RICE Prioritization Example: A Quarter's Product Backlog
Six competing ideas scored honestly, with a ranking that surprised the team and one deliberate override.
Six ideas competed for one quarter of a five-person team: a checkout redesign, an onboarding fix, enterprise SSO, a dashboard dark mode, a long-standing export bug, and an AI feature the CEO liked.
Two things happened when the numbers went in. The unglamorous onboarding fix beat everything by a factor of two, and the AI feature could not honestly score above 40% confidence — so it went to a two-week discovery instead of the roadmap.
Reach
- Onboarding fix — all ~1,200 new signups/quarter
- Checkout redesign — 4,000 buyers/quarter reach that screen
- Export bug — 310 users/quarter file or hit it
- Dark mode — ~2,000 active users would see it
- Enterprise SSO — 40 accounts (but two are renewals)
- AI summaries — unknown; assumed 1,500, no data
Impact
- Onboarding fix — 3 (massive): unblocks first value entirely
- Checkout redesign — 2 (high): removes the top funnel drop-off
- Export bug — 1 (medium): real pain, workaround exists
- Dark mode — 0.5 (low): delight, no behaviour change
- Enterprise SSO — 2 (high) for the accounts affected
- AI summaries — 2 claimed, evidence thin
Confidence
- Onboarding fix — 80%: drop-off data solid, impact inferred
- Checkout redesign — 90%: funnel data plus five user tests
- Export bug — 100%: fully understood
- Dark mode — 90%: certain effect, certainly small
- Enterprise SSO — 100%: contractual
- AI summaries — 40%: below the line → two-week discovery instead
Effort
- Onboarding fix — 0.5 person-months
- Checkout redesign — 2 person-months
- Export bug — 0.25 person-months
- Dark mode — 1 person-month
- Enterprise SSO — 3 person-months
- AI summaries — 4+ person-months (estimate unreliable)
Ranked Backlog
- 1. Onboarding fix — (1200×3×0.8)/0.5 = 5,760 → do first
- 2. Checkout redesign — (4000×2×0.9)/2 = 3,600
- 3. Export bug — (310×1×1.0)/0.25 = 1,240 — cheap, slot it in
- 4. Dark mode — (2000×0.5×0.9)/1 = 900 → next quarter
- 5. SSO — (40×2×1.0)/3 = 27 — OVERRIDE to #3: renewal contracts require it (reason recorded)
- AI summaries — not scored onto roadmap; 2-week discovery to earn a confidence number
What this canvas reveals
- The winner (onboarding fix) won on the divisor: near-total reach and massive impact over almost no effort. Small-effort ideas with real reach are chronically underrated.
- SSO scored last by arithmetic and shipped anyway — an explicit override, written down, because two renewal contracts required it. That is RICE working, not failing.
- The AI feature was not rejected; it was demoted to research. Below 50% confidence, the correct backlog is the discovery backlog.
Block by block
1Reach
- Onboarding fix — all ~1,200 new signups/quarter
- Checkout redesign — 4,000 buyers/quarter reach that screen
- Export bug — 310 users/quarter file or hit it
- Dark mode — ~2,000 active users would see it
- Enterprise SSO — 40 accounts (but two are renewals)
- AI summaries — unknown; assumed 1,500, no data
2Impact
- Onboarding fix — 3 (massive): unblocks first value entirely
- Checkout redesign — 2 (high): removes the top funnel drop-off
- Export bug — 1 (medium): real pain, workaround exists
- Dark mode — 0.5 (low): delight, no behaviour change
- Enterprise SSO — 2 (high) for the accounts affected
- AI summaries — 2 claimed, evidence thin
3Confidence
- Onboarding fix — 80%: drop-off data solid, impact inferred
- Checkout redesign — 90%: funnel data plus five user tests
- Export bug — 100%: fully understood
- Dark mode — 90%: certain effect, certainly small
- Enterprise SSO — 100%: contractual
- AI summaries — 40%: below the line → two-week discovery instead
4Effort
- Onboarding fix — 0.5 person-months
- Checkout redesign — 2 person-months
- Export bug — 0.25 person-months
- Dark mode — 1 person-month
- Enterprise SSO — 3 person-months
- AI summaries — 4+ person-months (estimate unreliable)
5Ranked Backlog
- 1. Onboarding fix — (1200×3×0.8)/0.5 = 5,760 → do first
- 2. Checkout redesign — (4000×2×0.9)/2 = 3,600
- 3. Export bug — (310×1×1.0)/0.25 = 1,240 — cheap, slot it in
- 4. Dark mode — (2000×0.5×0.9)/1 = 900 → next quarter
- 5. SSO — (40×2×1.0)/3 = 27 — OVERRIDE to #3: renewal contracts require it (reason recorded)
- AI summaries — not scored onto roadmap; 2-week discovery to earn a confidence number