Capability Prioritization

RICE = (Reach × Impact × Confidence) / Effort

• Reach (1–10): How many users/use cases affected

  • 10: All users, core workflow
  • 5: Most users, common workflow
  • 3: Some users, occasional workflow
  • 1: Few users, rare workflow

• Impact (0.5–3): How much it improves user experience/business

  • 3: Massive (core value proposition)
  • 2: High (significant improvement)
  • 1: Medium (nice to have)
  • 0.5: Low (minor improvement)

• Confidence (20%–100%): How confident in Reach/Impact estimates

  • 100%: Data-backed or validated
  • 80%: Reasonable assumptions
  • 50%: Educated guess
  • 20%: Pure speculation

• Effort (0.5–10): Person-weeks of work

  • 0.5: Trivial (<1 day)
  • 1: Small (1–2 days)
  • 3: Medium (1 week)
  • 5: Large (2–3 weeks)
  • 10: Very large (1+ month)

Framework 2: MoSCoW Classification

• MUST: Core flow cannot complete without it, revenue-critical, legal/compliance, blocks other MUSTs
• SHOULD: Significantly improves UX but workarounds exist, high demand but not universal
• COULD: Low effort high delight, small user segment, enhancement to existing capability
• WON'T: Out of scope, deferred to Phase 2, solved externally

Framework 3: Value vs. Complexity Quadrant

Value Score (0–10): User impact (0–4) + Business impact (0–4) + Strategic importance (0–2) Complexity Score (0–10): Technical complexity (0–4) + Dependencies (0–3) + Unknowns/Risk (0–3)

Quadrants:

• Quick Wins: Value >7, Complexity <4 → Do First
• Strategic Bets: Value >7, Complexity >7 → Plan Carefully
• Fill-ins: Value <4, Complexity <4 → Do Later
• Time Sinks: Value <4, Complexity >7 → Avoid/Defer

Framework 4: Kano Model

• Basic: Expected features; absence causes dissatisfaction (e.g., error messages, password reset)
• Performance: More is better; satisfaction increases linearly (e.g., faster load times)
• Excitement: Unexpected delighters (e.g., AI suggestions, personalization)
• Indifferent: User doesn't care either way
• Reverse: Some users prefer NOT having this

Priority order: Basic → Performance → Excitement → Indifferent/Reverse

Process

Step 1: Gather Capabilities

Collect all generated capabilities from the current spec.

Step 2: Apply RICE

For each capability, estimate Reach, Impact, Confidence, Effort. Calculate score. Rank.

Step 3: Apply MoSCoW

Classify each capability. Challenge every MUST — is it truly non-negotiable?

Step 4: Apply Value vs. Complexity

Score value and complexity. Assign quadrant. Flag Time Sinks.

Step 5: Apply Kano

Classify each capability by satisfaction type.

Step 6: Generate Matrix

Combine all four frameworks into a unified prioritization report.

Step 7: Generate Wave Plan

Group into MVP (Wave 1), Post-MVP (Wave 2), Future (Wave 3+). Estimate timeline.

Output Format

# Capability Prioritization Matrix
## Project: [Project Name]

## Priority Tier 1: MVP MUST-HAVES (Wave 1)

### CAP-NNN: [Title]
- **RICE**: [Score] (Rank #N)
- **MoSCoW**: [MUST|SHOULD|COULD]
- **Quadrant**: [Quick Win|Strategic Bet|Fill-in|Time Sink]
- **Kano**: [Basic|Performance|Excitement]
- **Recommendation**: [INCLUDE IN MVP | DEFER | AVOID]
- **Rationale**: [Why]

## Priority Tier 2: MVP SHOULD-HAVES (Wave 2)
[...]

## Priority Tier 3: POST-MVP (Wave 3+)
[...]

## Wave Planning

### Wave 1 (MVP - N weeks)
1. [Capability] (Week N)
[...]
**Total Effort**: N person-weeks

### Wave 2 (Post-MVP - N weeks)
[...]

## Key Insights
- **Quick Wins Identified**: [list]
- **Strategic Bets**: [list]
- **Time Sinks to Avoid**: [list]

## Confidence Levels
- High (100%): [capabilities]
- Medium (80%): [capabilities]
- Low (50%): [capabilities]

Quality Checks

• Every capability has all 4 framework scores
• RICE scores are justified, not arbitrary
• MUST-haves are truly non-negotiable (challenge each one)
• Time Sinks (low value, high effort) are flagged for reconsideration
• Wave 1 (MVP) is achievable in a reasonable timeframe
• Quick Wins are scheduled early for fast validation

Confidence Scoring

• 1.0: All inputs (reach, impact, effort) based on data or prior experience
• 0.9: Most inputs are reasonable estimates
• 0.8: Some inputs are educated guesses
• < 0.8: Too many unknowns — recommend discovery/estimation workshop

PTE-Specific Guidance

When prioritizing capabilities for the Precision Tone Engine:

• Enhancement workflow capabilities (tone rewrite, modal flow) are always MUST — core value prop
• Credit/subscription capabilities are revenue-critical — MUST for billing integrity
• Slack integration capabilities are MUST for MVP; Teams capabilities are SHOULD/COULD (parity in progress)
• Multi-language support is SHOULD for Basic+ tiers
• Analytics and custom workflows are COULD (Unlimited tier features)
• Consider Slack's 3-second trigger_id constraint when estimating complexity
• Account for the async Redis pub/sub pipeline when estimating effort for enhancement features
• Reference existing pricing tiers when scoring business impact