Before the session
Bring a short list of tickets with enough detail to explain the outcome you want. Include acceptance criteria and known dependencies. Pick a small, familiar piece of completed work as a reference so the team has something concrete to compare against.
In Pointable, create a room, choose a deck and share the invite link. Add tickets individually, paste a list or import a CSV or TSV. If the backlog comes from Jira, export it first. Pointable does not connect to Jira or sync estimates back.
Choose an expiry that gives you enough time to finish and export your results. Add a room password if you need one, and share it within your team.
Run a voting round
- Read one ticket and explain what finishing it means.
- Let everyone ask questions before choosing a card.
- Vote privately, considering the effort, complexity and uncertainty relative to your reference ticket.
- Reveal together and ask people with different estimates to explain their assumptions.
- Clarify the scope, vote again if useful and record the estimate the team agrees.
Someone facilitating the meeting can join as a visitor if they only need to watch. Keep the focus on understanding the ticket rather than persuading everyone to pick the same number.
A worked story-point example
Imagine a ticket to add a filter to a results page. Your reference ticket was a simple text change that the team called one point. The filter needs interface work, a query change and tests.
The votes come back as 3, 3, 5 and 8. The person choosing 8 expects a new API endpoint. The person choosing 3 knows an existing endpoint already supports the filter. Checking that assumption is more useful than taking the average.
Once the team confirms the endpoint and agrees which tests belong in scope, it votes again. It might settle on 3 or 5. Neither is a universal answer for filters. The estimate belongs to this team, this scope and its current understanding.
Choose a deck your team understands
Pointable offers Fibonacci, modified Fibonacci, powers of two and T-shirt sizes. Fibonacci is a reasonable starting point for a team already using story points. T-shirt sizes can help with an early conversation about relative size.
Agree what the scale means before voting. A five-point ticket is not automatically five days of work. Avoid comparing point totals between teams with different reference work.
For background on numeric scales, see Mountain Goat Software’s explanation of Fibonacci estimation.
What to do when votes differ
Try asking: “What work are you including?”, “What feels uncertain?” and “What would make this smaller?” A question-mark vote is a useful signal to pause and clarify.
If the discussion uncovers an unanswered dependency, record the question or split out an investigation. You can return to the estimate later. Repeated voting cannot resolve missing information.
Pointable suggests the nearest card to the average for numeric decks after a reveal. Treat it as a starting point. The host can change the selection before agreeing the estimate.
Finish with a useful record
Before closing the meeting, check that agreed estimates are recorded and unresolved questions have an owner. Copy the results or export a CSV before the room expires, then update your backlog where it lives.
After your next session, ask whether the conversation surfaced surprises earlier. That gives you a more useful improvement to work towards than simply finishing the voting faster.
Create a planning poker room or try the Pointable demo before inviting your team.