Standups That Are Worth the Fifteen Minutes

A daily standup earns its fifteen minutes when somebody leaves it doing something different from what they planned to do when they joined. An afternoon re-pointed at the thing that is actually at risk, a blocker picked up by the one person who can clear it, two people discovering they are both halfway through the same migration. If none of that happens, you have run a status broadcast, and a status broadcast belongs in a written channel where people can read it in ninety seconds and skip the parts that do not concern them.

Most standups fail on altitude rather than format. The team inherits the three questions - what did you do yesterday, what are you doing today, what is blocking you - and answers them at task level, which is the one level nobody else in the room needs. Nine people, sixty seconds each, and everybody has sat through eight updates that will not change a single decision they make before lunch.

What follows is a shape that genuinely fits inside fifteen minutes, a way of handling blockers so they get removed rather than announced, what actually changes for remote, hybrid and asynchronous teams, how the format adapts outside engineering, and an honest two-week test for whether yours has quietly become theatre.

What the fifteen minutes is actually for

A standup is a coordination meeting, not a reporting meeting. That distinction is not pedantry - it decides who the audience is. In a reporting meeting the audience is the manager, and eight people take turns performing diligence for one person. In a coordination meeting the audience is the other people doing the work, and you only say the things that might change what they do today.

That leaves three legitimate outputs, and only three. A shared picture of where the work stands against what the team said it would deliver. A reallocation of effort, where somebody moves onto the item that is at risk. And a short list of blockers, each leaving the meeting with a named owner and a date, so that somebody is responsible for clearing it before tomorrow.

Everything else is optional or belongs elsewhere. Yesterday's completed tasks are already visible on the board. Today's plan is only interesting at the points where it collides with someone else's. Demo previews, architecture debates, sprint planning and wellbeing conversations are all legitimate activities that are not this activity, and quietly pushing them into the fifteen minutes is the most common reason a standup runs to thirty-five.

The test to apply for a fortnight is simple. At the end of each standup, ask whether anything moved - not whether the updates were informative, but whether a plan, an owner or a priority actually changed. Two weeks of nothing moving is not a sign that you need a better facilitator. It is a sign that this team is not especially interdependent right now, and the meeting should go weekly or go away.

Why the three questions stop working

The three questions were designed for a co-located team of about seven, standing up, with the work physically in front of them on a wall. Remove the wall and the standing, scale to fourteen people across three time zones, and the questions turn into a sequence of individual performance reviews delivered to a manager who did not ask for them.

Two things go wrong at once. The answers drift into the past tense, because describing yesterday is safe and committing to today is not. And the group grows past the point where the marginal listener has any stake in the marginal update - the standard conditions for effort to leak out of a group, where each person contributes slightly less because responsibility is spread thinner across more people.

You can watch the drift happen over a couple of sprints. Updates get longer as people justify their day rather than coordinate it. Blockers get softened to 'nothing major', because naming one in front of eleven colleagues is socially expensive. Somebody starts debugging out loud and six people go silent for four minutes. The meeting is still fifteen minutes on the calendar and twenty-eight in practice, and the two people who needed to talk have not talked.

None of that is fixed by asking people to be more concise. It is fixed by changing what the meeting walks through, who it is addressed to, and what happens to a blocker after it is spoken.

The four numbers that keep a standup honest

These are working constraints rather than research findings, and they are worth stating to the team out loud, because a constraint nobody has named cannot be enforced without it feeling personal. The fifteen minutes is a hard stop: if the meeting is allowed to run when it is interesting, it will run every day.

The group size matters more than the per-person discipline. A round of eight at sixty seconds is eight minutes and leaves seven for blockers and re-planning. A round of fourteen is fourteen minutes with nothing left, which is precisely how standups turn into pure broadcast. If you are above ten, the fix is to split the group, not to speed everybody up.

A fifteen-minute shape that holds

The ratio is the point. Seven minutes on the work, three on blockers, three on moving effort around, one on closing cleanly. Most standups spend fourteen minutes on the first block and thirty seconds on the rest, which is why they feel informative and change nothing.

