Planning Poker Explained for Agile Estimation
Planning poker is a consensus-based technique agile teams use to estimate the size of backlog items: each team member privately selects a numbered card representing their size estimate, everyone reveals their card at the same time, and the team discusses any significant differences before converging on an agreed estimate.
Quick facts
- Cards typically follow a Fibonacci-like sequence (1, 2, 3, 5, 8, 13, 20, 40, 100) to discourage false precision at larger sizes.
- Simultaneous reveal prevents anchoring — no one's estimate is influenced by seeing others' numbers first.
- Estimates from planning poker are typically expressed in story points, not hours.
Planning poker vs simple team discussion
| Planning Poker | Open Discussion Estimation | |
|---|---|---|
| How estimates are formed | Simultaneous private reveal, then discussion of differences | Verbal estimates given openly, one at a time |
| Anchoring risk | Low — no one sees others' estimates before committing to their own | High — early estimates can bias later ones |
| Time required | Slightly longer due to structured rounds | Can be faster but less rigorous |
| Best for | Teams wanting more independent, unbiased input | Very small teams or simple items needing quick estimates |
How a planning poker session works
The team reviews a backlog item together, asking clarifying questions about scope and requirements. Each team member then privately selects a card representing their size estimate — physical cards, or more commonly today, a digital tool for remote teams. All cards are revealed simultaneously. If everyone's estimate is close, the team quickly converges on an agreed number. If there's a significant spread — say, some team members estimate a "3" while others estimate a "13" — this triggers a discussion specifically about why the estimates differ so much, often surfacing hidden complexity or misunderstanding about the item's scope.
Why simultaneous reveal matters
If team members share estimates one at a time out loud, the first number spoken tends to anchor everyone else's thinking, even unconsciously — a junior developer might defer to a senior colleague's estimate rather than voicing their own genuine assessment. Planning poker's simultaneous reveal removes this bias, surfacing each person's independent judgment before any discussion happens, which often reveals genuinely useful disagreement that a sequential discussion would have suppressed.
A worked example
A team is estimating a story about adding a new payment method. Four team members reveal cards showing 3, 3, 13, and 5. The wide spread prompts discussion: the developer who estimated 13 explains they're aware the payment provider's API requires a complex reconciliation step the others hadn't considered. This context changes the group's understanding of the story's real complexity, and after discussion, the team converges on an estimate of 8 — a more accurate number than either the low or high initial guesses, reached specifically because the disagreement was surfaced and discussed rather than resolved by simple averaging.
Common mistakes with planning poker
- Averaging estimates instead of discussing significant differences. The real value of planning poker comes from the discussion triggered by disagreement, not from a mathematical average of initial guesses.
- Letting a senior or vocal team member reveal or speak first, undermining the simultaneous-reveal structure meant to prevent anchoring.
- Using planning poker for every single backlog item regardless of size or clarity, adding unnecessary process overhead to items that are already well understood and quick to estimate informally.
- Treating the resulting number as a precise commitment rather than a relative size estimate carrying inherent uncertainty.
FAQ
What card values are typically used in planning poker? Most teams use a Fibonacci-like sequence (0, 1, 2, 3, 5, 8, 13, 20, 40, 100) plus a "?" card for genuine uncertainty and sometimes an infinity symbol for items too large or unclear to estimate at all.
Can planning poker be done remotely? Yes — many digital tools (like Planning Poker Online, or built-in features in tools like Jira) replicate the simultaneous-reveal mechanic for distributed teams effectively.
How long should a planning poker session take? This depends on the number of items being estimated, but most teams aim to keep individual item discussions brief (a few minutes each) once the technique becomes familiar to the team.
Is planning poker only for Scrum teams? No — while it originated in the Scrum and agile community, any team doing relative-size estimation for a shared backlog can use the technique, regardless of the specific methodology label they use.