Field note · October 28, 2025
The question that kills the most projects
If this were 100% accurate, what would change, and what is that worth?
I ask this in every kick-off now:
If this analysis or model were perfect — 100% accurate, available instantly — what would we do differently, and what is that worth?
It is not a rhetorical device. It is a genuine scoping question, and the answers sort projects into three groups fast.
Group one: a specific, vivid answer. "We would call the top 200 at-risk accounts each week instead of guessing, and each save is worth about £4,000 annually." Build it. You now also know the shape — a ranking problem, top-200, weekly — which is different from the churn probability model someone was about to specify.
Group two: a vague answer. "We would have better visibility." This usually means the requester wants reassurance about a number rather than a decision changed. That is a legitimate need and it is almost always served better by a threshold alert than by a dashboard or a model — an hour of work instead of two months.
Group three: no answer. Nobody can say what would change. Stop. This has happened maybe a quarter of the time and every instance has been a project that would have been built, delivered, admired briefly, and then maintained forever by someone who did not want it.
The question is uncomfortable to ask because it can sound like resistance. Two things make it land: ask it curiously rather than sceptically, and ask it about everything, including your own ideas, so it reads as process rather than as an objection to that particular request.
The version I use when there is no time for the full conversation: "what will you do differently depending on what I find?" Same information, shorter, harder to hear as a challenge.