Skip to main content

Estimation

Scraut supports async story point estimation using GitHub emoji reactions — no planning poker tool required.

How it works​

  1. A team member comments on an issue with: estimate this
  2. The Estimation Tally workflow fires
  3. The issue gets a voting comment with emoji options
  4. Team members react with the appropriate emoji
  5. After the voting period, the workflow tallies votes and applies the winning sp:N label

Triggering an estimation vote​

Option 1: Comment on the issue​

On any GitHub issue, add a comment:

estimate this

Option 2: Trigger from the Actions UI​

  1. Go to Actions → Scraut — Estimation Tally
  2. Run workflow with the issue number

The voting comment​

Scraut posts a comment to the issue:

🗳️ Story Point Estimation

Vote using reactions on this comment:

👍 = 1 sp    ❤️ = 2 sp    😄 = 3 sp
🎉 = 5 sp    🚀 = 8 sp    👀 = 13 sp

Votes close when the SM applies a label manually,
or trigger `estimation-tally` again to tally current votes.

Team members react to the comment — one reaction per person.

Tallying votes​

When the SM is satisfied with participation, trigger the tally:

  1. Go to Actions → Scraut — Estimation Tally
  2. Fill in the issue number
  3. Run workflow

Scraut counts the reactions, determines the winner (majority or median), and applies the sp:N label.

Example tally:

  • 2 votes for 😄 (3 sp)
  • 1 vote for 🎉 (5 sp)
  • Result: sp:3 label applied

Scenario: Team estimates a new story async​

Characters: Alice (dev), Bob (dev), Charlie (SM), and a new issue #52

Steps:

  1. Charlie creates issue #52: "As a user, I can export my data as CSV"
  2. Charlie comments: estimate this
  3. Scraut posts the voting comment with emoji options
  4. Alice sees the issue in her GitHub notifications, reacts with 😄 (3 sp)
  5. Bob checks the issue during his coffee break, reacts with 🎉 (5 sp)
  6. Charlie reacts with 😄 (3 sp)
  7. After everyone votes, Charlie triggers the tally
  8. 2 votes for 3 sp, 1 vote for 5 sp → sp:3 label applied
  9. Issue is ready for sprint planning

Total time: ~5 minutes async per person. Zero meetings.

Tips for good async estimation​

  • Vote independently — react before reading others' reactions to avoid anchoring
  • Discuss outliers — if votes are split 1 vs 8, drop a comment explaining your reasoning
  • Re-estimate after clarification — if the story changes significantly, trigger a new vote
  • Use the LLM — comment what complexity is this? and the bot will give its analysis before voting