Planning Poker in Jira: The Complete Guide for Scrum Teams (2026)

Planning Poker is the most widely adopted estimation technique in agile software development. It helps teams reach consensus on story point estimates through structured voting and discussion. But for teams that live in Jira, running Planning Poker in an external tool creates unnecessary friction.

This guide covers everything you need to know about running Planning Poker natively inside Jira — from setup and card set selection to advanced workflows for distributed teams.

What is Planning Poker?

Planning Poker (also called Scrum Poker) is an estimation technique where team members independently select a card representing their effort estimate for a user story or task. All cards are revealed simultaneously to prevent anchoring bias, and any significant disagreements are discussed before re-voting.

The technique was first described by James Grenning in 2002 and popularized by Mike Cohn in his book “Agile Estimating and Planning.”

Why Run Planning Poker Inside Jira?

Most agile teams already manage their backlog, sprints, and work tracking in Jira. Running estimation in a separate tool introduces several friction points:

• Context-switching: Moving between Jira and an external tool breaks focus
• Data duplication: Issue details must be copied or linked externally
• Manual data entry: Estimates must be transferred back to Jira fields
• Account management: Team members need separate logins for the estimation tool
• Onboarding overhead: New team members must learn an additional tool

A Jira-native Planning Poker solution eliminates all of these by embedding the estimation experience directly into the Jira workflow.

Setting Up Planning Poker in Jira

Step 1: Choose a Jira Planning Poker Plugin

The Atlassian Marketplace offers several Planning Poker plugins; our comparison of Planning Poker apps for Jira covers the main ones. Key features to evaluate:

Feature	Why It Matters
Multiple card sets	Different teams prefer different scales
Direct field writing	Saves estimates to Jira story point fields automatically
Refinement sessions	Estimate multiple issues in a structured session
Estimation method	Most voted, average, median or highest, so the room matches how the team decides
Real-time updates	Shows live voting status and card reveals
Issue preview	View issue details without leaving the estimation room

Agile Toolbox for Jira supports all of the above and adds retrospective boards in the same app, so one install covers both ceremonies. Every setting is described in the room configuration docs.

Step 2: Configure Your Card Set

Choose the card set that matches your team’s estimation style:

Fibonacci (most common): 0, 1, 2, 3, 5, 8, 13, 21, 34 The increasing gaps between larger numbers reflect the natural uncertainty in estimating larger work items. This is the default choice for most Scrum teams.

T-Shirt Sizes: XS, S, M, L, XL, XXL Best for teams that prefer relative sizing without numeric precision. Useful during early-stage backlog grooming before items are refined enough for numeric estimates.

Power of 2: 0, 1, 2, 4, 8, 16, 32, 64 A cleaner doubling scale that some teams prefer for its mathematical consistency.

Custom: Define your own values For teams with specific scales (e.g., half-day increments, ideal hours, or company-specific point systems).

Step 3: Configure the Target Field

Set up the Jira field where estimates will be saved. Most teams use the built-in Story Points field, but you can target any numeric custom field.

Step 4: Save Default Configurations

If your team uses different card sets for different issue types (e.g., Fibonacci for stories, T-shirts for epics), save default configurations per project and issue type. This eliminates repeated setup for recurring estimation patterns.

Running a Planning Poker Session

Preparing the Refinement

1. Create a new refinement session from the Refinements tab on your Jira project page
2. Add issues from the backlog or an upcoming sprint
3. The refinement appears on the board’s Refinements tab for everyone who can see the board, so there is nothing to send around. The refinements docs walk through the view.

During the Session

1. Open the first issue — the Planning Poker room displays the cards
2. Each team member selects their estimate privately
3. When all votes are in, reveal the cards
4. Review the results: average, median, and individual votes
5. If estimates are close, accept the consensus and save
6. If estimates diverge, discuss and re-vote
7. Move to the next issue

After the Session

The estimates are saved directly to the Jira story point field. No manual data entry needed. The refinement session stays accessible for reference.

Distributed Teams: Do the Reading Async, Keep the Vote Live

Agile Toolbox rooms are live: the team votes together and reveals together. For teams spanning several time zones, that does not have to mean a long meeting. Split the work instead:

1. The Scrum Master creates the refinement from the backlog a day or two ahead. It sits on the board’s Refinements tab, so everyone reads the issues in their own time.
2. Questions go into the Jira issue comments before the session, where they stay attached to the work.
3. The live session is short and only for voting: pick a window that overlaps for everyone, run the round timer, and skip the discussion on every round labelled Unanimous.
4. Talk only about the rounds labelled Split, and re-vote.

Two settings help here. The “?” card lets someone abstain without skewing the result, and “Allow offline voting” keeps a participant’s chosen card through a dropped connection so the round does not restart. There is a longer treatment of the trade-offs in Synchronous vs. asynchronous Planning Poker in Jira.

Best Practices

1. Prepare issues in advance. Don’t estimate issues cold. Create the refinement at least a day before the meeting so team members can review the issues.

2. Keep sessions focused. Estimate 8–12 issues per session maximum. Beyond that, estimation quality drops.

3. Discuss, don’t average. When estimates diverge, the conversation is more valuable than the number. Ask the outliers to explain their reasoning.

4. Use relative estimation. Start with a well-understood reference story (“this is a 3”) and estimate others relative to it.

5. Time-box each issue. 3–5 minutes per issue. If consensus doesn’t emerge, either break the story down or flag it for more research.

Getting Started

Install a Planning Poker plugin from the Atlassian Marketplace and try it with your next refinement. Most plugins offer free trials.

Agile Toolbox for Jira includes both Planning Poker and retrospective boards, so you can cover both ceremonies with a single plugin. Get Agile Toolbox for Jira — free forever for up to 10 users.

Related reading

• Demystifying story points: what the numbers mean and how velocity uses them.
• T-shirt sizing vs. story points: when a coarser scale is the better choice.
• Estimating an issue: the documentation for a single Planning Poker room.