Your Innovation Pilots Are a Graveyard.

Time to Stop, Bury them with a distinct Ceremony.

Most innovation pilots in large organisations are not experiments. They are expensive theatre. A team is assembled, a budget is allocated, a vendor is selected, and a steering committee is formed. Six months later, a slide deck lands in front of a board that politely nods. Then nothing happens.

Innovation consultants refer to this as “failing fast.” There is nothing fast about it. And the failure was implicit from day one.

The goal was never to learn. The goal was to look innovative.

95% of AI pilots never reach production, according to MIT Media Lab research. The numbers are no better for the waves of Blockchain, digital twins, and IoT platforms. We have spent a decade chasing technology trends as if the technology itself would do the work. It does not. It never did.

The problem is not that organisations pilot too little.
It is that they pilot the wrong things, for the wrong reasons,
and structurally in the wrong way.

Startups Know Something Enterprises Have Forgotten

There is a useful comparison hiding in plain sight. Startups that actually scale do not run innovation pilots. They make a conscious, deliberate bet on one problem worth solving, and they go and solve it. The entire team is aligned around a single question: Does this create value for a paying customer, and can we prove it this month?

That discipline comes partly from necessity. When you have 18 months of funding runway, you do not have the luxury of a portfolio of exploratory pilots with no deadline for a go or no-go decision. But it also comes from structural clarity. In a startup, the distance between the person who has the idea and the person responsible for commercial delivery is essentially zero. Innovation and business development are the same conversation, often held by the same person.

In a mid-size or enterprise company, that distance is enormous. There is an innovation team, a product team, an IT department, a transformation office, a commercial organisation, and somewhere further along, there is the customer. A pilot can live comfortably inside that structure for a year without ever getting close to any of those last two. The people running the pilot are measured on outputs: a demo, a report, and a proof of concept. The people responsible for revenue are measured on something else entirely and rarely get involved until it is far too late.

In a startup, an innovation person is the same as the business development person. In a large company, they are barely on speaking terms.

This is the pilot trap. Not a failure of ambition or intelligence. A failure of structure. Pilots in large organisations are architecturally separated from the moment of value. They exist in a protected space that is just far enough from the business to feel safe and just far enough from the customer to be useless.

Kill More Pilots, Earlier

The counterintuitive prescription is not to run better pilots. It is to kill more of them, much sooner, and concentrate the energy you free up on the handful that are actually worth crossing the chasm for.

Geoffrey Moore’s chasm is usually discussed in the context of taking a product from early adopters to the mainstream market. But there is a chasm that comes even earlier for enterprise innovation, and it is less discussed. It is the gap between a pilot that works in a controlled environment and an innovation that is actually embedded in how the business operates and delivers value to customers. Most pilots never cross it. They die quietly, or worse, they are kept on life support because nobody wants to be the one to cancel them.

The honest thing to do is to make that go/no-go decision as early as possible. Not at the end of six months when budgets are spent, and people are emotionally invested. Why not after 2 days or after the first week? Ask the one question that matters: is there a credible path from what we have learned here to something a customer would pay for or a process that demonstrably reduces cost or risk? If the answer is not a clear yes, stop. Completely. Free the people and the budget.

Keeping a mediocre pilot alive does not just waste money.
It steals the attention that a promising one needs.

The organisations that are genuinely effective at innovation are not the ones running the most pilots. They are the ones that are ruthless about concentrating their crossing-the-chasm energy on the two or three bets that have real traction. You cannot cross the chasm with ten horses. You need one very good one, and everything pointed in the same direction.

Stop Blaming the Technology

Blockchain did not fail your supply chain transformation. AI is not failing your customer service efficiency improvement. The technology works perfectly in environments designed to use it. What failed was the assumption that deploying a technology is the same as solving a business problem.

When I ask leaders why they started an AI pilot, I hear one of two answers: because the board asked us to, or because our competitor was doing it. Neither is a business reason. Both are fear in disguise.

Pippi Longstocking said she had never tried something before, so she was sure she could do it. That spirit is worth celebrating. But Pippi had a business goal: she wanted to go to school, be a fine lady, or bake pancakes. She had a destination. She was not running a pilot to explore whether destinations existed.

Technology is not the answer to a question you have not asked yet.

Before you launch anything, answer two questions honestly: what specific customer or operational problem are we solving, and what does success look like in the next 30 days? (Not 18 months). If you cannot answer both in two sentences, you are not ready, and you should not start.

Fail Fast Actually Means: Decide Fast

The innovation industry sold a beautiful lie. Fail fast, learn faster. Iterate, pivot, disrupt. The Silicon Valley gospel found its way into the PowerPoints of every corporate innovation department, and somewhere in the translation, the learning got dropped, and only the failing remained.

Real “failing fast” looks like this. You define a hypothesis on Monday. You test it with real users by Wednesday. You have data by Friday. You make a decision the following Monday: kill it or scale it. That is one week. Not a quarter. Not a financial year.

What most organisations practise is slow-motion failure with expensive consultants and milestone reviews. The budget runs out before the learning begins. By the time a pilot is complete, the problem it was supposed to solve has either been addressed by a competitor or quietly removed from the agenda.

Fail fast should mean: make a decision in days,
not deliver a presentation in months.

The cost of a failed pilot is rarely the direct cost of the pilot itself. The real cost is the confidence it drains from teams, the cynicism it plants across the organisation, and the credibility it burns with the customers who were never asked whether this was the right problem to solve in the first place.

Be Successful Fast

There is a companion to failing fast that nobody talks about. Succeeding fast. Not succeeding in six months once the business case is approved and the change management plan is finished. Succeeding this week, with what you have now.

Every organisation I have worked with has one thing in common. Somewhere in the business, a team is doing something that genuinely works. A process that runs cleanly. A customer interaction that consistently lands well. A data flow that actually produces decisions. The question is why the rest of the organisation does not look like that team.

The answer is almost always the same. That team did not wait for permission. They found a real problem, built the smallest possible solution, measured the outcome, and improved it. They did not call it an innovation pilot. They called it their job.

Organisations that have genuinely transformed are not the ones that ran the most impressive transformation programmes. They are the ones who made relentless small bets on real problems, killed the ones that did not work quickly, and doubled down on the ones that did. The chasm was crossed not in one dramatic leap but in dozens of small, deliberate steps, each one grounded in something a customer or operator actually needed.

Transformation is not a programme. It is the sum of many small decisions to stop tolerating things that do not work.

Three Things to Do on Monday Morning

1. Cancel one pilot that has been running for more than four months without a customer-facing result. No discussion, no review meeting. Cancel it, free the people, and redirect the energy somewhere with a clearer line to value.
2. Find the most frustrated person in your organisation. The one who complains loudest about a broken process. Give them three weeks and one developer. No steering committee, no architecture review. Solve the thing, measure it, and show the result.
3. The next time a vendor pitches a blockchain solution, an AI platform, or a data mesh architecture, ask one question first: Can you show me, using our data, solving our specific problem, in the next two weeks? Watch what happens.

Technology that delivers real value does not need six months to prove it can. If it does, the problem was never the technology.