A standup exists to surface blockers, not to report progress. Keep it to 15 minutes, hold it at the same time daily, and structure it around work rather than people - walk the board right to left so the closest-to-done items get attention first. The failure mode is the status meeting in disguise: everyone recites what they did, nobody says what is stuck, and the manager is the only person who benefits. If your standup would be equally useful as a written message, make it a written message and reclaim the time.
The daily standup is the most widely adopted and most frequently pointless meeting in agency work.
The intended version takes ten minutes, surfaces the two things that are stuck, and lets the team unblock each other before lunch. The common version takes twenty-five minutes, consists of six people describing their work to a manager while the other five wait their turn, and produces nothing that could not have been read.
The difference is not discipline or facilitation skill. It is what the meeting is understood to be for.
This guide covers what a standup is actually for, the format that works for agency teams, when to run one at all, how to handle multiple concurrent client projects, and the async alternative.
What it is for
A standup surfaces blockers. That is the whole purpose.
Everything else - progress, coordination, visibility - is a by-product, and every one of those by-products is better served by something else. Progress is better tracked on a board. Coordination is better done in the conversation the standup triggers. Visibility for managers is better served by a written update.
If you accept that framing, the format follows. The question is not "what did everyone do yesterday" but "what is stopping anyone from making progress today, and who can fix it."
The diagnostic: if nothing in your standup ever changes anyone's plan for the day, it is a status meeting. Status meetings are a legitimate thing to want, but they do not need to be daily, do not need to be synchronous, and do not need the whole team standing around.
The format that works
Fifteen minutes, hard stop. Same time every day.
Walk the board, not the people
The most useful single change, and it is counterintuitive.
The standard format goes person by person: what I did, what I am doing, what is blocking me. That structure optimises for individual accountability, which is why it drifts into status reporting - each person is implicitly justifying their day.
Walking the board instead means going through the work, right to left - starting with whatever is closest to done and moving back toward not-started.
Three things this fixes:
It prioritises finishing over starting. Attention goes first to the item one review away from delivery rather than the one someone began yesterday. Agencies chronically have too much in progress and not enough finished, and this format pushes against that daily.
It makes stuck work visible. An item sitting in "awaiting client feedback" for nine days is impossible to miss when you walk the board and trivially easy to miss when you go person by person, because the person waiting on it has other work to talk about.
It removes the performance. Nobody is reporting on themselves, so nobody needs to sound busy.
Three questions, about work rather than people
For each item on the board:
- Is this moving?
- If not, what is it waiting on?
- Who can unblock it, and by when?
That third question is where the value is, and it is the one most often skipped. A blocker named and not assigned is a blocker that will be named again tomorrow.
Take detail offline, visibly
Two people getting into implementation detail is the most common way a standup runs long. Name it and move on: "let's pick that up after - Priya, Dev, five minutes?"
Say it out loud rather than letting it happen after, so the group knows it is handled rather than dropped.
Who attends
The people doing the work, and whoever can remove blockers. Not stakeholders who want visibility - give them the written update instead. A standup with observers becomes a performance, reliably and quickly.
When not to run one
Standups are not universally appropriate, and agencies frequently adopt them because they are a recognisable practice rather than because the conditions apply.
Run a daily standup when: several people are working on the same project simultaneously, work is genuinely interdependent, and things change day to day.
Do not run one when:
People are working solo on separate client projects. Four people on four unrelated projects have nothing to unblock for each other. Their standup is four monologues. Twice a week is plenty.
The work is long-cycle. If a task takes six days, a daily standup produces five days of "still going." Move to twice weekly.
It has become a status meeting and cannot be recovered. Better to stop it than to keep a meeting everyone resents.
Mindful QA's observations on agency delivery make a related point about agencies specifically: agency work often does not fit the rigid ceremony structure product teams use, because the unit of work is a client engagement rather than a sprint. Adopting the ceremonies without the underlying structure gets the cost and not the benefit.
Handling multiple concurrent client projects
The specific agency problem. A product team runs one standup because there is one product. An agency has eight clients and people spread across several.
Three workable models.
Per-project standups, short and only when active. Ten minutes per project, attended only by people on it. Most accurate, most calendar overhead. Works when projects have three or more people.
One team standup, walked by project. Go through active projects rather than people, and only the people on each project speak for that portion. Fifteen minutes covers four or five projects. Best default for most small agencies.
A weekly portfolio review instead. For teams where everyone works solo across projects, a single weekly session covering the health of every active project is more useful than any daily meeting. This is closer to capacity planning than to a standup, and for that team shape it is the right instrument.
