A pilot is supposed to reduce risk.

Yet in many organisations, the pilot is little more than a step on the way to a decision that has already been made. The board approves a limited trial, everyone has an idea of what it is supposed to prove, and when the results come back messy, disappointing or simply unclear, the learning gets explained away.

The rollout continues.

The risk was never really being tested. The decision had already been made.

That is the real problem with pilots. They do not fail because they are too small. They fail when an organisation is not prepared to change its mind because of what it finds.

A useful pilot needs to be able to produce an answer you do not want.

Before it starts, there should be some agreement about what would make you continue, what would make you stop, and what would leave you needing to learn more. The people running it need to be able to say that something is not working, that people are finding it difficult, or that the expected benefit simply isn't there.

And leadership needs to be prepared to act on that information.

Imagine a public sector team testing an AI tool to help handle citizen enquiries. The system performs well in testing, but once it is used in real work, staff spend almost as much time checking and correcting its responses as they save.

That is not a failed pilot.

That is the pilot doing its job.

It has found something the original assessment could not: the tool may work, but it does not yet work well enough in the environment where it is actually needed.

The organisation now has useful choices. It could narrow what the tool is used for, improve the information it works from, change the process around it, give staff more support, or decide that the benefit is not worth the effort.

What matters is that the decision changes because of what was learned.

That is where the real value of a pilot sits. Not in producing a positive result, but in replacing an assumption with something you have actually learned.

A leadership team that can say, *we tried it, this is what happened, and this is what we are changing because of it* builds trust in evidence. People learn that raising a problem can improve a decision rather than derail it.

But if everyone knows the rollout will happen regardless of what the pilot finds, they learn something else: the evidence was never really going to matter.

That matters across the organisation. Finance needs to know whether a pilot has genuinely reduced the downside. People need to know whether feedback can actually change the approach. Technology needs to know whether something that works in isolation works in the real environment.

A pilot cannot answer all of those questions if it is only designed to prove that the technology works.

The harder part is creating enough space for the answer to be uncertain.

So the next time someone proposes a pilot, ask a simple question:

**What would we need to see for us to decide not to continue?**

If nobody can give you a credible answer, it isn't really a pilot.

**It's a rehearsal for a decision you've already made.**