of this easier, and that is useful. Yet the part of the conversation that stayed with me was not the tool. It was the timing.

Teams do not postpone customer conversations simply because research is difficult to organize. Sometimes that is the genuine reason. At other times, operational friction provides a convenient explanation for a deeper problem: the project has already progressed too far, and worse, resources have already been invested in developing a solution for what I would call merely a “possible innovation candidate.”

By the time research is proposed, there is often more than an idea. A concept has been developed, a prototype exists, and a roadmap is beginning to take shape. Several internal supporters have attached their names to the initiative. Someone has written a business case, secured initial funding, or presented the opportunity to senior management. The team has invested time, the sponsor has invested credibility, and the organization has gradually started treating the project as something that exists rather than something that might deserve to exist.

Then someone asks for user research.

The activities may still look the same. Participants are recruited, interviews are conducted, observations are analyzed, and findings are presented. However, the job that research is being asked to perform has changed.

Early research asks what the organisation needs to learn before it commits. Late research is more likely to ask whether enough evidence can be found to justify what the organization has already started.

That is the moment research risks moving away from discovery and towards approval-seeking. It becomes solutionism with a research layer added afterwards.

The same interview can serve a different purpose

The difference between early and late research is not primarily methodological. The same interview guide, recruitment process, and analysis method could be used at either stage. What changes is the organizational environment in which the evidence will be interpreted.

Early in an initiative, negative evidence threatens an idea. Later, the same evidence may threaten a prototype, a roadmap, a team’s objectives, a sponsor’s judgement, and a narrative that has already travelled through the organization.

This matters because people do not assess new information independently of what they have already invested. Barry Staw’s foundational work on escalation of commitment showed that decision-makers can continue committing resources to an unsuccessful course of action, particularly when they feel personally responsible for the original decision [1]. Arkes and Blumer later described the sunk-cost effect as the tendency to continue an endeavor because money, time, or effort has already been invested, even though those past investments should not determine whether further investment is justified [2].

Late research enters precisely this environment. The team is no longer learning only about customers, markets, and business models. It is also negotiating with everything that has already been spent and promised.

This does not mean that the team deliberately manipulates the research. Most people involved will be sincere. They want to learn, they care about the project, and they may genuinely believe that further development will reveal its potential. Yet sincerity does not remove bias. A team can ask an honest question while unconsciously preferring the answer that allows it to continue.

The expense of late research therefore does not lie mainly in recruiting participants or conducting interviews. It lies in the amount of accumulated commitment that the findings must overcome.

Innovation work is learning before commitment

I have encountered this pattern several times in my professional work.

I have been invited into innovation and venture projects that appeared advanced from the outside. The teams had prototypes, presentations, roadmaps, market estimates, and a coherent story. Considerable work had clearly taken place, and the quality of the artifacts often created the impression that the initiative itself was equally mature.

Then we began examining what the project was built on.

Which assumption was the prototype intended to test? What were potential customers doing instead? What would make them abandon an established workaround? Who experienced the struggle, who controlled the budget, and who would carry the operational and political cost of switching? Which observations supported the current direction, and what finding would cause the team to reconsider it?

That is usually when the character of the conversation changes.

The problem is rarely that the team is weak or careless. Most teams are responding rationally to the system around them. They have been rewarded for building something, aligning stakeholders, making progress visible, refining the narrative, and defending the next stage of funding.

All of those activities can be useful. None of them proves that the initiative deserves to exist or continue.

They demonstrate that the team has been active. They do not demonstrate that consequential uncertainty of feasibility has been reduced.

This distinction separates innovation work from delivery work. When the customer problem is sufficiently understood, the relevant market mechanisms are known, the business model has been established, and the primary challenge is reliable execution, the organization is doing delivery. The work may still be difficult and uncertain in many practical ways, but the underlying direction is no longer the main question.

Innovation begins where material uncertainty remains.

A good innovator is therefore not someone who already knows the answer. A good innovator knows how to learn before the organization makes being wrong unnecessarily expensive.

That does not mean arriving without expertise. People responsible for innovation should understand customer research, business model logic, assumptions, experiments, evidence, and decision gates. They should know how to distinguish an interesting signal from a convenient anecdote. What they should not do is assume that their experience allows them to know the specific opportunity before reality has had a chance to challenge it.

This broader view matters because business model validation must test the whole system, not only the customer problem or the first solution. A promising signal in one component does not establish that the connections among demand, switching, delivery, revenue, and economic viability will hold.

Research by Camuffo, Cordova, Gambardella, and Spina supports this view. In a randomized controlled trial involving 116 Italian startups, entrepreneurs who were taught to formulate explicit hypotheses and test them systematically performed better and were more likely to pivot than those following more conventional approaches to market feedback [3]. A later replication across 759 firms and four randomized controlled trials found that the scientific approach increased idea termination and was associated with a more selective pattern of strategic change, rather than either refusing to change or pivoting repeatedly [4].

