Where the Daily Meeting Belongs in a Cadence

The question usually arrives with the calendar already half-drawn: what a business meeting cadence of daily, weekly and monthly should each contain. The daily gets assumed first, because it is the tier everyone has seen somewhere. Then the weekly inherits whatever the daily could not finish, the monthly inherits whatever the weekly could not finish, and within two quarters the three meetings are running one agenda at three speeds, with three grades of fatigue in the room.

The tiers are not speeds. Each one exists to absorb a different class of change, and the class is set by how long a problem can sit before it costs something. A blocked task costs a day. A trend bending the wrong way costs a month. Putting either in the wrong room is not a scheduling error; it is the reason the room stops being worth attending.

The question each tier can answer

A daily answers what is in the way right now. A weekly answers whether the work is on the path that was agreed, and what changes if it is not. A monthly answers whether the path itself is still right, given four weeks of evidence that no single week contained. Nothing travels upward that a lower tier could have resolved, and nothing travels downward that the lower tier has no authority to decide. That is the entire design. The difficulty is that only the weekly is genuinely universal — the other two have to be argued for.

A daily meeting is worth its cost in one condition: the unit of work is smaller than a week, and it can be blocked between one day and the next. An engineering team shipping in two-day increments meets that condition. A support operation working a live queue meets it. An incident meets it, which is why nobody has ever had to justify the frequency of an incident call.

A leadership team whose smallest unit of work is a week does not meet the condition, and a daily placed there will fill itself with the only material available, which is status. Fifteen minutes of eight people confirming that things are proceeding is not a cadence tier. It is a weekly with worse attendance economics, run five times.

A test that settles it in one question

Before adding a daily, ask what would actually go wrong if it were skipped for a week. The answers sort cleanly, and the sorting is usually the end of the debate:

• Something would sit blocked for four days that could have been cleared in four minutes. The daily is earned.
• Someone senior would feel less informed. That is a reporting need, and reporting has cheaper formats than a recurring meeting for eight people.
• Nobody in the room can name what would go wrong. The meeting is being added for reassurance, and it will be quietly cancelled within a quarter, often by whoever proposed it.

The third answer is the one we hear most, and it is not a failure of nerve. It is information: the team does not have a daily-sized problem, and building a tier for one it does not have costs more than the empty slot it fills.

The most common reason a daily appears is not that the work got faster. It is that the weekly stopped resolving things. Items arrive at the weekly, get discussed at length, and leave undecided. The following week they arrive again, slightly more urgent and slightly more expensive, and somebody proposes meeting daily so that nothing has to wait a full week for attention.

The weekly was not failing on frequency. It was failing because nobody in the room had the authority to close an item, and meeting more often does not confer authority. What follows is predictable: the unresolved items now circulate five times a week instead of once, the weekly becomes a summary of the dailies, and the cadence turns into a set of meetings about meetings. That is the point at which most teams abandon the redesign and go back to the calendar they had.

Meeting more often does not confer the authority to decide. It only lets the same item go unresolved five times a week instead of once.

The symptom that separates the two

A frequency problem looks like items arriving late: the team knew on Tuesday, and the room was eleven days away. An authority problem looks like items arriving on time and leaving unchanged. The first is fixed by adding a tier. The second is fixed by naming who decides, which is cadence built from decision rights rather than from meeting frequency and a different piece of work entirely. Adding a daily to an authority problem is the most expensive available way of not solving it.

If the daily is the tier most often added without cause, the monthly is the one most often collapsed into the weekly, usually by running the weekly with a longer deck once every four weeks. That loses the thing a monthly is for. Four weeks is the shortest window in which a trend separates from noise. A conversion rate that dipped in week two and recovered in week three said nothing worth acting on. The same rate down in three weeks out of four is a signal, and it was invisible to every weekly that produced it.

A monthly that reviews the last four weekly packs has not produced a monthly view; it has produced a longer weekly. The tier is doing its work when it can act on something no single week held — a drift, a cost line bending, a segment changing shape — and when it can leave the week's execution alone entirely.

What each tier hands upward

• Daily to the weekly: blocks that could not be cleared at the working level, with what was already tried.
• Weekly to the monthly: decisions made, and the ones deferred for want of evidence, which is what the monthly then goes looking for.
• Monthly to the quarter: the pattern now stable enough to change a plan over.

Read in the other direction, the same list says what may not travel. The monthly does not adjudicate a blocked task. The daily does not reopen a decision the monthly made. A cadence stays tolerable for as long as those two rules hold, and the weekly, monthly and quarterly tiers divide along the same logic once the daily question has been settled.

Most of the cadences we are asked to redesign do not need a new tier. They need the existing tiers to stop carrying each other's material, and that is diagnosable from artefacts already sitting in the business: two months of agendas, the minutes, and the list of items that appeared in three or more consecutive meetings without resolving. The repeat list is usually short, and it names the problem faster than any workshop will.

A tier is worth adding only when a class of decision has nowhere to go. Every other version of the problem is an ownership question wearing a calendar's clothes. That is the order the work runs in: read the cadence that already exists, find the decisions with no home, and change the calendar last. Where that thinking becomes work is consulting, and the neighbouring notes in operating cadence cover the tiers on either side of the daily.