Skip to content

Development Velocity

We measure development throughput in story points and track it through velocity and burndown. Because we can quantify it, we can spot bottlenecks and improve delivery predictability over time.

A story point measures the complexity and effort of a feature, bug, or requirement - not raw hours. We estimate in points because complexity is more stable and comparable across people than time estimates.

  • Points are assigned during refinement and confirmed at Sprint Planning, AI-assisted then human-decided (see Project Manager).
  • We use a Fibonacci-style scale (1, 2, 3, 5, 8, 13) - the gaps force a deliberate “is this really bigger?” conversation.
  • An item larger than ~13 points is split before it enters a sprint.
PointsRough meaning
1–2Trivial / well-understood, low risk
3–5Normal story, some unknowns
8Complex, multiple components or notable risk
13Large - split if possible before committing

Velocity is the sum of story points for items that reach “Done” in a sprint. Averaged over the last 3–6 sprints, it becomes our forecasting baseline for how much to commit to next.

xychart-beta
    title "Team velocity (points completed per sprint)"
    x-axis [S1, S2, S3, S4, S5, S6]
    y-axis "Story points" 0 --> 45
    bar [28, 31, 30, 35, 33, 38]

We watch the trend, not any single sprint: a steady or rising average with low variance signals a healthy, predictable team.

The burndown tracks remaining points against the ideal trajectory, day by day. It is the team’s early-warning system: a flat line mid-sprint means work is stuck.

xychart-beta
    title "Sprint burndown (remaining story points)"
    x-axis [D1, D2, D3, D4, D5, D6, D7, D8, D9, D10]
    y-axis "Remaining points" 0 --> 40
    line [40, 36, 32, 28, 24, 20, 16, 12, 8, 0]
    line [40, 38, 36, 31, 30, 22, 18, 11, 5, 0]

The straight line is the ideal burn; the second is a typical actual. Persistent gaps above the ideal line are discussed at the Daily Scrum.

MetricWhat it tells usHealthy signal
Velocity (3–6 sprint avg)Sustainable throughputStable or gently rising
Velocity variancePredictabilityLow and shrinking
Burndown shapeIn-sprint flowTracks near the ideal line
Points per developerLoad balance & blockersEven, no silent overload
Spillover (carried points)Over-commitment / blockersLow and trending down
Scope change mid-sprintGoal protectionRare; handled next sprint

Because velocity is measured, it is improvable. Typical levers we act on, decided at the Retrospective:

  • Smaller stories → smoother burndown and fewer carryovers.
  • Reducing spillover → more accurate commitment, less context-switching.
  • Removing recurring blockers (environments, reviews, dependencies) surfaced by points-per-developer and burndown stalls.
  • AI assistance in build and review compresses cycle time per point (see AI Tooling Map).

The goal is predictability first, speed second - a team that reliably hits its commitment is more valuable than one that is occasionally fast.