Most written standups die the same way. They start as three lines in a channel, they work for about six weeks, and then someone suggests a short weekly call to go over what people have been posting. Within a month you have the posts and the call, which is more overhead than the standup you replaced.
The format below is built to resist that. It is short on purpose, and most of the work is not in the format at all — it is in the three failure modes at the bottom, each of which arrives disguised as an improvement.
The format
Three lines, posted by each person, in a channel, once a day.
Yesterday: Finished the account-merge dry run. Found four legacy account types we had not scoped. Today: Writing the migration path for legacy type 1. Blocked: Need Priya’s call on whether types 3 and 4 ship at all. Asked Tuesday, no reply yet.
That is the whole artifact. No mood check, no confidence rating, no link to a board, no emoji key.
Two rules attached to it:
“Blocked” is either a name and an ask, or the word “no”. Not “waiting on design”. A person, a thing, and how long you have been waiting. Anything vaguer is a status colour, and status colours in a daily post are noise.
“Yesterday” is what changed, not what you did. Nobody needs your hours. They need to know what is now true that was not true yesterday, which is often a shorter list than the day felt like, and occasionally an empty one. “Yesterday: no change, still on the merge path” is a legitimate post and a team where it never appears is a team that has learned to perform.
The posting rule
One deadline, one channel, and the deadline is the part people skip.
Post before 11:00 in your own morning. Not “sometime in the day”. A daily artifact with no time attached becomes a trickle, and a trickle cannot be read — someone checking the channel at 09:30 sees a third of the team and has to come back twice more.
The rule that makes it survivable on a spread-out team is that the deadline is local. Everyone posts in their own morning; nobody sets an alarm for someone else’s timezone. What you lose is a single moment when the whole picture exists at once. What you get is a channel that is complete by the end of the earliest afternoon, which is enough for every decision a standup was ever actually used for.
Read once, at a time you choose. The counterpart rule, and the one that protects the format. If reading the channel means watching it, you have built an interruption rather than removed one.
Why three lines and not five
Every field you add is paid for daily by every person on the team, forever, and the cost compounds in a way that is hard to see because each individual field looks cheap.
More importantly: length determines whether it is read. A three-line post from eight people is one screen. Add a “confidence” field and a “risks” field and it is three screens, at which point nobody reads it, at which point somebody proposes a call to go over it. The format’s shortness is not an aesthetic preference. It is the mechanism.
If a field would only matter once a month, it does not belong in a daily post. It belongs in the weekly update, which has room for it and a reader who is expecting it.
The three ways it becomes a second standup
None of these arrive labelled. Each one arrives as a reasonable suggestion from a reasonable person.
1. The read-back call
Someone proposes fifteen minutes to “go through the posts together”, usually because they noticed the posts were not being read. It sounds like a small addition. It is the original meeting, restored, with a written pre-read nobody now needs to write carefully because it will be narrated anyway.
The tell: anyone reads their own post aloud, or a call opens with “so, what’s everyone working on”.
The repair: if the posts are not being read, that is the problem to fix, and a call does not fix it — it removes the incentive to fix it. Ask the specific question in the channel instead, quoting the post you are asking about. Three people watching a fourth answer a question they could have read is the exact cost the standup was cancelled over.
2. The thread that swallows the decision
Someone replies to a post with a question, the question is good, and eleven messages later a real decision has been made inside a thread on a Tuesday post that nobody will ever find again.
This is the most damaging of the three, because it is genuinely productive. Work happens. It becomes invisible while it happens, and three weeks later two people have different accounts of what was agreed and neither can prove theirs.
The tell: a thread on a standup post is longer than four replies, or contains the words “let’s go with”.
The repair: threads are fine for clarification and nothing else. The moment a thread produces a decision, the decision gets written down somewhere durable and the thread gets a final reply saying where it went. The standup channel is a log of state, not a record of decisions, and treating it as both is how you end up with neither.
3. The completeness ratchet
Fields accrete. Somebody adds “risks” after a surprise. Somebody adds a link to the board so the posts are traceable. Somebody adds a weekly summary post, then a monthly rollup, and one day a manager is producing a report from the standup channel and the posts start being written for that reader instead.
Once a daily post has a downstream reader with authority, it stops being a coordination artifact and becomes upward reporting, and people write it accordingly — carefully, defensively, and at length. That is the death of the format, and it happens without anyone deciding it.
The tell: you edit your standup post before publishing.
The repair: keep the audiences separate on purpose. The daily post is for the people doing the work. Upward reporting is a different artifact with a different cadence and a different reader — the one who has nine minutes a month for you — and it is written by one person, not extracted from eight.
The version that failed
Standup — Tuesday
Yesterday: Continued working on the migration piece, had a few conversations with the platform folks about the approach, and did some investigation into the account types issue. Also picked up two review requests and helped Dario debug the staging deploy. Today: More migration work, hopefully get the legacy path started. Also have the planning session at 2. Blockers: None really, just the usual dependencies. Might need some input from design at some point. Confidence: 🟡
Eighty words, and the only fact in it is the planning session.
“Continued working on” describes effort, not change — a reader learns nothing that was not true on Monday. “Might need some input from design at some point” is a blocker written so that nobody has to act on it, which means it will still be there on Friday and will surface as a surprise in the following week. The amber circle carries no information at all; it is a mood.
The post is not lazy. It took longer to write than the good one. It is written to look thorough to an imagined critical reader, which is the ratchet in action — this is what a standup post looks like once someone senior started reading the channel.
What it costs
Latency. A blocker posted at 09:00 gets an answer at 14:00, sometimes tomorrow. In a room the answer was thirty seconds away. That is a real loss and no format recovers it — you are trading response time for eight people’s uninterrupted mornings, and on some days that is a bad trade.
The unblocked-by-accident conversation is gone. In a standup, somebody occasionally mentions a thing in passing and someone else says wait, we solved that. Written posts produce less of this, because people report what they judge relevant and the incidental detail is the first thing cut. You lose some real value here and it is worth being honest that no rule recovers it.
Discipline is required indefinitely. The three failure modes above are not one-time hazards. They recur, each time proposed by someone acting in good faith, and someone has to keep saying no to reasonable suggestions. That job is unglamorous and it never ends.
When it should be a meeting
When more than about a third of the posts are blocked on each other, the team has a coordination problem that a written format is currently making worse. Meet, unpick it, go back to posting.
When the work is genuinely novel — a first week on an unfamiliar system, an incident, a design that has not settled — written standups are the wrong instrument. They are a good log of a known shape of work and a poor medium for figuring out the shape.
And if two people are the entire dependency graph, drop the format. Two people do not need a standup in any medium; they need each other’s attention twice a day, which they already have.