Planning poker for Jira Cloud

Planning poker, inside Jira.

Your team votes privately, reveals together, and the agreed estimate is saved straight to the Jira issue. No extra tabs, no copying numbers across.

Coming soon to the Atlassian MarketplaceSee the features

How it works

How a session works

Four steps, all on one page in Jira.

  1. Step 1: Prepare

    Create a session and build its queue ahead of time with type-ahead search, a filter builder or JQL.

  2. Step 2: Vote

    Everyone sees the current issue in full and picks a card privately. Nobody sees the votes yet, including the manager.

  3. Step 3: Reveal

    The session manager reveals the cards together, with the average, median, agreement and vote distribution.

  4. Step 4: Save

    The manager picks the agreed card or types a value and saves it to the Jira field you chose. The queue moves on.

Features

Everything a refinement session needs

From building the queue to writing the number back to Jira.

  • Real-time sessions

    Run live planning poker inside Jira Cloud. Prepare upcoming sessions and revisit completed ones from the lobby.

  • Clear roles

    Session managers run the session, participants vote and observers follow along without voting.

  • Issue queue

    Add issues by search, filters or JQL. Reorder by drag and drop or from the keyboard.

  • Private voting

    Everyone sees the current issue in full. Votes stay hidden until reveal, even from the manager.

  • Reveal and statistics

    Average, median, agreement, range, vote distribution, consensus detection and a suggested card.

  • Save to Jira

    Pick a card or type a value, save it to the issue and move on to the next one automatically.

  • Card templates

    Fibonacci, Modified Fibonacci, T-shirt sizes and Powers of 2, or build your own deck.

  • Your estimation field

    Write to a number field, Original estimate or a text field. Set a default and override it per session.

Explore all features

Estimation Insights

See how your estimates held up

Compare each estimate with the actual effort, measured as time logged or cycle time between two statuses.

  • Accuracy and team pace (hours per point)
  • Under- and over-estimated issues
  • Rounds per issue and a calibration table
About Insights

How each issue appears in the report

Estimated: 5 points · Actual: 11h

Example format only.

Data and permissions

Your data stays in Atlassian

Pokerly is built on Atlassian Forge. It has no external servers and keeps its data in Atlassian's cloud.

Read the data and privacy statement
  • No external backend

    App data lives in Forge-hosted storage within your Atlassian environment. No external data egress.

  • Runs as the signed-in user

    Jira reads and writes use each person's own permissions, so people only see issues they're allowed to see.

  • Minimal permissions

    read:jira-work, write:jira-work, read:jira-user andstorage:app. No external permissions.

Bring planning poker into Jira

Estimate together, on the issue, and keep the number where your team already works.

Coming soon to the Atlassian MarketplaceRead the getting-started guide