Agile Estimating and Planningwriting

mike-cohnplanningstory-pointsestimationvelocity
2005-11-01 · 2 min read · Edit on Pyrite

mike-cohn's comprehensive treatment of estimation and planning in Agile contexts, published in 2005. The book codified story-points and velocity as planning tools, providing the quantitative infrastructure for Agile planning that user-stories-applied established qualitatively.

Story points and velocity

story-points are a unitless measure of the size and complexity of a user story, estimated relative to other stories rather than in hours. mike-cohn's book explains the cognitive basis for this: humans are better at relative estimation than absolute estimation. Estimating "this story is twice as big as that one" is more reliable than estimating "this story will take 14 hours."

velocity — the number of story points a team completes per iteration — is the empirical feedback mechanism. After a few iterations, a team's velocity stabilizes, providing a reliable basis for forecasting. This connects to empirical-process-control: instead of planning based on estimates made before work begins (which are unreliable), teams use measured velocity to forecast based on what they actually deliver.

The book covers:

  • Why traditional estimation fails and the case for relative sizing
  • Planning poker (planning-poker): the estimation technique using Fibonacci numbers
  • Release planning: using velocity and story count to forecast release dates
  • Iteration planning: selecting stories for each sprint
  • Tracking and communicating: burndown charts, burnup charts
  • Planning poker

    planning-poker — team members simultaneously reveal estimate cards to avoid anchoring bias — is described in this book and became a widely adopted practice. mike-cohn formalized a practice that had been used informally in XP teams. The Fibonacci sequence (1, 2, 3, 5, 8, 13, 21...) used for cards reflects the increasing uncertainty in larger estimates.

    Influence

    The book's treatment of story points and velocity became the de facto standard for Agile planning in scrum teams worldwide. Practices from this book are embedded in major Agile project tracking tools (Jira, Rally, VersionOne). The book also influenced release planning in safe-scaled-agile-framework and other scaling frameworks.

    The approach has critics: story points are often misused as productivity metrics, gaming velocity becomes a problem, and some practitioners (including kent-beck himself) have questioned whether story points add more value than simpler counting approaches.

    See also: user-stories-applied for the requirements companion to this planning book.