Perspective #04 · Governance
The quarter we stopped agreeing on what "done" meant
Most execution problems are alignment problems wearing an execution costume.
Teams rarely disagree about the plan. They disagree earlier than that — about what finishing it would even look like.
Every quarter, somewhere in the world, a leadership team sits down to review why last quarter's roadmap slipped. Almost every time, the diagnosis reads the same way: priorities changed, resources were tight, dependencies surfaced late. All true. None of it is the actual cause.
The actual cause, in the rooms I've sat in, is almost always earlier and quieter than that. Somewhere in the first two weeks of the quarter, two senior people privately disagreed about what "done" meant for the same line on the roadmap — one meant shipped to the first customer, the other meant rolled out to the full base — and neither said so out loud, because the plan had already been agreed and reopening it felt like a step backwards.
That disagreement doesn't go away. It resurfaces in every steering meeting as a scheduling argument, a scope argument, a resourcing argument — anything except what it actually is. By the time it's named, three months and a great deal of goodwill have usually been spent arguing about the wrong layer of the problem.
The fix is unglamorous: a single, short definition of "done" per priority, written down before work starts, with one person accountable for defending it if it gets renegotiated later. It sounds too simple to matter. In the organisations where I've introduced it, it has been the single change that shortened steering meetings the most — not better facilitation, not clearer OKRs, just an agreed answer to a question everyone assumed had already been asked.