The mistake is running one standup for everything at a size where most attendees are irrelevant to most of the content. Six people sitting through eight projects when each works on two is forty minutes of collective time to deliver ten minutes of value.
The async alternative
For distributed teams, or any team where the standup is not earning its time, written works and often works better.
How it runs: each person posts a short update in a channel by an agreed time. Same three questions, focused on blockers. Anyone blocked tags whoever can unblock them.
What it is better at: timezones, focus - no mid-morning interruption to everyone's deep work - and creating a searchable record. It also produces better-considered updates, because writing forces clarity.
What it is worse at: the unplanned conversation. The genuine value of a synchronous standup is frequently the exchange that happens when two people realise they are working on overlapping things, and that does not happen in a channel.
The hybrid that usually wins: async written updates daily, one short synchronous session weekly for the conversations that need to happen live. Most agency teams land here after trying both.
The rule worth applying: if your standup could be replaced by a written message with no loss, replace it. Fifteen minutes a day across six people is roughly ninety hours a year. That is a substantial amount of billable capacity to spend on a meeting nobody would miss.
Adapting it to your team's shape
Three team shapes, three different answers.
Co-located, single project. The textbook case, and the standard format works. Fifteen minutes, board on screen, walk right to left. This is the only shape where a daily synchronous standup is unambiguously correct.
Distributed across timezones. Async written updates, posted by an agreed time, with one short synchronous session weekly. The written format is genuinely better here - it produces more considered updates and creates a searchable record - and the weekly live session covers the conversations that need to happen out loud.
Everyone on separate client projects. The most common agency shape and the one where daily standups make least sense. Four people on four unrelated projects have nothing to unblock for each other, so the meeting is four monologues. Replace it with a twice-weekly session walked by project, or a single weekly portfolio review closer to capacity planning.
The test that cuts through all of it: would anyone's day change because of what was said? If the answer is consistently no, the format is wrong for your team shape regardless of how standard it is.
Making the board carry its weight
A standup is only as good as the board it reads from, and most of the effort in improving standups is really effort in improving the board.
Three properties matter.
States that reflect reality, including waiting. Most boards have some variant of to-do, doing and done. Agency work needs at least one more: awaiting client. Without it, work blocked on someone else's decision sits in "in progress" and makes the team look slow when it is not their throughput problem. That single column change frequently improves both the standup and the conversation with the client.
One item per meaningful piece of work, not per tiny task and not per whole project. Too granular and the board becomes noise; too coarse and nothing ever moves, so the standup has nothing to read.
Visible age. How long has this item been in this state? Nine days in review is invisible on most boards and is exactly the thing a standup should catch.
Get those three right and the standup largely runs itself, because the board surfaces the blockers before anyone speaks.
The three-minute standup
Worth naming as a target rather than an accident. On a good day, when nothing is blocked and everything is moving, a standup should take three minutes and end.
Teams resist this. Fifteen minutes are in the calendar, everyone is present, and finishing in three feels like the meeting failed. It is the opposite - a short standup means the board is accurate, the work is flowing and the blockers are being caught elsewhere.
Say it explicitly the first few times: "nothing blocked, we're done, get your morning back." After a fortnight it stops feeling odd, and the meeting acquires a useful property - people start looking forward to it being short, which means they arrive having already checked the board.
What to do with what the standup surfaces
A standup generates two kinds of output and they need different homes.
Immediate blockers get an owner and a date in the meeting. That is the point of the meeting.
Recurring blockers need somewhere else to go. If "waiting on client approval" appears every day for two weeks, that is not a standup item, it is a client conversation about approval turnaround. The standup surfaces the pattern; it cannot fix it.
Watching for the difference is the most valuable thing a facilitator does. A team that raises the same blocker eleven times has a process problem that daily discussion is quietly normalising, and the fix belongs at the kickoff of the next project rather than in tomorrow's standup.
Facilitating one well
The difference between a fifteen-minute standup and a twenty-five-minute one is almost entirely facilitation, and it is a small set of learnable behaviours.
Start on time regardless of who is missing. Waiting teaches people that the start time is notional, and within a fortnight nobody arrives on time. Starting without them is uncomfortable once and then solves itself.
Keep the board on screen, always. The moment the meeting becomes people talking without a shared artefact, it drifts into status reporting. The board is what keeps the conversation about work rather than about individuals.
Interrupt early and warmly. "Let me stop you there - is this one blocked, or just in progress?" Facilitators who let a tangent run for ninety seconds before intervening have already lost the time. Interrupting at fifteen seconds feels rude and is not, because everyone in the room wants the meeting to be short.
