Estimation field guide

Why Story Points Often Lead to Unproductive Discussions

When two developers choose different numbers, the disagreement can reveal missing context. The waste starts when the team debates the numbers without exploring their meaning.

Recognize a stalled discussion

“It is clearly a 3.” “It cannot be less than an 8.” If the next round repeats those positions, the team may be negotiating labels rather than learning about the story. A changed number alone is not evidence of better understanding: someone may simply have stopped arguing.

Useful disagreement produces new information: an overlooked test case, an external dependency or a different interpretation of the acceptance criteria. Listen for whether people can explain what changed between votes. If nothing changed except the social pressure to agree, pause the estimation.

Three reasons the same story gets different estimates

  • Different scope: one person includes migration, testing and review while another considers only the happy-path implementation. Write down what is included before comparing numbers.
  • Different knowledge: someone knows a reusable component exists; someone else remembers a difficult dependency. Share the evidence instead of assuming the other person is wrong.
  • Different definitions: “large” can mean difficult logic, lots of routine work or high uncertainty. Separate complexity, effort and risk to find the actual disagreement.

An early estimate from a senior colleague can also become an anchor. Independent initial estimates help, but the discussion afterward still needs space for people to explain a different view.

Replace another vote with a better question

  • “Which implementation, test or operational task is included in your estimate?”
  • “What do you know about this story that the rest of us may not know?”
  • “Is this difficult work, a lot of work, or work we do not understand yet?”
  • “What would we need to clarify before your assessment changes?”

Invite people with differing estimates to explain their reasoning without making them defend a number. Summarize the new information, update the story and only then decide whether another vote is useful.

Use the export story to test the approach

Illustrative example — not a customer report

For a CSV export, the lower estimate assumes a small dataset and existing permission checks. The higher estimate includes large exports and field-level access rules. Agreeing to 5 points hides that mismatch. Defining supported data volumes and naming the required access rules resolves part of it. Any remaining uncertainty can become an investigation or a separate story.

Finish with understanding, not forced consensus

Record the assumption that changed, the decision made and any unresolved question. If the story remains too uncertain, defer its estimate with an owner for the next action. Consonus can make rating differences visible across three dimensions, but the team still decides what those differences mean. A useful retrospective question is: could we explain our final estimate to someone who missed the meeting?

Ready for clearer estimates?

Bring one story to Consonus. Compare your team’s complexity, effort and risk ratings, discuss the different assumptions and agree on your next step.

Free during public preview. No payment details required.