Slow Decision Making Isn't Caution: The Decision Clock

Your last pricing decision took five weeks. Nobody wasted a week. The analysis was real, the two options were genuinely close, and everyone involved had other work that could not wait either. Reconstruct it afterwards and every individual step looks defensible.

Is slow decision making a sign of caution?

Not reliably, and the assumption is worth testing before you build anything on it.

The caution story is appealing because it flatters everyone: if decisions take time here, it must be because the stakes are high and we do our homework. It is also unfalsifiable as stated, explaining every delay equally well, which means it predicts nothing.

Kathleen Eisenhardt put it under load. Her inductive study of eight microcomputer firms asked how executive teams make rapid decisions in a fast-moving industry, and the pattern runs against the intuition: fast decision makers use more, not less, information than slow decision makers, and develop more, not fewer, alternatives, through what she called a two-tiered advice process. She reports that conflict resolution and the integration of strategic decisions with tactical plans are critical to decision pace, and she proposes that fast decisions built on this pattern lead to superior performance (Eisenhardt, 1989).

Read the scope before the conclusion. Eight firms, one US industry, the 1980s, an inductive design built to generate propositions rather than test them at scale. I do not read that as settling what happens in a forty-person company in Cluj or Copenhagen. I read it as removing one defence: slowness is not by itself evidence that more information was gathered, because in the sample where somebody looked, the better-informed teams were the faster ones.

Which leaves the question open rather than answered. If the extra weeks are not buying information, what are they buying? Sometimes something real: a board member on holiday, a legal opinion, a customer who has to be asked. Sometimes nothing at all. The difference is measurable, and almost nobody measures it.

Why does a decision sit after it is ready to close?

Three structures account for the waiting interval in the engagements I run, and each maps to a different dimension of the decision system. None of them is about effort.

No named closer. The decision is ready, everyone can see it is ready, and no single person holds the right to end it. So it waits for a room, a consensus, or a moment when enough of the relevant people are in one conversation to make closing it feel safe. This is Decision Clarity, and of the three it is the one I find cheapest to fix and the one I most often see mistaken for a people problem. The giveaway is that it closes instantly once somebody with standing says “we are doing option B”, and everyone's private reaction is relief rather than surprise.

An input with no deadline attached. There is a named closer, and the closer cannot act alone. They need a consult, a ruling on a seam between two functions, or a sign-off from above. That is often correct design. What is almost never designed is the clock on it. A consult right without a deadline is an open-ended veto held by whoever is slowest to reply, and it stays invisible because nobody refused anything. This is Escalation Discipline, and the fix is a stated window rather than a removed requirement: the consult has four days, after which the closer closes.

One room closes everything. The calendar becomes the queue. If the only forum that can close a decision meets weekly, a decision that becomes ready on Wednesday waits for Monday, and one that becomes ready on Monday afternoon waits six days. Spread arrivals evenly across the week and the average wait after readiness is half the cadence, before anybody deliberates about anything. That is arithmetic about the cadence rather than a finding about your team, and I rarely see it counted, because it never shows up as anybody's delay. This is Leadership Load Balance, and it is why leadership meetings that do not produce decisions cost more than the hour they occupy.

Decision Authority Dependency sits across all three. We score it alongside the five dimensions as a moderator, never inside them. The more of the system actually runs through one person, whatever the org chart says, the longer the waiting interval gets for everything that person touches. A founder who delegated a decision class on paper but kept the habit of being consulted has not shortened the queue, only made it harder to see, because the delay now arrives as courtesy rather than as escalation. This is where the Delegation Envelope earns its keep: a limit nobody wrote defaults to zero, and a closer unsure where their authority stops waits, which is free for them and expensive for you.

Three causes, three different fixes, and no way to tell which you have without the second stamp.

How do you time your own decisions?

Five steps, about ninety minutes, and you need your calendar and your sent mail rather than your memory.

1. Pick the last five decisions you closed personally. Not the five most painful ones, which bias the result toward drama. The last five, in order, whatever they were. Two of them being small is information rather than a flaw in the sample.
2. Reconstruct the three stamps from the record. Search the thread for when the question first appeared, find the message where the last needed input arrived, and find the message where you closed it. Days are precise enough. Where you cannot find the ready stamp, that is itself a finding: a decision whose readiness nobody noticed was never waiting on information.
3. Total the two intervals separately. Gathering days against waiting days, across the five. Resist averaging them into one number, because the average is what you have been carrying around already.
4. For every long waiting interval, name which of the three structures caused it. No named closer, an input with no deadline, or one room closing everything. If two apply, write both; they stack rather than compete. If none applies, you have found a fourth cause and it deserves more attention than my three.
5. Change one structure and measure the same five classes next month. One, not three. The second measurement exists to tell you whether the fix worked, and three simultaneous changes make that unreadable.

The shape below recurs across the engagements I run. It is not one company's story, and I offer it as an illustration of a pattern rather than as a measured result. The waiting total comes out larger than the team expected, the biggest single contributor is the weekly forum nobody counted as a bottleneck because it always starts and ends on time, and the hardest conversation is not about the forum. It is about the two decisions where the closer was named, available and sure of the answer, and still waited to say it out loud in front of other people.

No cadence change will touch that last one. The same mechanism sits under chronic escalation: closing a decision alone carries a cost that waiting for company does not, and nobody needs that explained to them.

So the question I would put to your leadership team this week is not how fast you decide. For the last decision that frustrated you with how long it took, when exactly did it become closeable? If nobody can answer, the five weeks were never five weeks of deciding, and the fix is not in the room where you have been looking.