An anonymised account of a real engagement: a large transformation programme exploring a new digital capability through an initially limited proof-of-concept.
The real problem
The pilot was successful enough to continue but never successful enough to justify scale. Yet nobody wanted to stop it. Each stage brought additional stakeholders, additional funding and additional expectations. The initiative slowly evolved from experiment to pilot to programme without a deliberate decision at any transition point.
What we did
We introduced explicit evidence gates. Each transition required answering three questions: what did we learn? What decision does that learning justify? What evidence would stop us continuing? The discussion shifted from enthusiasm to evidence.
The outcome
Several planned investments were delayed or stopped until additional evidence existed. The organisation avoided scaling uncertainty simply because momentum had accumulated.
What we’d do differently
We would define kill criteria before the pilot began. Once an initiative develops political support, stopping becomes significantly harder.
What this proves
The purpose of a pilot is not to prove something works. It is to discover whether it deserves to continue.