Prioritization

PM Prioritization Framework

A comprehensive prioritization skill for product managers to evaluate features, initiatives, and tasks using proven frameworks.

Overview

This skill helps PMs make data-driven decisions by applying the right framework for their specific context. All outputs are provided in clean, tabular format for easy sharing with stakeholders.

When to Use This Skill

Use this skill when you need to:

• Prioritize a product roadmap
• Decide which features to build next
• Allocate team resources across initiatives
• Make tradeoff decisions between competing priorities
• Communicate priorities to stakeholders

---

Frameworks Included

1. RICE Scoring

When to use: When you have quantitative data and want an objective score

Best for: Feature prioritization, roadmap planning with multiple stakeholders

Requires: Estimates for reach, impact, confidence, and effort

Formula: (Reach × Impact × Confidence) / Effort

Components:

• Reach: How many users/customers affected in a time period
• Impact: Effect per person (Minimal=0.25, Low=0.5, Medium=1, High=2, Massive=3)
• Confidence: Estimate confidence (50%, 80%, 100%)
• Effort: Person-months required

Output Format:

| Feature | Reach | Impact | Confidence | Effort (PM) | RICE Score | Rank |
|---------|-------|--------|------------|-------------|------------|------|
| Feature A | 5000 | 2.0 | 100% | 3.0 | 3333 | 1 |
| Feature B | 3000 | 1.0 | 80% | 2.0 | 1200 | 2 |

---

2. Impact/Effort Matrix (2×2)

When to use: When you need quick visual prioritization or have limited data

Best for: Sprint planning, quick decisions, getting team alignment

Requires: Relative assessment of impact and effort (High/Low or Small/Medium/Large)

Quadrants:

• Quick Wins (High Impact, Low Effort) → