A decision that reopens has not been made, and the time spent making it the first time is spent again.
Why it matters when the plan changes
Velocity is usually treated as a cultural attribute and is mostly a structural one. The first explanation for a slow decision is that the leader lacks pace, but the more common causes sit in ownership: authority split between equal parties, evidence that arrives late, or a question that crosses a boundary nobody owns. These sit at the level of ownership, not design, so the org chart looks fine while the decision stalls. Each cause is fixable by assignment, not exhortation, and each is invisible in a discussion about pace.
The tension is with quality and commitment. Deciding fast without the parties who must execute it produces a decision that reopens in practice, costing the time twice. Donald Sull, Rebecca Homkes and Charles Sull found only 9% of managers say they can rely on colleagues in other functions and units all the time, which is where cross-unit decisions unravel. Naming one accountable owner, the RACI convention, lets a decision move without waiting for consensus that was never coming.
In practice
A leadership team is told it is too slow and responds by shortening meetings and deciding faster. Three of the next five decisions are reopened within a month, because the people who had to deliver them were not in the room. The organisation now decides quickly and holds nothing, which is slower than before.
Evidence
Cross-unit commitments are the least reliable part of execution, which is where decisions stall or fail to hold.
Donald Sull, Rebecca Homkes and Charles Sull, Why Strategy Execution Unravels and What to Do About It, Harvard Business Review (2015)The standard responsibility convention makes one party individually and ultimately accountable, which is what unblocks a stalled decision.
Responsibility assignment matrix (RACI), Wikipedia (2026)
What it cannot tell you
Decision velocity measures elapsed time from question to decision holding, and says nothing about whether the decision was correct. A fast decision reached without the evidence or the people needed to execute is not a good decision merely because it was quick, and the measure cannot distinguish speed from soundness on its own.
Questions
By the elapsed time from a question arising to a decision being made and holding, counted on decisions that actually matter. Counting decisions announced rather than decisions that held will overstate velocity in exactly the organisations where reopening is the underlying problem.
Usually because authority is split between parties with equal standing, because the evidence needed arrives after the moment it was useful, or because the question crosses a boundary nobody owns. Rarely because an individual is indecisive, though that is the first explanation offered.
No. A decision made without the people who have to execute it will be relitigated in practice, which costs more than the time saved. The target is a decision that is made once and holds, not the shortest possible interval to an announcement.
Name one accountable owner for each decision that recurs at a boundary, and write down what happens when the consulted parties disagree. Most persistent slowness comes from a small number of unassigned recurring decisions rather than from any general cultural pace.
Consensus cultures trade speed for commitment, which is a real trade rather than a failure. They become a problem when consensus is required for decisions that do not need it, and the fix is to name which classes of decision require agreement and which require only consultation.