These studies should not be interpreted as proof that every experiment produces a better venture or that uncertainty can be eliminated. Their more relevant contribution is that explicit hypotheses and disciplined tests can help entrepreneurs recognize when an idea should be changed or abandoned.

The central innovation skill is not being right immediately. It is learning fast and early enough that being wrong remains affordable.

What an advanced project may still be hiding

Organizations often mistake the visible development of a project for the reduction of uncertainty behind it.

A polished concept can feel like evidence because it makes the opportunity easier to imagine. Internal alignment can feel like market validation because several influential people now support the idea. A prototype can create the impression that the team has moved closer to a viable business, even when the prototype has not tested the assumptions most likely to make that business fail.

Internal alignment can keep a weak project alive for a long time. It can protect the team, unlock further budget, and make the initiative appear increasingly official. What it cannot do is create demand outside the organization.

By the time research begins, the project may no longer be treated as a hypothesis. It has become someone’s work, someone’s quarterly objective, someone’s internal promise, and sometimes someone’s route to recognition.

That changes the politics of learning.

Early research asks whether the organization should commit. Late research asks whether the organization can still afford to be honest.

This is why positive feedback collected after a prototype already exists must be interpreted carefully. Such feedback is not worthless. It can reveal confusing language, missing functionality, unexpected objections, relevant use cases, and aspects of the customer context that the team has overlooked.

It is nevertheless easy to ask positive feedback to carry more weight than it deserves.

This is why user research can produce socially convenient answers rather than decision-grade evidence. The problem is not necessarily dishonesty; the interview setting itself can reward politeness, abstraction, and self-presentation.

People are often polite when they are shown a new idea. Some enjoy participating in innovation discussions and want to be helpful. Internal sponsors may naturally remember the statements that support the direction they already favour. Teams may interpret general interest as evidence of future adoption.

None of this is equivalent to a customer changing behavior, reallocating a budget, replacing an existing workaround, accepting the cost of switching, or taking a meaningful risk.

The later the research occurs, the easier it becomes to collect evidence that improves and protects the existing story instead of testing whether the story is true.

Early research does not need to prove the market

The argument against early user research is frequently framed in terms of cost and speed. Research is described as expensive, slow, or too academic for the pace at which innovation teams are expected to move.

Thorough research can certainly be demanding. Some questions require broader samples, careful recruitment, specialist expertise, quantitative analysis, or observation over an extended period. A few interviews cannot establish market size, predict adoption rates, prove willingness to pay, or demonstrate that a business will become commercially viable.

However, early research does not need to begin as a definitive study. Its first job is usually to reveal whether the team is building on assumptions that nobody has examined.

In many contexts, a focused team can conduct an initial set of exploratory conversations within a week. Participants may come from existing customers, prospects, professional networks, industry communities, or groups already using an alternative solution. Research platforms may reduce the administrative burden, but the value does not come from speed alone. It comes from asking a clear question of the right people before the answer becomes politically inconvenient.

Eight or ten interviews do not prove that a market exists. They can still reveal that the original problem is poorly framed, that customers experience it differently from the way the team imagined, or that the proposed solution conflicts with how purchasing and switching decisions actually occur.

Research on qualitative saturation is relevant here, although it is often applied too broadly. Guest, Bunce, and Johnson analyzed sixty interviews from a particular study involving a relatively homogeneous population. They found that basic elements of the main themes appeared within the first six interviews and that saturation occurred within the first twelve in that dataset [5]. Their findings do not establish a universal minimum or maximum number of interviews. Sample requirements depend on the question, the diversity of participants, the purpose of the research, and the quality of the analysis. The narrower conclusion is that focused qualitative research can reveal meaningful patterns relatively quickly under appropriate conditions.

That is the role of an early research cycle. It does not purchase certainty. It purchases a better decision about what should happen next.

Assume that two people spend five days preparing, conducting, and synthesizing an initial research cycle, at an internal or equivalent external cost of $1,000 per person per day. The cost would be approximately $10,000.

That amount does not buy proof, but it may buy an important decision point. It can show whether the problem is sufficiently important to justify further investigation, reveal what customers do today instead, and expose whether the struggle creates meaningful cost, delay, risk, frustration, or political difficulty. It can also reveal whether switching would be easy, expensive, unrealistic, or dependent on people who have not yet been involved.

Most importantly, it requires the team to make its assumptions visible.

The team must explain what it currently believes, which conditions need to be true for the initiative to deserve further investment, what would count as a meaningful signal, and what evidence would cause it to change direction or stop.

These questions appear straightforward, but they remove the places in which weak projects hide. Without explicit assumptions, teams can debate features, preferences, opinions, and optimistic projections for months. Once the assumptions are stated clearly, the conversation becomes more demanding because the team has to explain what it believes about reality before reality is asked to respond.

Research is a gate before it becomes a report

The value of early research does not lie primarily in the final report, the workshop output, or the customer quotations selected for a presentation.

Its value lies in the decision it makes possible.

