Every product team has more ideas than time. Without a system, the winner is whoever argues loudest — usually the highest-paid person in the room. RICE replaces the shouting match with a number.
The formula
Score = (Reach × Impact × Confidence) ÷ Effort
- Reach — how many users does this touch per quarter? A real number from analytics, not a feeling
- Impact — per user: 3 = massive, 2 = high, 1 = medium, 0.5 = low, 0.25 = minimal
- Confidence — 100% = solid data, 80% = some evidence, 50% = educated guess. Below 50%? Stop scoring, go validate
- Effort — person-weeks across everyone: engineering, design, QA, you
Worked example
You PM a note-taking app. Three ideas compete for next quarter:
A — Dark mode. Reach 6,000 × Impact 0.5 × Confidence 0.8 ÷ Effort 3 weeks = 800
B — Offline sync. Reach 2,500 × Impact 2 × Confidence 0.8 ÷ Effort 8 weeks = 500
C — Fix onboarding drop-off. Reach 4,000 × Impact 2 × Confidence 1.0 ÷ Effort 2 weeks = 4,000
The onboarding fix wins by a mile — even though it’s the least glamorous idea and nobody argued for it in meetings. That’s the point: RICE surfaces quiet, high-leverage work that loud opinions overlook.
Where beginners go wrong
- Inflating numbers for pet ideas — tweak inputs until your favorite wins and you’ve built a shouting match with extra steps
- Treating scores as truth — RICE informs the decision, it doesn’t make it; a lower score can still win for strategic reasons, just write down why
- Scoring alone — score as a team; the debate about inputs is where real alignment happens
- Never re-scoring — confidence changes as you learn; stale scores are worse than none
Do this today
Take three ideas from your team’s backlog (or three improvements for any app you use) and score them. Surprised by the result? Good — that’s the framework earning its keep. A ready-made RICE scorecard with a decision log is in the PM Docs Toolkit.