Prioritization Scorer

Score and rank backlog items consistently using proven frameworks. Eliminates gut-feel prioritization and creates a defensible, shareable ranking for planning and stakeholder conversations.

When to Use

• "Help me prioritize these features"
• "What should we build next?"
• "Score this backlog"
• "RICE score for..."
• Before quarterly planning or sprint planning
• When debating between multiple initiatives
• Stakeholder asks "why are we doing X instead of Y?"

Frameworks

RICE (default for consumer/mobile)

Score = (Reach × Impact × Confidence) / Effort

ICE (fast and lightweight)

Score = Impact × Confidence × Ease

Weighted Scoring (custom criteria)

Define your own criteria and weights. Good for when you need to factor in strategic alignment, technical debt, or other org-specific concerns.

Process

1. Collect the Backlog

Ask for items to score. Accept:

• A pasted list of features or initiatives
• Reference to knowledge/products/*/roadmap.md
• A brain dump ("I'm thinking about X, Y, and Z")

2. Choose the Framework

Recommend based on context:

• RICE — best when you have rough reach data (analytics, user counts)
• ICE — best for quick gut-check ranking when data is sparse
• Weighted — best for cross-functional buy-in where multiple criteria matter

3. Score Each Item

Walk through each item and estimate each factor. For consumer/mobile:

• Reach: use MAU, DAU, or segment size from knowledge/products/*/metrics.md
• Impact: reference user research from memory/feedback/ or knowledge/customers/
• Confidence: factor in whether there's data, research, or just intuition
• Effort: break into eng/design/QA if possible

Flag items where confidence is below 50% — these may need an experiment before committing.

4. Rank and Discuss

Present the ranked list. Then:

• Identify the top 3 (your "must do" tier)
• Flag any scoring surprises (high-effort items that scored well, pet features that scored poorly)
• Call out clusters (multiple items serving the same goal — could they be combined?)

5. Sanity Check

Ask the user:

• Does this ranking feel right? If not, which item feels wrong and why?
• Are there strategic reasons to override the score? (document why)
• Any items that are prerequisites for higher-scored items?

6. Log the Prioritization

Write to memory/decisions/YYYY-MM-DD_prioritization.md so future planning can reference it.

Output

## Prioritization: [Context]

**Date:** YYYY-MM-DD
**Framework:** RICE / ICE / Weighted
**Items scored:** [count]

### Rankings

| Rank | Item | Reach | Impact | Confidence | Effort | Score |
|------|------|-------|--------|------------|--------|-------|
| 1 | [item] | [R] | [I] | [C] | [E] | [score] |

### Recommended Focus (Top 3)
1. **[Item]** — [why this matters]
2. **[Item]** — [why this matters]
3. **[Item]** — [why this matters]

### Needs Validation (Low Confidence)
- [Item] — confidence is [X]%, recommend: [experiment / research / spike]

### Overrides & Notes
- [Any strategic overrides with rationale]

### Dependencies
- [Item X] must ship before [Item Y]

Principles

• Consistent > precise — the value is relative ranking, not absolute numbers
• Low confidence = experiment, not debate — if you can't estimate impact, test it
• Document overrides — if you override the score, write down why (future you will thank you)
• Re-score quarterly — context changes, scores should too