Planning Poker with Jira

Before you start

• A Jira Cloud account with access to the project you want to estimate.
• A Planner Poker account (free) — the Jira connection is stored per user.
• A refined backlog. Estimation is not the place to write acceptance criteria.

1. Connect your Atlassian account

Open your profile in Planner Poker and choose Connect with Jira. You are sent to Atlassian's consent screen, where you approve read access to your projects and issues via OAuth 2.0 (3LO). No password is ever entered into Planner Poker, and the access tokens are encrypted at rest. You can revoke the connection at any time from your profile or from your Atlassian account settings.

2. Create a room for the session

Give the room a recognisable name (for example "Sprint 42 refinement"), pick your scale — Fibonacci or T-shirt — and optionally set a voting timer to keep rounds tight. Share the room link or six-character code with the team; they join with just a display name, no account required.

3. Import the sprint into the task playlist

In the task playlist, choose Import from JIRA. Select the project and board, and the active sprint is loaded automatically — or switch to all issues if you are grooming the wider backlog. Tick the issues you want, import them, and they appear in the playlist as an ordered queue with their titles and descriptions.

Drag items to reorder before you begin. Estimating related stories back to back keeps the team's mental context loaded and produces more consistent numbers.

4. Estimate through the queue

Each item becomes the active task in turn. Everyone votes privately, cards reveal at once, and the average, minimum and maximum appear immediately. Discuss the outliers, agree a value, and move to the next item. The session keeps a full history, so the summary at the end lists every task, every participant's vote and the agreed estimate.

5. Write the estimates back to Jira

At the end of the session, use the agreed values from the summary to update the story points field on each issue. Because the summary lists the Jira issue keys alongside the agreed numbers, the update is mechanical — and it is worth doing the same day, while the reasoning behind each number is still fresh.

Practical tips for Jira-driven sessions

• Import at most 15–20 issues. Attention drops sharply after an hour, and tired teams inflate estimates.
• Filter out sub-tasks. Estimate at the story level; sub-tasks double count.
• Keep a reference issue in the queue. Re-estimating a known story at the start recalibrates the whole team.
• Park blocked items. If a story needs a decision from outside the room, skip it rather than guessing — a guess becomes a commitment in the sprint plan.
• Use the timer. Two minutes per round is enough for most refined stories and stops one issue from consuming the session.

Troubleshooting the connection

• "Jira session expired" — Atlassian refresh tokens rotate; reconnect from your profile in one click.
• No projects listed — your Atlassian user needs Browse Projects permission on the site you selected during consent.
• No active sprint — the board must be a Scrum board with a started sprint; otherwise switch the import to all issues.