The Operating Cadence a Company Is Running

Every company runs an operating cadence, whether or not anyone chose it. There is a leadership meeting on Monday, a business review at month end, a planning cycle each quarter, and a budget once a year. Between them they decide when a number gets looked at, when a plan may be altered, and when someone has to account for a miss. Almost nobody calls this strategy. It is the calendar.

But the calendar sets an upper bound on how fast the business can respond to anything. A company that reviews commercial performance monthly cannot correct a commercial problem in three weeks, however capable its people are, because the moment at which the problem becomes visible to the people who can act on it does not arrive until the fourth.

What the calendar is actually scheduling

An operating cadence is not an administrative preference; it is the resolution at which a company can see itself. A quarterly review resolves problems that last a quarter. Anything that appears and reverses inside six weeks reaches that forum only as an anomaly already corrected, described by the person who corrected it. The forum receives a story instead of a signal, and it receives it too late to test.

Most cadences are not designed against this. They are inherited: from an investor's reporting template, from the last company the operations lead worked in, from the shape of the financial year. Each of those encodes a business with its own decision speeds. None of them encodes this one.

The rhythm that works at forty people is one where the leadership meeting is also the decision. Everyone with context is in the room, the discussion and the resolution occupy the same hour, and nothing has to be written down for the decision to hold, because the people who will execute it heard it made.

At three hundred people that same meeting still runs, at the same time, with roughly the same people in it. It is no longer the decision. It has become a status readout followed by a set of unresolved items that get settled in side conversations across the following week, which is where the decision now actually happens, unrecorded and unevenly witnessed.

The meeting still runs at the same time with the same people. It is no longer the decision.

The symptom presents as a discipline problem

The company reads this as behaviour. People are not preparing. The agenda is loose. Decisions are not getting made, so the fix is a firmer chair, a pre-read, a decision log. The rituals get tightened around a forum whose job has quietly changed, which produces a well-run meeting that is still not where the decisions are.

The honest version of the diagnosis is duller. The company outgrew the assumption that one room could hold enough context, and nobody moved the decision to a forum sized for the new company, because there was never a week in which that was the most urgent thing to do.

Cadences accumulate. A recurring forum is created to resolve one specific thing, at a frequency that matched the urgency of that thing: a launch, an integration, a quarter that went badly. The thing gets resolved. The forum stays, because cancelling it requires someone to say out loud that it is no longer needed, and its cost is spread thinly across other people's weeks rather than concentrated anywhere visible.

• A weekly forum created during an incident, still weekly two years after the incident.
• A monthly review that exists because a former investor once asked for a monthly pack.
• A quarterly planning cycle inherited whole from a company with a longer sales cycle and a different cash position.
• A steering group that outlived the programme it steered and now steers whatever is nearest.

The load is real even when the meeting is empty

None of these is expensive alone, and each is defensible in isolation, which is why they are never removed together. In aggregate they set the company's response time. A leadership team's capacity to decide is not consumed by its decisions; it is consumed by the standing commitments that surround them. The calendar fills, the attention available for anything unscheduled falls, and the business becomes structurally worse at the one thing it did not plan for, which is the category most decisions of consequence fall into.

This is the same subtraction problem that shows up whenever a company adds faster than it removes. Cadence is where it is easiest to see, because every addition is in writing, in the calendar, with names attached.

The useful question is not how often the leadership team should meet. It is this: for each decision the company makes repeatedly, how long does it take to find out whether the last one was right? That interval is the only defensible input to a frequency, and it is knowable.

A pricing change in a self-serve product returns a readable signal in days. A change to an enterprise sales motion returns one in two or three quarters, because the cycle it affects is that long. Reviewing both monthly means reviewing one too slowly to steer it, and the other so often that the team is reading noise and acting on it.

Different loops, different rhythms

Most companies are running three or four distinct loops at genuinely different lengths, with a single cadence laid flat across all of them.

• The demand loop, which returns a signal in days to weeks.
• The delivery loop, which returns one in weeks to a quarter.
• The commercial model loop, covering price, segment and channel, which returns one in two to four quarters.
• The capability loop, hiring and skills, which returns one in a year or more.

Matching frequency to loop length is unglamorous and it is most of the work. It usually produces fewer forums rather than more: a fast one that reads the fast signal and is allowed to act without escalating, and a slow one that can afford to be slow because nothing needing a decision this week is waiting behind it.

It also makes the reversible and irreversible split operable. Reversible decisions belong in the fast loop, where being wrong costs a cycle. Irreversible ones do not belong in a recurring meeting at all; they belong to a session convened for them.

The companies that fix this do not redesign the calendar first. They start with the decisions the business actually makes on repeat, and the last four times each one was made: where it was decided, by whom, and how long after the evidence arrived. The distance between where decisions are supposed to happen and where they do is the whole finding, and it is usually available in an afternoon.

Removing before designing

What follows is subtraction. A cadence improves almost entirely through the removal of forums rather than the design of better ones, and the removals are political in a way the design work is not. Every standing meeting is somebody's visibility, and one or two of them are the only place a function gets heard at all.

This is part of what we do in consulting work, and the part teams most often want to carry themselves afterwards, which is where training belongs: reading the decision record before touching the calendar, matching each forum to the loop it is meant to close, and naming what is being removed rather than letting it lapse. It sits inside the argument we keep returning to under operating cadence — that a company's real operating model is whatever its calendar is built to notice.