The team may continue because the initial evidence supports further investigation. It may deepen the research because the problem appears material but remains insufficiently understood. It may change direction because a critical assumption has weakened, or it may stop because the opportunity no longer justifies additional investment.

This is a different way of understanding research. It is not an activity attached to a project once the project is already underway. It is a gate placed before the project becomes unnecessarily expensive.

Stage-Gate thinking is useful here, provided that it is not reduced to governance theatre. Its relevant contribution is the discipline of making an explicit decision before additional resources are committed. More recent descriptions of gating systems have also emphasized that these mechanisms need to become faster, more adaptive, and more compatible with iterative development rather than functioning as rigid sequential approval processes [6].

The gate I mean is therefore small, fast, and close to the evidence. Its purpose is not to demonstrate compliance with a process. It is to make it harder for a team or sponsor to hide behind the quality of a presentation.

The relevant cost comparison is not $10,000 for research against zero dollars for doing nothing. It is $10,000 bet now against months in which a multidisciplinary team continues moving in the wrong direction.

A team that keeps designing, building, coordinating, presenting, and governing a weak initiative consumes resources even when no separate research budget is visible. Depending on the organization, the cost of another months may include salaries, technology, prototypes, external partners, management attention, and the opportunity cost of not pursuing something stronger.

Early research appears expensive because it has a visible price. Continuing without learning often appears inexpensive because the cost is distributed across normal project activity.

That accounting illusion makes research look costly and weak projects look cheap.

The cost of not knowing compounds

A late diagnosis is rarely expensive only because the diagnosis itself costs more. It is expensive because the underlying problem has had time to grow.

The same pattern appears in product and innovation work. The later a weak assumption is examined, the more budget, identity, politics, and internal logic become attached to it. Stopping no longer feels like a useful learning outcome. It feels like failure.

Teams therefore continue improving the prototype, sharpening the narrative, and searching for favorable reactions. They find groups that respond positively and develop explanations for those that do not. They call the activity validation, although it may gradually become a negotiation with sunk cost.

A team can be sincere while protecting a weak project. A sponsor can request evidence while rewarding confirmation. An organization can celebrate learning while punishing the first person who recommends stopping.

For that reason, early user research is not only a research topic. It is a capital-allocation topic.

The relevant question is not whether the team has produced something. It is whether the next unit of capital, attention, time, and credibility deserves to be committed.

The sponsor’s real decision

Sponsors therefore have a different job from the one they are often asked to perform.

Their task is not merely to assess whether the team has built something impressive or presented a compelling opportunity. It is to decide whether the next investment will reduce consequential uncertainty or protect costs that have already been incurred.

That question changes the funding conversation.

Before approving a prototype budget, a pilot, a detailed roadmap, or another quarter of development, the sponsor should ask for the assumptions rather than the ambition. The market-size slide and polished concept can wait until the team has explained what must be true.

Which claims are supported by direct observation, and which remain interpretations? What do customers currently do instead? What would make them switch? Which stakeholder experiences the problem, which one controls the budget, and which one can prevent adoption? What finding would cause the team to stop?

An organization is allowed to place a strategic bet even when evidence is limited. Some decisions cannot be tested fully in advance, and uncertainty can never be removed entirely. The organization should nevertheless describe the decision honestly.

When critical assumptions remain unexamined, another round of development is not evidence-based progress.

It is a bet hidden inside delivery language.

Once a project has acquired a team, budget, roadmap, and story, it becomes difficult to stop for reasons that have little to do with customer demand. Escalation of commitment appears in meetings through phrases such as “one more iteration,” “we need better messaging,” “customers do not understand it yet,” or “we have already invested too much to stop.”

Any of those statements may be correct. The problem is that they are often accepted before being tested.

When everyone holds a hammer

There is also a team-design problem beneath all of this.

Many innovation teams are built predominantly around people who are good at creating solutions: engineers, designers, product managers, domain specialists, and business builders. These capabilities are valuable and eventually indispensable.

However, when a team is strong only at shaping solutions, it may remain weak at determining whether the underlying problem deserves a solution in the first place.

That is not a minor imbalance. It allows a team to move quickly in the wrong direction while looking highly productive from the outside.

Not every initiative needs a formal research department before work can begin. Yet someone involved must possess the competence and independence required to investigate the problem, question the framing, distinguish claims from evidence, and recognize when a plausible opportunity is not strong enough.

When those capabilities are missing, the organization should develop them or introduce them before its commitment grows. Otherwise, what appears to be speed may simply be the extension of a weak project’s life.

Early research is not expensive because it delays progress. It is valuable because it reveals whether what the organization calls progress deserves to continue.

Keeping uncertain projects alive without learning is expensive.

And when everyone is holding a hammer, the customer’s struggle has little chance of being understood as anything other than a nail.

If your team needs to determine whether an opportunity deserves further investment, my innovation strategy consulting provides an independent review of the customer evidence, weakest assumptions and next decision gate.

Paid application: The Early Research Gate

Read more