Name the parking lot out loud. "Priya, Dev - can you two pick that up straight after?" Said publicly, so the group knows it is handled rather than dropped, and so the two people involved actually do it.
End when you are done, not when the time is up. A standup that finishes in five minutes on a good day is working correctly. Filling the remaining ten minutes because they are in the calendar is how a standup becomes a status meeting.
Rotate the facilitator. Prevents it becoming a manager's meeting and spreads the sense of ownership. It also reliably improves the format, because each person removes something they personally find pointless.
What to do about blockers that are not yours
A large share of agency blockers are external - waiting on a client decision, a third party, a supplier. The standup surfaces them and cannot resolve them, which is where teams get demoralised.
Three responses that help.
Distinguish "blocked" from "waiting" on the board. Blocked means someone here can act. Waiting means the ball is with someone else. Different columns, different conversations - and it stops the team feeling responsible for delays that are not theirs.
Give every waiting item an owner for the chase. Not the person doing the work necessarily - whoever owns the client relationship. Something waiting on a client for nine days needs a conversation, not another day of patience.
Escalate on a schedule rather than on frustration. Agree that anything waiting more than three days gets chased, and more than a week gets raised with the approver. Making it a rule removes the judgement call about whether it is too soon to nag, which is what causes the delay.
The pattern across all three: the standup's job is to make the waiting visible and assign the chase. It cannot make the client answer.
Standups and the rest of the delivery rhythm
A standup is one of four rituals and it works best when the others are doing their jobs, because a standup that carries the weight of all four becomes a long meeting.
The kickoff establishes who approves what and how fast. Without it, the standup fills with "who do we ask about this?"
The weekly client update handles progress communication. Without it, the standup becomes a status meeting because status has nowhere else to go.
The retrospective handles recurring problems. Without it, the same blocker gets discussed eleven times in standups without anyone changing the process that produces it.
The board holds state. Without a good board, the standup becomes the only place project status exists, which makes it long and makes anyone who missed it uninformed.
If your standup is running long, the cause is frequently one of those four being absent rather than the standup itself being badly run. Fixing the upstream ritual shortens the meeting more reliably than any facilitation technique.
The failure modes
It becomes a status report to a manager. The commonest. Symptom: people address the manager rather than each other. Fix: walk the board, and have the manager stop speaking first.
It runs long. Fifteen minutes means fifteen. Ending on time, every time, is what makes people willing to attend.
Blockers get named but not assigned. A blocker without an owner and a date recurs indefinitely. This is the single highest-value discipline in the whole meeting.
Nobody says they are stuck. The serious one, and it is cultural rather than procedural. If admitting a blocker feels like admitting failure, the meeting cannot do its job. The fix is the same as in a retrospective: the most senior person goes first and says what they are stuck on.
It replaces the board. If the standup is where project state lives, it is doing a job a board does better and asynchronously. The standup should be reading from the board, not substituting for it.
Replacing it with something better
If the diagnosis is that your standup is not earning its time, four alternatives are worth knowing, each suited to a different problem.
The written daily update. Same three questions, posted in a channel by an agreed time. Suits distributed teams and anyone whose deep work is being fragmented by a mid-morning meeting. The trade-off is losing the unplanned conversation, which is genuinely the best thing about a live standup.
The twice-weekly walk of the board. Tuesday and Thursday, twenty minutes, all active projects. Suits agencies where people work solo across several clients, because the daily version produces nothing but the weekly version is too slow to catch a stalled item.
The weekly portfolio review. Forty-five minutes, every active project reviewed for health rather than progress: on track, at risk, blocked, and what capacity looks like for the coming fortnight. This is closer to capacity planning than to a standup and it is the right instrument for a small agency running many small engagements.
Nothing, plus a good board. Genuinely viable for very small teams where everyone can see the state of the work and speaks constantly anyway. The risk is that blockers stay silent because there is no moment that asks for them, so it works best where raising problems is culturally easy.
The mistake is keeping a daily standup because it is what teams do, while everyone quietly agrees it is not useful. That costs roughly ninety hours of billable capacity a year across a six-person team, and the cost is invisible because it is spread fifteen minutes at a time.
What good looks like
Fifteen minutes. Same time daily. The board on screen, walked right to left.
Most items get five seconds - "moving, fine." Two or three get a real conversation, each ending with a name and a date. Anything detailed is taken offline out loud.
Nobody reports on themselves. The manager speaks least. And on a day when nothing is blocked, it takes four minutes and everyone gets on with their work - which is the point, and is also why a standup that consistently fills its fifteen minutes is worth examining.