task-ops
A daily standup meeting that has gone wrong looks almost the same as one that works. Eight people are on the call. Everyone speaks in turn. It ends within twenty minutes. Nobody complains out loud. And yet the same blocker surfaces on Thursday that surfaced on Monday, the person who needed help waited three days to ask for it, and the schedule slipped without anyone noticing until the deadline was a week away.
The usual fix is to tighten the meeting: cut the time, stand up literally, ban laptops. That treats the symptom. The standup turns into a status report when the status is not visible anywhere else, and the only way to get it is to make each person recite it. Fixing that is a question about the board, not about the meeting.
Most teams inherit the standup from Scrum without reading the source, which is short and more permissive than the folklore around it. The 2020 Scrum Guide describes the Daily Scrum as a 15-minute event for the Developers of the Scrum Team, held at the same time and place every working day of the Sprint. Its stated purpose is to inspect progress toward the Sprint Goal and adapt the Sprint Backlog, adjusting the upcoming planned work.
Two things in that description are routinely ignored. The first is who it is for. The event is for the people doing the work, and the Product Owner or Scrum Master participate only when they are actively working on Sprint Backlog items. A standup where six people report to one person who then reports upward is a different meeting wearing the same name.
The second is what the format is. The guide states that the Developers can select whatever structure and techniques they want, as long as the event focuses on progress toward the Sprint Goal and produces an actionable plan for the next day of work. The three questions are not in the 2020 text at all. They were in earlier versions and survived as habit. Teams that treat them as a rule end up defending a form that the rulebook stopped requiring.
That matters practically. If the three questions are producing a recital, a team following Scrum is free to replace them tomorrow. The two constraints worth keeping are the 15 minutes and the same time every day, because both exist to make the event cheap enough that skipping it is never tempting.
The classic form asks what each person did yesterday, what they will do today, and what is in the way. The reason it caught on is that it is impossible to get wrong, and a new team needs a form it cannot get wrong.
The failure starts when the board becomes accurate enough that the first question is redundant. If a card moved from In Progress to Review yesterday, saying so out loud adds nothing. The answer becomes a performance of effort rather than information, and people start padding it because a short answer sounds like a slow day.
The second failure is that the questions are addressed to the room rather than to the work. Person by person, the meeting covers eight separate plans, and nobody holds the shape of the whole. The dependency that matters, where one person's Thursday depends on another person's Wednesday, sits between two reports and gets mentioned by neither.
The alternative that fixes both is to walk the board instead of the roster. Start at the column closest to done and move backward, card by card, and ask what is needed to move each one. Work that has stalled announces itself. Work that nobody is touching becomes visible in a way that no round of personal updates will produce, because a card with no owner has nobody to speak for it.
The third question, about blockers, is the one to keep in every format. It is also the one most often answered dishonestly, and the reason is usually that the last three blockers raised in the meeting were not resolved. A blocker question only gets honest answers when raising a blocker visibly causes something to happen the same day. If naming one just adds a line to someone's notes, people stop naming them and route around the obstacle privately, which costs days.
Fifteen minutes times eight people is two hours of paid attention a day, which is real but not the largest number in play. The larger cost is the interruption. A standup at 11:00 splits the morning in half, and for anyone doing work that needs a long uninterrupted stretch, the hour before it is spent knowing it is coming.
This is why the time of day is worth more argument than the length. Right after the start of the working day is the common answer, and it works when everyone starts at roughly the same hour. It works badly on a team with flexible hours or two time zones, where it pushes some people to shape their day around a meeting that adds nothing to their morning.
The second hidden cost is the meeting after the meeting. A standup that surfaces a real problem cannot solve it in the remaining four minutes, so it produces a second conversation with two or three people, and that conversation is where the work actually gets unblocked. Teams that treat this as a failure try to solve everything inside the fifteen minutes and end up at forty. Teams that treat it as the intended output name the follow-up out loud, say who is in it, and move on.
The third cost is the one nobody counts: the standup can become the only moment when the shared picture exists. If the board is only trustworthy for the fifteen minutes after the meeting, then everyone spends the remaining twenty-three hours working from memory, and the schedule that gets reported upward is a reconstruction. A board that holds the schedule, the cards and the conversation in the same place is what makes the meeting optional rather than load-bearing.
The format follows from how the team is spread out, not from preference. Each one buys something and gives something up.
| Format | Works when | Gives up | Time cost per person |
|---|---|---|---|
| In person, at the board | Everyone is in one place on the same days | Nothing is recorded for anyone absent | 15 minutes |
| Video call, all on at once | Hours overlap by at least two hours | Side conversation, reading the room | 15 minutes plus context switch |
| Written, posted by a deadline | Hours overlap little, or focus time is scarce | Immediate back and forth | 5 to 8 minutes, no interruption |
| Written plus one call a week | Most weeks are routine, some are not | Daily face time | Lowest total |
The written form is the one most often adopted for the wrong reason and abandoned for a good one. Adopted because the meeting felt like a waste, abandoned because nobody reads the posts. A written standup only works with two rules attached: a posting deadline, and a named person whose job that morning is to read all of them and start the one conversation they imply. Without the reader, written standups become a log that satisfies the ritual and informs nobody.
