Comparing Velocity Across Different Sprint Lengths
How to normalize and compare velocity when switching between 1-week, 2-week, and 3-week sprints. Includes conversion formulas and trade-offs.
Detailed Explanation
Velocity and Sprint Length
If your team changes sprint length (e.g., from 2 weeks to 1 week), velocity needs to be normalized for comparison. The relationship is not simply linear.
Why It Is Not Linear
A 2-week sprint does not produce exactly 2x the velocity of a 1-week sprint because of:
- Sprint overhead -- planning, review, retro take the same time regardless of sprint length
- Context switching -- shorter sprints have proportionally more ceremony time
- Flow efficiency -- longer sprints allow deeper focus on complex stories
Typical Overhead by Sprint Length
Sprint Length Ceremony Time Effective Dev Time
1 week ~4 hours ~32 hours (80%)
2 weeks ~6 hours ~74 hours (93%)
3 weeks ~8 hours ~112 hours (93%)
4 weeks ~10 hours ~150 hours (94%)
Normalization Formula
To compare velocities across different sprint lengths:
Normalized velocity = (Velocity / Sprint days) x Reference days
Example:
2-week sprint velocity: 30 points
Normalized per week: 30 / 10 = 3.0 points/day
1-week sprint velocity: 12 points
Normalized per week: 12 / 5 = 2.4 points/day
The 2-week sprint produces 25% more per day because of lower overhead.
When to Change Sprint Length
| Consider shorter sprints | Consider longer sprints |
|---|---|
| Rapidly changing requirements | Complex, interconnected stories |
| Need faster feedback | High ceremony-to-work ratio |
| Small, independent stories | Deep research or exploration |
| Stakeholder wants frequent demos | Team is disrupted by frequent planning |
Practical Advice
When switching sprint lengths, discard historical velocity and re-establish a new baseline over 3-5 sprints. Do not try to mathematically convert old velocity -- too many factors change.
Use Case
Use this guide when debating whether to change sprint length, or when merging teams with different sprint cadences.