Establishing a Velocity Baseline for New Teams

Building a Velocity Baseline from Scratch

New teams face a chicken-and-egg problem: you need velocity to plan sprints, but you need completed sprints to have velocity. Here is how to bootstrap the process.

Step 1: Run a Calibration Sprint

Pick a modest set of well-understood stories. Do not over-commit. The goal is to establish a data point, not to impress anyone.

Sprint 0 (calibration):
  - Pulled 30 estimated points
  - Completed 22 points
  - Baseline velocity: 22

Step 2: Use a Reference Story

Before estimating anything, the team agrees on a reference story -- a task that everyone understands. Assign it a known point value (e.g., 3 points) and estimate everything else relative to it.

Step 3: Apply the "Yesterday's Weather" Rule

After sprint 1, plan sprint 2 with the same capacity. After sprint 2, use the average of sprints 1 and 2. By sprint 5, you have a meaningful baseline.

Sprint 1: 22 pts  →  Plan sprint 2 for ~22
Sprint 2: 26 pts  →  Average: 24, plan sprint 3 for ~24
Sprint 3: 24 pts  →  Average: 24, plan sprint 4 for ~24
Sprint 4: 28 pts  →  Average: 25, baseline is stabilizing
Sprint 5: 25 pts  →  Average: 25, ready for forecasting

Common Pitfalls

• Planning too ambitiously in sprint 1 -- this leads to carryover and discourages the team.
• Comparing to other teams -- each team has its own scale.
• Skipping retrospectives -- early sprints are learning opportunities; retrospectives help calibrate.

When Is the Baseline Ready?

When the standard deviation drops below 25% of the mean, your baseline is stable enough for forecasting.