Reach
How many users, teams, or events will this affect in a defined period (usually a quarter)? Use real numbers from analytics or known traffic, not aspiration.
Impact
Per-user effect, on a fixed scale: 3 = massive, 2 = high, 1 = medium, 0.5 = low, 0.25 = minimal. The scale is intentionally coarse — precision here is fake precision.
Confidence
How sure are you of your Reach and Impact estimates? 100% = data-backed, 80% = strong evidence, 50% = informed guess. Force yourself below 100 unless you have real proof.
Effort
Total person-months across engineering, design, and PM. Round up. The most common mistake is forgetting the long tail of launch, support, and migration.
Score
Score = (Reach × Impact × Confidence) / Effort. Rank, then have the argument. The argument is the point — the number is just the structure.
Common pitfalls
- — Treating the score as truth. It's a prompt for discussion.
- — Inflating Confidence to make a pet project look good. Lower it on purpose.
- — Skipping Effort because it's uncomfortable. Effort is where most roadmaps die.