# Planning Poker

URL: https://softwaredictionary.org/terms/planning-poker
Category: Teams & Process
Last updated: 2026-10-03
Pronunciation: PLAN-ing POH-ker

In short: Planning poker is a team estimation technique in which members privately pick estimate cards for a task, reveal them at once, then discuss the differences.

## What is planning poker?

The technique was described by James Grenning in 2002 and popularized by Mike Cohn's book on agile estimation. The team looks at a user story, asks the product owner questions, and then each person chooses a card, usually from a modified Fibonacci sequence such as 1, 2, 3, 5, 8, 13, 20 and 40, representing story points.

Everyone reveals at the same moment, which prevents anchoring, where the first number spoken pulls everyone else toward it. If the estimates are close, the team takes the common value. If they differ widely, the people with the highest and lowest cards explain their reasoning, and these conversations often uncover hidden work, risks or misunderstandings before anyone writes code.

The growing gaps between numbers reflect that big tasks are harder to estimate precisely, so there's no point arguing whether something is a 19 or a 21. Many teams treat a very high card as a signal to split the story. Remote teams use online planning poker tools that hide votes until everyone has chosen.

A common misconception is that the numbers are the main value. The estimates are rough and only meaningful within the same team; the real benefit is the shared understanding the discussion creates. Teams that can't agree after one short round usually need more information, not more voting.

## Key takeaways

- Planning poker is a team technique for estimating effort.
- Everyone chooses a card privately and reveals at the same time.
- Cards follow a growing sequence such as 1, 2, 3, 5, 8, 13.
- Large differences trigger discussion that exposes hidden work.
- The conversation matters more than the exact number.

## Frequently asked questions

**Why does planning poker use Fibonacci numbers?**

Because the growing gaps reflect growing uncertainty. It's realistic to tell a 2 from a 3, but not a 20 from a 21, so the scale forces estimates to stay rough for big tasks.

**What happens if estimates differ a lot?**

The people with the highest and lowest estimates explain their reasoning, the team discusses, and then everyone votes again. Big differences often reveal different assumptions about the work.

**Do you have to use planning poker in Scrum?**

No. Scrum doesn't prescribe any estimation technique. Planning poker is popular, but teams also use t-shirt sizes, simply count stories, or skip estimates and rely on small, similar-sized items.

---

Software Dictionary: https://softwaredictionary.org/ · https://softwaredictionary.org/llms.txt