How to choose a planning poker tool: 10 questions to ask

The right planning poker tool is the one your team uses without thinking about it. At a minimum it should keep votes hidden until everyone reveals at once, offer the deck you estimate with, let people join in seconds and get the agreed estimate into your backlog without retyping. Beyond that, the best choice depends on how your team works: in a room or remote, live or async, in Jira or elsewhere.

This guide gives you 10 questions to ask of any tool, why each one matters and how to check it in a short trial. It does not rank products; it helps you judge them against your own team. If you are new to the technique, read what is planning poker? first.

The 10 questions at a glance

Start with how your team estimates

Before comparing tools, write down five facts about your team. They turn a long feature comparison into a short list of must-haves.

• Size: how many people vote, and how many watch?
• Location: all remote, all in one room, or hybrid?
• Rhythm: live sessions, async voting across time zones, or both?
• Scale: story points, T-shirt sizes or hours?
• Backlog: where do the items live, and where must the estimate end up?

Then mark each of the 10 questions as a must-have, a nice-to-have or irrelevant. A small team in one office estimating from a spreadsheet needs a very different tool from two teams sharing a Jira backlog across three time zones.

Questions about voting

1. Are votes really hidden until the reveal?

The private vote is the whole point of planning poker: it stops the first number anyone sees from anchoring everyone else. Check two things. On screen, people should see who has voted but not what they voted. Underneath, ask whether votes are sent to other people’s browsers before the reveal or only afterwards. A vote that is merely hidden on screen can be found by anyone curious enough to open their browser’s developer tools.

2. Does it have the decks and special cards you use?

Look for your scale (modified Fibonacci, T-shirt sizes, hours or whatever your team uses) and the special cards you rely on, especially ? for “not sure”. Check how the tool counts a ? in its statistics: it should register as a vote without dragging the average. If you need a custom scale, test that specifically. Planning poker cards and decks explained covers the options.

3. What does it show after the reveal?

Useful tools show the spread, not just an average: the highest and lowest votes, the most common vote, the median and a simple chart, so the facilitator can see at a glance whether to discuss or move on. Check who chooses the final estimate. A tool that records the average automatically encourages teams to skip the conversation that makes an estimate worth having.

Questions about running sessions

4. How quickly can people join?

Every minute spent on sign-ups and passwords at the start of a session is a minute of estimating lost, and a reason for a guest or contractor to give up. Time how long a new person takes to get from a link to a first vote, on a laptop and on a phone. Check whether guests need an account, whether anything has to be installed, and whether the tool works on your company’s network and browsers.

5. What can the facilitator control?

The facilitator, often called the moderator, needs to reveal the votes, start a re-vote, choose the final estimate and move to the next item. Look also for a round timer, a role for people who watch without voting, and a way to hand over control if the facilitator’s connection drops. Missing controls show up as awkward pauses in the middle of a session.

6. Does it suit remote, hybrid and async work?

For remote and hybrid teams the tool replaces the table, so it has to work for everyone at once and cope when someone’s connection drops and comes back. If your team spans time zones, check for an async mode: can people vote before a due date, add a comment explaining their card, and review the results together later? Async planning poker explains when that pays off, and remote planning poker covers the live version.

Questions about your backlog and your data

7. Does it connect to where your backlog lives?

Copying items into a tool and estimates back out is slow and invites mistakes. If your backlog is in Jira or another tracker, check whether the tool can pull items in and write the agreed estimate back to the right field. Ask whose permissions it uses to write: each person’s own, or a connection an admin set up, and whether your tracker’s history shows who saved each estimate. Test it on a real board before rolling it out; planning poker in Jira compares the ways to set this up.

8. Where is your data kept, and for how long?

Backlog items can contain customer names, unreleased plans and details of security fixes. Find out what the tool stores (names, item titles, descriptions, votes), where it stores it and for how long, especially for sessions started without an account. Check who can join with a link, and whether links can be guessed. Data protection rules such as the UK GDPR expect personal data to be kept no longer than it is needed, so a clear retention period and an easy way to delete a session are good signs.

Questions about access and cost

9. Can everyone on the team use it?

Check that the tool works with a keyboard alone, that colour is not the only way it shows who has voted, that text has enough contrast, and that a screen reader announces the reveal. The W3C’s Web Content Accessibility Guidelines (WCAG) 2.2 are the usual benchmark, so ask the supplier how the tool measures up. Better still, invite a colleague who uses assistive technology to join the trial.

10. What does it cost as you grow, and where are the limits?

Many tools are free to start, with limits that matter later: how many people can join, how long sessions are kept, and whether history and integrations are included. Work out when you would hit each limit, what the next step costs and how it is charged (per user, per team or per site), and whether you can export your estimates if you switch. A free tool you outgrow within a month can cost more in disruption than a paid one.

How to test a tool in 15 minutes

Feature lists only go so far. A short trial with two or three colleagues answers most of the 10 questions:

1. Create a session and join from a second device. Use a phone or a private browser window, as a guest. Time how long it takes.
2. Vote and look before the reveal. On the second device, check what you can see of the other person’s vote.
3. Play the awkward cards. Vote ?, re-vote, switch decks, and check how the statistics handle each one.
4. Connect your backlog. Import two real items from a test board, save an estimate and check it landed in the right field.
5. Read the data policy. Find the retention period and how to delete a session.
6. Run one real refinement session. Then ask the team one question: would you happily use this every week?

Run the same trial on each tool on your shortlist, with the same people, and the differences become obvious quickly.

On planningpoker.org The tool on this site is one option. It is free in the browser, and anyone can start a guest session without an account: it holds up to 8 people, the host included; the others join by link with just a name, and the session is deleted after 24 hours. Other people’s votes are not sent to anyone’s browser before the reveal. A free Laya account keeps sessions and adds Jira issues, Laya boards and Async Poker.