| name | feature-prioritization |
| description | Use when the user wants to decide what to build next, rank features, cut scope, or make trade-offs between competing requests. Also use when the user mentions 'what should we build,' 'prioritization,' 'RICE,' 'ICE,' 'backlog grooming,' 'MVP scope,' 'feature ranking,' 'what to cut,' or 'MoSCoW.' |
Feature Prioritization
Prioritization is not about ranking what the loudest voice asked for. It is about maximizing value delivered per unit of effort against a defined outcome.
RICE Scoring
Score each candidate feature on four dimensions:
| Dimension | Definition | Scale |
|---|
| Reach | How many users affected per quarter? | Raw number |
| Impact | How much does it move the target metric per user? | 3=massive / 2=high / 1=medium / 0.5=low / 0.25=minimal |
| Confidence | How confident in reach and impact estimates? | 100% / 80% / 50% |
| Effort | Person-months to build? | Raw number (lower = better) |
RICE score = (Reach × Impact × Confidence) / Effort
Higher score = higher priority.
ICE Scoring (Faster Variant)
For rapid prioritization without hard data:
| Dimension | Question | Score 1-10 |
|---|
| Impact | How much does this move the goal if it works? | — |
| Confidence | How confident are we it will work? | — |
| Ease | How easy is it to implement? | — |
ICE score = (Impact + Confidence + Ease) / 3
ICE is a gut-check tool. RICE is a data tool. Don't confuse them.
Kano Model
Classify features before scoring:
- Must-haves — Expected. No delight if present; strong dissatisfaction if absent.
- Performance — More = better. Direct correlation with satisfaction.
- Delighters — Not expected but create delight. No dissatisfaction if absent.
Rule: Ship all must-haves first. Then performance features. Then delighters.
MoSCoW for Scope-Cutting
| Category | Definition |
|---|
| Must have | MVP fails without this |
| Should have | High value, can launch without if pressed |
| Could have | Nice to have, first to cut |
| Won't have | Explicitly out of scope this cycle |
MoSCoW works best when "Won't have" is as long as "Must have."
Common Rationalizations
| Rationalization | Reality |
|---|
| "The customer asked for it" | Customer requests are signals, not specifications. Score them like everything else. |
| "This is a quick win" | "Quick" is effort, not impact. Score impact first — low-effort, low-impact work is still waste. |
| "We should do everything on the list" | Every yes is a no to something else. The list must have a cut line. |
| "RICE is too complex" | Show the scoring. Invisible prioritization produces invisible trade-offs. |
Verification
