Feature Prioritization (RICE/ICE)

Also known as: RICE FrameworkICE ScoringFeature ScoringPrioritization MatrixWSJFMoSCoW

RICE Score = (Reach × Impact × Confidence) ÷ Effort

💡The Concept

Feature prioritization is the discipline of deciding WHAT to build and in WHAT ORDER using a repeatable, data-driven framework instead of gut feeling or whoever shouts loudest. The RICE framework scores each feature on Reach (how many users), Impact (how much it moves the needle, 0.25-3x), Confidence (how sure you are, 0-100%), and Effort (person-months). RICE Score = (Reach × Impact × Confidence) ÷ Effort. The ICE variant uses Impact, Confidence, and Ease (inverse of effort). Teams using structured prioritization ship 50% fewer 'wasted' features.

⚠️The Trap

The biggest prioritization trap is the HiPPO problem — Highest Paid Person's Opinion wins. In organizations without a framework, 64% of features are prioritized by executive request rather than data. Another trap: overweighting 'Reach' and building for the majority while ignoring high-value power users. A feature used by 5% of users who generate 40% of revenue may score higher than a feature for 80% of users who are on free plans.

🎯The Action

Score every feature request with RICE before it enters your roadmap. Create a shared spreadsheet: Feature | Reach (users/quarter) | Impact (0.25-3x) | Confidence (%) | Effort (person-weeks) | RICE Score. Stack rank by score. Review the top 5 and bottom 5 — if any bottom-5 feature 'feels' wrong, challenge your scoring inputs. Commit to building only the top 3 RICE items per sprint.

⚡Pro Tips

Confidence is the most underrated input. A feature with 3x impact but 20% confidence (score: 0.6) should lose to a feature with 1x impact and 90% confidence (score: 0.9). Run a quick customer validation before committing to low-confidence features.

Add 'Revenue Impact' as a tiebreaker column. If two features have similar RICE scores, the one that drives measurable revenue wins. This prevents analysis paralysis.

Never let any single person score alone — use team-average scoring to reduce individual bias. Product, Engineering, and Sales should each contribute scores.

🚫Common Myths

✗Myth: “The framework will tell you exactly what to build”

✓Reality: RICE/ICE generates a ranked list, but it's a decision SUPPORT tool, not an oracle. The framework helps you articulate WHY you chose something — when the CEO asks 'why aren't we building X?', you can point to its score.

✗Myth: “Small features should be skipped because they have low Reach”

✓Reality: Small features with high Confidence and low Effort can have massive RICE scores. Bug fixes and UX polish often have extremely low effort (0.5) and high confidence (100%), making their RICE score competitive with big features.

🧪