One question, a half day, and a budget that moved
A Mittelstand manufacturer was three weeks from signing a 1.8M euro platform migration. One question killed it in a half day, and the budget went somewhere that actually paid.
I have sat on both sides of these calls: as the engineer who once spent fourteen months polishing a “strategic modernisation” the business never asked for, and as the person who now will not approve a rebuild until it survives five plain questions. None of the five are original. Question three is Bezos’s one-way versus two-way door. Question five is lean-startup hypothesis testing. Question one is the burning-platform test every change consultant has sold since the nineties. The value is not novelty. It is forcing the conversation before the vendor deck decides for you.
The five
-
What breaks if we do nothing? Name a cost, a risk, a deal you lose. If you cannot, you have an itch, not a decision.
-
Business constraint or engineering annoyance? “The code is ugly” is not a reason. “We cannot onboard a new plant without three weeks of manual config” is.
-
One-way door or reversible? Most teams agonise over reversible calls and rush the ones they can never undo.
-
Who owns this in eighteen months? If you cannot name the person and team today, you are building debt with a launch party.
-
What is the smallest version that proves the bet? Not the MVP. One workflow. One plant. If the small version does not pay, the big one never will.
Back to that migration
Question one had no answer the business could name. The system was ugly, not failing. So we left it in place and pointed the budget at the real bottleneck: onboarding a new plant that took three weeks of manual config. Eighteen months on, that is down to under a week, and they have brought two more plants online without new hires. The migration would have touched none of it. Figures disguised at the client’s request; the shape is exact.
That is the difference the questions make. They do not tell you not to rebuild. They tell you whether the rebuild is aimed at the thing that actually hurts, or at the thing that is merely unpleasant to work in. The same discipline sits underneath the build-versus-buy-versus-modernise decision: most of the cost of getting it wrong is committed before a line of code is written, in the framing of the problem.
One caveat the framework crowd skips
Five questions do not beat office politics on their own. That migration got killed cleanly because the sponsor was willing to be asked in public and someone in the room had the standing to ask. The questions surface the truth. Someone still has to act on it. If the honest answer to question one is inconvenient for whoever proposed the programme, the framework will not make that answer welcome. It only makes it visible.
And yes, this is also how we sell: the short assessment is a real service. The questions work whether you call us or not, which is the only reason I trust them. If you are three weeks from signing a rebuild, run all five first. My money is on number four being the one your team skips. If you would rather have someone ask them with you, talk to TrueForge before the deck decides for you.