Strategy and planning
RICE scoring
Reach × Impact × Confidence ÷ Effort — a comparable score that structures debate, not a verdict.
Intercom popularized RICE as a simple, comparable score:
RICE = (Reach × Impact × Confidence) / Effort| Factor | Meaning | Typical unit |
|---|---|---|
| Reach | How many people/events in a period | users per quarter, transactions/month |
| Impact | Effect on each person reached | 3 = massive, 2 = high, 1 = medium, 0.5 = low, 0.25 = minimal |
| Confidence | How sure you are of the estimates | 100% high, 80% medium, 50% low |
| Effort | Person-months (eng + design + PM) | 0.5, 1, 2, … |
Scores are relative, not absolute truth. Calibrate Impact and Effort as a team so "2" means the same thing to everyone. Use scores to structure debate, then apply judgment. Intercom's own guidance: frameworks inform decisions; they do not replace them.
Walking through five candidates
A B2B project-management product with ~40,000 monthly active users. One quarter of capacity (~3 person-months across eng for "one big bet" or several smaller ones).
| Idea | Reach (users/quarter) | Impact | Confidence | Effort (person-mo) | RICE |
|---|---|---|---|---|---|
| A. Onboarding checklist to first project | 8,000 new users | 2 | 80% | 1.0 | 12,800 |
| B. Slack notifications for task assign | 12,000 | 1 | 70% | 1.5 | 5,600 |
| C. Custom roles for enterprise | 400 | 3 | 90% | 3.0 | 360 |
| D. Dashboard performance (p95 4s → 1s) | 25,000 | 0.5 | 90% | 2.0 | 5,625 |
| E. Public roadmap / changelog page | 5,000 | 0.5 | 50% | 0.5 | 2,500 |
Arithmetic for A: (8000 × 2 × 0.8) / 1.0 = 12,800. For C: (400 × 3 × 0.9) / 3.0 = 360.
What the scores say: onboarding wins by a wide margin. Slack and performance are close; custom roles look terrible on RICE despite "strategic" enterprise pressure.
What the scores miss: C might unblock $2M of pipeline that finance cares about. D might be a guardrail — if p95 latency is killing retention, Impact of 0.5 was a lie. Dependencies matter: Slack might need notification infra that also unlocks email digests.
Example
Recalculate C if "reach" is accounts in negotiation, not MAU. Suppose 15 deals this quarter, each worth roughly as much as 500 MAU in LTV terms — you might score reach as 15 × a deal-impact scale instead of 400 users. RICE is only as good as the unit you choose. Pick the unit that matches how value actually shows up.
For the top 3–5 RICE winners that would consume a quarter, spend ~30 minutes each on a deeper pass (what Lenny's Newsletter contributors call DRICE): bottom-up impact estimate, task-level effort, explicit hypothesis. Lightweight ICE stays fine for a weekly experiment backlog; expensive bets deserve the deeper look.
Common mistakes
- Precision theater: debating Impact 1.7 vs 1.8 for three hours.
- Gaming Confidence to win the ranking.
- Wrong reach unit. Enterprise deals are not 400 MAU.
Check yourself
Onboarding RICE is 12,800; custom roles is 360. Ship roles because sales is loud? Not on RICE alone. If you override, write why (pipeline, regulation). Do not pretend the score said yes.
Why might dashboard perf's Impact of 0.5 be a lie? If p95 of 4s is killing retention, it is a guardrail, not a "minimal" nice-to-have.
Key takeaways
- RICE = Reach × Impact × Confidence ÷ Effort. Calibrate units as a team.
- Scores structure debate. They miss pipeline, guardrails and dependencies unless you choose honest units.
- Deeper-pass the expensive bets. Do not fake precision on the rest.
Further learning
RICE: Simple prioritization for product managers — IntercomPrioritization is a challenge for any product manager. How do you decide what to work on first? Here is a simple scoring system to help prioritize well.The Intercom Blog
Introducing DRICE — Lenny's NewsletterBoost your experiment win rate and help your team drive more impactlennysnewsletter.com
Next: ICE, Kano and WSJF.