Run it in this order for a fortnight before you adapt it. The temptation is to keep the round of people and add the board walk on top, which gives you a twenty-two minute meeting and a team that quietly stops attending. The board walk replaces the round; it does not supplement it.

Walk the board, not the people

The single highest-leverage change you can make is to stop going around the circle. Instead, pull the board up on screen and walk the work items, starting from the one closest to done and moving backwards towards the ones not yet started. Whoever is on an item speaks to it for up to a minute; everyone else stays quiet.

Right to left matters because it puts finishing ahead of starting. Teams that walk left to right spend their attention on the exciting new work and let three nearly-finished items sit in review for a week. Walking from the right forces the question that actually unblocks delivery: what is stopping this one thing crossing the line today?

It also removes the confession round. When you go person by person, everyone with a quiet day has to manufacture something worth saying, and the people whose contribution is not visible as tickets - the ones doing support rotation, review, mentoring, sales enablement - sound like they did nothing. Walking items means silence is structural rather than personal, and nobody performs busyness to fill it.

The tell that this is working is that some people barely speak on some days and nobody minds. The tell that it is not working is that you have simply renamed the round: eight items, eight people, in the same order every morning. If that happens, look at whether your board reflects the real work or is a per-person to-do list with a Kanban skin on it.

Make a blocker cost something

Blockers are the part of the standup that justifies doing it live at all. Status can be written; a person publicly committing to unstick another person is much harder to do asynchronously, and it degrades fast when there is no witnessed moment attached to it.

The reason blockers get softened is that raising one implies you have been stuck and did not sort it yourself. Lower that cost deliberately: raise your own blockers first as the lead, use the word 'stuck' without hedging it, and never respond to a blocker with a question that sounds like an audit.

Choosing a format for your team's shape

Pick the format from the shape of the team, not from what the framework diagram shows. The daily live call is the right default for a small team whose work genuinely overlaps, and the wrong default for six specialists who each own a separate surface and have not needed each other since August.

The two-tier pattern deserves a warning. It works, but only if the squad standups stay the substantive ones and the leads sync is deliberately dull - a five-item cross-squad dependency check, nothing more. The moment the leads sync starts making decisions, the squad meetings become ritual and attendance rots within a month.

Remote, hybrid and asynchronous standups

Remote standups need a driver and a shared screen. Someone shares the board, moves the cursor down the items, and says the next name explicitly - 'that is done then, Sam, the export job'. Without an explicit handoff, every transition costs three seconds of nobody-goes-then-two-people-go, which across eight items is nearly a minute of dead air and a general sense that the meeting is limp.

Hybrid is harder and mostly fails in one specific way. Four people sit in a room around one laptop, three dial in, and the in-room four have a fast, half-audible conversation that the remote three catch about sixty per cent of. If you run hybrid standups, put everyone on their own device with their own audio, including the people sitting together, or accept that the remote members will stop raising blockers within a fortnight.

Asynchronous standups work well for teams whose work does not collide daily, but they need stricter rules than live ones, not looser. A good written update is two lines: whether the commitment is still on track, and what you need from a named person. Nobody needs your task list. Set an explicit response expectation - blockers answered within four hours - and have one person read the whole thread each morning and escalate anything nobody replied to.

The honest trade-off is that asynchronous formats lose the moment where somebody visibly takes on your problem in front of the group. Teams that go fully async and keep delivering almost always retain one live session a week where blockers and re-planning happen out loud. It is the live commitment, not the daily cadence, that does the work.

Standups outside engineering

The format travels well as long as you can name a unit of work to walk. Sales walks the deals that move this week, not the calls people made. Support walks the queue: oldest ticket, escalations, and anything a customer has now chased twice. Marketing walks the campaigns in flight with a launch date inside the next fortnight. Operations walks incidents and the two processes currently changing.

Where non-engineering standups struggle is that there is often no board at all, so the meeting defaults to a round of people describing activity. The fix is to spend twenty minutes once building a five to eight item list of live commitments with a name and a date against each. That list is your board. It does not need a tool; a pinned document works, provided it is genuinely maintained and not quietly abandoned in week three.

