Why the worst constraint decides
Here is a commitment: twelve German-speaking agents, go-live in September. Sales confidence is high. The workforce plan is solid. The margin clears the floor. And the platform integration the service depends on is not safe until nineteen days after the contractual go-live.
What is the health of this commitment?
If you average, it looks fine — four strong dimensions, one weak one. If you ask the only question that matters — can we actually keep this promise on this date? — the answer is no. The technology constraint binds, and no amount of strength elsewhere compensates, because the business does not experience averages. It experiences whichever constraint fails first.
Convergence, not averaging
The honest computation is convergence: take every function’s clear-by date — the day the workforce is staffed, the day the platform is safe, the day the contract allows — and commit to the earliest date that clears all of them. One dimension moves, the answer moves. That is not pessimism; it is how delivery actually behaves.
The consequences of taking constraints seriously
- Recommendations become dates, not scores. “Commit for 19 September” is actionable. “76% healthy” is not.
- Every option shows what blocks it. In a Decision Room, an option that cannot clear its constraints is marked blocked — with the constraint named, so the argument is about reality rather than optimism.
- One unvalidated assumption can hold everything. If a launch-critical assumption is unmeasured, it binds — and should. The alternative is discovering it in production, with a penalty clause attached.
The worst constraint deciding is uncomfortable exactly once: before you commit. Averages are comfortable twice — before you commit, and then never again.
Common questions
What is a binding constraint?
A binding constraint is the single requirement that currently prevents a commitment from being safe — the workforce plan that staffs after the launch date, the platform integration that completes after the contractual go-live, the licence pool with zero seats available. Other dimensions can be perfect; the binding constraint decides.
What does convergence mean in decision-making?
Convergence takes every function's clear-by date — when hiring is staffed, when technology is safe, when the contract allows — and recommends the earliest date that clears all of them. It is the opposite of averaging: one late dimension moves the whole answer.
Why is averaging dangerous for commitments?
Because an average lets strong dimensions cancel a fatal one. A commitment scoring 90 in four areas and 20 in one averages to 76 — which looks committable. But the business does not experience the average; it experiences the 20. The constraint fails, and it takes the commitment with it.