Standup Smith
Coming soon

Nine standup question sets that fit in a DM

If a question cannot be answered in two sentences, it does not belong in a daily standup. Use these sets in Standup Smith, or in any bot that asks three questions.

A daily standup has one job: surface stuck work before lunch. It is not a status novel, a therapy session, or a demo. Remote teams make this worse by adding icebreakers, emoji check-ins, and a fourth question "for fun." Fun belongs in a social channel. The standup DM should take about 2 minutes.

These sets are written for an engineering manager of 8 to 40 people. Each set is exactly three questions. That matches Standup Smith's cap of three questions per standup. If you run a bot that allows more, still stop at three. People pad.

1. The default shipping set

Use this when the team ships product every week.

  1. What did you merge or hand off since the last check-in?
  2. What will you merge or hand off before the next one?
  3. What is blocked, and who are you waiting on?

The verbs are merge and hand off, not "work on." "Work on" hides a day of spinning. If nothing merged, the honest answer is "nothing merged; still on the query plan." That is useful. "Making progress on the query plan" is not.

2. The incident week set

Swap this in when a severity-1 incident is open. On Business and Scale you can pause the default standup and run this one instead. Team has 3 standup slots, so you can keep the default and add this one without deleting anything.

  1. What did you do on the incident since the last check-in?
  2. What is the next concrete action, and who owns it?
  3. What do you need from another team in the next 4 hours?

Notice the 4-hour window. Daily is too slow in an incident. You may still run a 10-minute voice call. The written answers are the record, so people who log in at 22:00 have the same facts.

3. The on-call set

For the rotating on-call pair, not the whole squad.

  1. What pages did you take, and which are still open?
  2. What will you hand to the next on-call?
  3. What runbook was wrong or missing?

This set turns a broken runbook into a blocker with an owner's name on it, which is the point.

4. The platform / infra set

  1. What dependency did you change (IAM, queue, cluster, build)?
  2. What will you change next, and who else breaks if you do?
  3. What is waiting on another squad's migration, token, or approval?

Question 2 is the risk line. On Team and up, the AI digest lifts those answers into its list of risks.

5. The product-engineering pair set

When a PM sits in the same standup as the engineers, do not add a fourth question for them. Change the three.

  1. What customer-facing change shipped?
  2. What decision is still open, and whose yes is missing?
  3. What will we cut if the plan slips by two days?

The cut question stops the standup from becoming a wish list.

6. The contractor / vendor set

  1. What did you deliver against the ticket numbers in the last cycle?
  2. What will you deliver next, with ticket numbers?
  3. What are you waiting on from us (access, review, staging data)?

Ticket numbers keep a 15-person contractor standup from turning into prose.

7. The new-hire first two weeks set

  1. What did you set up or read?
  2. What will you try to ship, even if it is a one-line fix?
  3. Who do you need 20 minutes with, and did you book it?

You can run this as a second standup on Team (3 standups) for the two new people only. Do not make senior engineers answer these.

8. The Gulf / split-weekend set

If the squad spans a US week and a UAE or KSA week, the "yesterday / today" wording fails on Friday.

  1. What did you finish in your last working session?
  2. What will you finish in your next working session?
  3. What is blocked, and who (name, timezone) are you waiting on?

On Business and Scale, attach the UAE or KSA holiday calendar so the bot does not DM anyone on a public holiday. The wording still matters on a normal Friday.

9. The "we talk too much" recovery set

If people write paragraphs, shrink the allowed answer in the question itself.

  1. One merged thing (or "none").
  2. One next thing (or "none").
  3. One blocker with a @name (or "none").

You can add "one sentence" in parentheses. Standup Smith does not cut answers short. The wording of the question does that job. Managers who want more detail can reply under that person's answer.

Rules that beat a clever fourth question

Do not ask mood. Do not ask hours. Do not ask "any wins to celebrate." Those belong in a weekly retro or a kudos bot.

Do not ask people to paste Jira URLs as the whole answer. A URL with no sentence is how blockers hide.

Do not rotate questions every day. Rotate weekly at most. People stop thinking when the prompt is a moving target.

If more than 20 percent of answers are "none / none / none" for two weeks, the work lives in a different tool or the questions are wrong. Change the questions. The weekly email on Team and up shows a different number, the participation percent, which tells you who stopped answering at all.

Copy any set into Standup Smith or any other standup bot. The async standup playbook, which we email when you confirm, has twelve sets: these nine plus sets for support, data, and security teams.

Founding members

Founding members get 30% off regular prices for as long as they stay subscribed. 30-day money-back guarantee once the product opens. No card needed to join.

No public date yet. Founding members get the invite first, by email, before anyone else.