The second difference is cadence. Many non-engineering teams do not have work that changes meaningfully every twenty-four hours. Three mornings a week is a perfectly respectable standup rhythm, and it beats a daily one that half the team has learned to attend with the microphone muted and another tab open.

The thirty-second pulse check

A standup tells you what is happening to the work and almost nothing about what is happening to the people doing it. Confidence in the sprint goal, load, and whether anyone is quietly drowning are all things that get answered with 'yeah, fine' in a round of eight and answered honestly in one anonymous tap.

One question, one tap, once or twice a week: how confident are you that we hit Friday's commitment, one to five. It costs ten seconds of meeting time, and it is the trend that carries the signal - a team that answers four, four, three, two over a week has told you something a round of verbal updates never would. Session Flo runs this as a live pulse check from the same room code you already use, with the responses anonymous and the trend visible across sessions.

Two rules make it useful rather than decorative. Show the result on screen immediately, so people see their answer land, and say something about it out loud - even 'that is the third day of twos, so we are re-scoping this afternoon'. A pulse check whose result is never mentioned is a form, and teams stop filling in forms by about the second week.

The five ways a standup dies

First, the status broadcast: everyone reports, nobody coordinates, and the information could have been read faster than it was spoken. Second, the manager audience: updates get addressed to one person, and the tone shifts from coordinating with peers to justifying a day. You can hear this one - people start sentences with 'so basically I just' and finish with an apology.

Third, the solving round, where two people debug in front of six who cannot contribute. Fourth, the growing room: someone invites a stakeholder for visibility, then two more, and now sixteen people attend a meeting designed for seven. Every extra attendee makes speaking slightly more expensive and listening slightly less rewarding, which is exactly how contribution erodes as groups grow.

Fifth, and most corrosive, the unclosed blocker. A team that names blockers for three weeks and watches none of them get cleared will stop naming them - not out of apathy, but from a rational read that saying it out loud costs something and buys nothing. Once that happens, the meeting is a fifteen-minute ritual that reports good news, and it will need more than a format change to recover.

Notice that four of the five are about who is in the room and what happens after the words, not about the words themselves. Almost nobody fixes a dying standup by tightening the script.

A two-week test for whether yours is working

Tick these off privately for ten working days rather than asking the team how they feel about standups, which reliably produces polite answers. Scores are not the point; the pattern of which line fails repeatedly is.

The last item is the strongest test available and the one most teams will not run. Make attendance genuinely optional for two weeks and see who keeps coming. A standup that people attend because it is in the calendar and a standup that people attend because they need what is in it look identical from the outside and are entirely different meetings.

If the honest answer is that half the team would stop coming, do not respond by mandating attendance. Respond by cutting the group, moving to three days a week, or narrowing the meeting to blockers only for a sprint and seeing whether anyone misses the rest of it.

When to stop running a standup

A daily standup is a coordination cost you pay because the work is entangled. When the work stops being entangled, the meeting stops paying for itself, and continuing it out of loyalty to a framework is how teams end up defending ceremonies they privately resent.

Three signals say it is time to change. Nothing has moved as a result of the meeting for a fortnight. The same three people speak substantively every day and the rest give one-liners. Or the team has drifted into parallel workstreams where the overlap is genuinely once a week. Any of those means go to two or three days, or to written updates with one live blocker session.

Kill it deliberately rather than letting it rot. Say what is replacing it - the channel, the board, the weekly - name who to go to when stuck, and put a date in the calendar to review the decision in six weeks. Teams that abandon standups quietly usually rediscover in about a month that nobody knows what anyone else is doing, and then reinstate a worse version of the meeting they just escaped.

The goal was never the ritual. It was fifteen minutes in which the team could see the same picture of the work, move somebody onto the thing that is at risk, and leave with fewer people stuck than when they arrived. Any format that does those three things is a good standup. Any format that does not is fifteen minutes you could give back.