The Ruins of Innovation
I have been part of innovation projects in startups and in corporates.
Some survived.
Most did not.
They were stopped when funding dried up, when priorities shifted, when leadership changed, when the “bubble” inside the company deflated. Budgets were reallocated. Teams dissolved. People left. Sometimes assets were sold. Often they were quietly buried.
In hindsight, many of these initiatives look like mini bubbles inside otherwise stable firms. A pocket of venture logic inside a performance machine. Capital allocated as bets. Stories told about optionality and growth. Forecasts stretching beyond the evidence available.
That part is not surprising.
Most bets do not pay off as promised. That is the math of uncertainty. If all of them worked, they would not be bets.
What is surprising is something else.
It is not that projects fail.
It is that almost nothing survives them.
The Hidden Output of Failed Bets
Every serious innovation effort produces more than a product.
It produces:
– New technical architectures
– Prototypes and partial systems
– Supplier relationships
– Customer conversations
– Pricing experiments
– Regulatory clarifications
– Internal political maps
– Hiring profiles
– Learning about what not to build
– Patterns of demand that almost worked
Call them artifacts. Call them residue. Call them ruins.
They are expensive.
And yet, once the project is stopped, these artifacts are rarely integrated into the core business. They remain attached to the project that died. When the project dies, they die with it.
The narrative becomes binary:
“It worked.”
Or
“It failed.”
But that framing ignores the intermediate layer — the accumulated knowledge embedded in assets, code, decisions, and scars.
Mini Bubbles Inside Corporates
When a corporate launches an innovation initiative, it often creates a temporary parallel universe.
Different metrics.
Different tolerance for loss.
Different decision logic.
For a period of time, the initiative lives under venture-style assumptions:
– Invest now, justify later.
– Optionality is value.
– Exploration before efficiency.
Inside that bubble, teams move differently. They hire differently. They think differently.
But when the external pressure returns — earnings season, cost cutting, new CEO, macro shock — the bubble bursts. The venture logic collapses back into performance logic.
And here the system shows its weakness.
The organization knows how to shut down a P&L line.
It does not know how to harvest a failed experiment.
So it defaults to the simplest move: stop, cut, forget.
The Brain Drain Effect
What I have seen repeatedly is this:
Projects are stopped.
People are let go or reassigned.
Technical systems are abandoned.
Entrepreneurial talent leaves.
The loss is not only financial.
It is cognitive.
You lose:
– The people who navigated uncertainty.
– The ones who learned where assumptions broke.
– The informal networks they built.
– The pattern recognition they developed.
Some firms try to sell the assets externally. Rarely with meaningful success. Most of the value is contextual. It only makes sense inside the organization that generated it.
But internally, the memory fades quickly.
The official takeaway becomes:
“We tried. It didn’t work. Let’s not burn our fingers again.”
The result is defensive learning (zero-sum learning). Not compounding learning (positive-sum learning).
Why Artifacts Do Not Survive
There are structural reasons for this pattern.
Ownership disappears.
When a project dies, no one “owns” its residue. Without ownership, artifacts decay.Performance systems reject ambiguity.
Core businesses optimize for predictability. Half-built systems and ambiguous insights do not fit clean KPIs.Careers are tied to success narratives.
Few leaders want to champion the remains of a failed initiative. It signals proximity to loss.Accounting logic treats everything as sunk.
Once written off, the mental model becomes: value = zero.
But that equation is wrong.
Financially sunk does not mean strategically worthless.
The Ruins as Strategic Capital
In archaeology, ruins are not garbage. They are data.
They tell you how a society thought.
What it valued.
Where it miscalculated.
What it mastered.
Innovation ruins are similar.
A failed AI initiative may have:
– Built data pipelines that the core business still lacks.
– Mapped customer pain points more precisely than marketing ever did.
– Identified regulatory bottlenecks before competitors.
– Clarified cost structures under real conditions.
These are not trivial side effects.
They are hard-earned insights.
The tragedy is not that bets fail.
The tragedy is that their learning does not compound.
The Venture Illusion
Part of the problem lies in how corporates borrow venture logic without adopting venture discipline.
In venture capital, a failed startup is not pure waste if:
– The team goes on to build something better.
– The investors apply pattern recognition to future bets.
– The ecosystem absorbs talent and knowledge.
Failure is distributed. Learning circulates.
Inside a corporation, failure is centralized. It gets absorbed into the cost line. The people often exit. The system resets.
There is no internal “portfolio memory” that improves the next allocation decision.
So every few years, the same enthusiasm returns. A new theme. A new lab. A new fund. The cycle repeats.
Without institutional memory, every wave feels like the first.
What Would Compounding Look Like?
If we take uncertainty seriously, then failed bets are expected.
The question is not:
“How do we avoid failure?”
It is:
“How do we design for learning survival?”
Some uncomfortable implications follow.
First: Every innovation project should produce a structured artifact archive.
Not a slide deck.
Not a celebration video.
But a decision log:
– Initial hypotheses
– Evidence gathered
– Assumptions falsified
– Unit economics tested
– Customer segments explored
– What surprised us
– What we would test next if capital returned
If this does not exist, the next team starts from zero.
Second: Talent from stopped initiatives should be redeployed deliberately, not randomly.
Those who have navigated uncertainty are rare assets. Reassigning them into purely operational roles wastes that muscle.
Third: Technical systems should be modularized early.
If architectures are built as experiments that can be reused in adjacent domains, their survival probability increases.
Otherwise they are too entangled with the failed narrative.
The Emotional Layer
There is also a psychological dimension.
Stopping a project hurts.
For founders, it can threaten existence.
For corporate innovators, it can threaten a career.
When personal identity is tied to a project, its termination feels like personal rejection.
So once it is stopped, there is a tendency to distance from it. To close the chapter quickly. To move on.
Revisiting the ruins requires emotional maturity.
It requires saying:
“We were wrong about X. But right about Y.”
“This architecture did not scale for that market. But it solves another internal bottleneck.”
“This pricing failed externally. But it revealed our cost illusions.”
That is a harder story to tell than “success” or “failure.”
But it is a more accurate one.
The Cost of Forgetting
If artifacts do not survive, the company pays twice.
First, in capital lost.
Second, in future ignorance.
The next initiative will:
– Relearn similar lessons.
– Repeat similar structural mistakes.
– Overestimate novelty.
– Underestimate constraints already discovered.
This is not creative destruction.
It is repetitive destruction.
And over time, it breeds cynicism.
Employees conclude that innovation is theatre.
Leadership concludes that innovation does not work.
Both are partially right — in a system that does not preserve learning.
A Different Definition of Return
Perhaps the return on innovation capital should not be measured only in revenue generated.
It should also be measured in:
– Reduction of future uncertainty
– Quality of subsequent allocation decisions
– Reusable capabilities created
– Optionality clarified, even if not exercised
A €5 million initiative that prevents a €50 million strategic mistake has created value. But only if that prevention is traceable and acknowledged.
If the insight disappears with the team, the money is just burned.
The Open Question
Innovation bubbles inside companies will continue.
Themes will rise and fall.
Technologies will attract capital and then normalize.
Boards will ask for growth.
Executives will fund bets.
That is not the problem.
The deeper question is this:
When the next internal bubble implodes in your organization, what will remain?
Will it be a cautionary tale?
Or a set of assets, insights, and people that raise the starting position for the next move?
Most firms know how to start innovation projects.
Very few know how to end them in a way that compounds.
The ruins are already there.
The decision is whether to treat them as debris or as foundation.
