Give Every Pilot a Decision Date

A pilot can keep a team busy long after it has stopped helping the organization learn. Agreeing how it will end is part of designing it well.

The pilot has a sponsor, a small team and a promising idea. There is a launch meeting, an update deck and a date for the first demonstration. What it does not have is a clear account of the decision the organization expects to make afterwards.

When that decision is missing, almost any result can become a reason to continue. Encouraging feedback means the team should expand. Weak feedback means it needs another iteration. Limited use means the pilot needs more promotion. Each explanation may be reasonable, but together they can make the work remarkably difficult to stop.

Decide what this pilot is meant to establish

A useful starting point is to name the uncertainty. Can a customer complete a task with the proposed service? Will a team adopt a different workflow? Can the organization deliver a new offer within an acceptable cost and level of effort?

These are different questions. A polished demonstration might help people understand a concept while telling you very little about whether they will use it. A short trial with enthusiastic volunteers may reveal usability problems without establishing demand across a wider business.

Write down what the experiment can answer, and what will remain unknown. That gives the results a fair scope and reduces the temptation to claim either too much success or too much failure.

Bring the decision into the brief

Before starting, agree a review date, the evidence to bring and who has authority to decide. Describe the possible next steps in plain language: continue with a larger test, change a specific part of the approach, pause until a named dependency is resolved, or stop.

Consider a hypothetical pilot for a new internal advisory service. Counting meetings may tell you that the team is active. A more useful review could examine whether the service helped its intended users resolve a defined problem, how much staff time delivery required, and whether there is an owner prepared to take responsibility for the next stage.

Agree the relevant standards before the findings arrive. They should reflect the stage and purpose of the pilot; an early experiment should not be judged as though it were an established business. Equally, calling something experimental should not exempt it from a decision.

Fund the next question deliberately

Writing for Strategyzer in 2023, Tendayi Viki argued for assessing innovation projects through evidence of progress toward a workable business model, rather than relying on projections or the appeal of an idea. That is a useful practitioner perspective on the allocation question: what has the work established that makes another investment worthwhile?

For a pilot team, the next request can be small and explicit. Here is what we learned. Here is what we still do not know. Here is the next test, the resource it needs and the decision it will support. That gives a sponsor something concrete to evaluate.

If the result is inconclusive, say why. An underpowered test, a changed assumption and a consistently weak response are not the same situation. Another iteration should have a reason that can be examined, rather than becoming the default outcome of every review.

Keep the learning when the work stops

Closing a pilot should leave a usable record: the question tested, the evidence collected, the limits of that evidence and the decision taken. Include any conditions that would make the idea worth revisiting. A future team should be able to understand the outcome without reconstructing it from a folder of presentations.

That record also helps separate the decision about an idea from a judgment about the people who tested it. A team that uncovers a serious obstacle early has produced information the organization needed.

A decision date creates a useful obligation for everyone involved: at this point, we will explain what we have learned and decide what deserves to happen next.