For distributed teams, the useful hybrid is written daily with one synchronous call a week, held on the day when the week's plan is set. The daily post covers the blocker question. The call covers the shape of the week, which is the part that text handles worst.
Three mechanics do most of the work, and none of them require anyone to stand.
Start at the scheduled minute with whoever is present. Waiting for the last two people teaches everyone that the start time is soft, and the meeting drifts later by a minute a week until it is a twenty-five minute meeting starting at 11:07.
Have the board on the screen, and let it drive the order. Screen sharing a Gantt view or a kanban board changes what people say, because a claim that does not match what is on the screen is visible to everyone. It also removes the need for anyone to describe the state of things, which is where most of the minutes go.
Park anything that needs a decision. The phrase to use is short and specific: name the topic, name the two or three people, name the time. A meeting where four topics get parked and four follow-ups happen is working correctly, even if it feels less satisfying than resolving things live.
One more habit is worth adopting deliberately. Rotate whoever runs it. When the same person runs the standup every day, the meeting becomes a report to that person, and everyone else optimizes their answer for the audience of one. Rotating breaks that, and it also spreads the knowledge of what a good standup feels like, which is hard to transfer any other way.
The path to a short standup is to move things out of it into places where they live better.
Progress on individual cards belongs on the board. If the board is current, nobody needs to say it. Getting it current is a habit problem and worth attacking directly: cards move when the work moves, not the night before the meeting.
Deadline collisions belong on a schedule view. A list of cards does not show that three deadlines land on the same Friday, and a person reading their own tasks aloud will never notice it. A Gantt or calendar view shows it without discussion.
Longer questions and decisions belong in a thread attached to the card, where the reasoning stays next to the work. A decision made in a call and repeated in chat is a decision that will be re-litigated in six weeks, because nobody can find it.
Handovers and cross-team dependencies belong in writing with a date. Standups are good at surfacing these and bad at holding them, because the mention lives only in memory.
What is left after all that is genuinely short. Blockers, the day's plan where it differs from what the board implies, and anything that changed outside the team. Five minutes covers it, and the meeting becomes worth attending because nothing in it is something anyone could have read.
There are a few reliable tells, and each one points at a different fix.
People report to one person, not to each other. Look at where everyone's eyes go, or who gets addressed on a call. The fix is rotating the facilitator and walking the board rather than the roster.
The same blocker appears twice in one week. This means blockers are being logged, not resolved. The fix is to name an owner and a next step before the meeting ends, every time, with no exception.
Nobody can say what the deadline is. If asked what ships next and when, and the answers vary by person, the meeting has been reporting activity rather than progress toward anything. The fix is a visible target on a shared schedule, referenced at the start of every standup.
The meeting is the only time the board is accurate. This is the most common one and the hardest to admit, because the meeting compensates for it well enough that the underlying problem stays hidden until a deadline is missed. The fix is to make updating the board cheaper than talking about it, which usually means fewer tools rather than better discipline. Teams carrying a board in one tool, a schedule in a spreadsheet and the conversation in chat are paying a tax three times a day, and the honest comparison is between that and one place that covers all three.
Pick one of two changes and give it two weeks. If the board is not trustworthy, stop trying to shorten the meeting and fix the board first, because every other improvement depends on it. If the board is trustworthy, replace the round of personal updates with a walk through the board from done backward, keep the blocker question, and see whether the meeting ends in seven minutes instead of twenty. A board where the kanban view, the schedule and the conversation sit in one place makes that switch straightforward, which is the case that Pinateca is built around.
The 2020 Scrum Guide sets the Daily Scrum at 15 minutes regardless of team size, which functions as a ceiling rather than a target. Teams with a current board often finish in five to eight minutes, and that is a sign of health, not of a meeting being skipped. If it consistently runs past 15 minutes, the cause is usually that decisions are being made inside it instead of parked for a smaller follow-up.
No. The three questions were in earlier versions of the Scrum Guide and are not in the 2020 text, which states that the Developers may choose whatever structure they want as long as the event inspects progress toward the Sprint Goal and produces a plan for the next day. Walking the board card by card, from the column closest to done backward, covers the same ground with less recital and surfaces unowned work.
The event is defined as being for the people doing the work, with the Product Owner and Scrum Master taking part only when they are actively working on items themselves. A manager who attends as an observer tends to turn it into a report, because everyone starts addressing their update to the most senior person present. If a manager needs the information, a shared board and a weekly summary meet that need without changing the meeting.
Use a written post with a fixed deadline, plus one synchronous call a week on the day the week's plan is set. The written form only works if one named person reads all the posts each morning and starts the conversation they imply, otherwise it becomes a log nobody reads. Keep the blocker question in the written form, since that is the part where a delayed answer costs the most days.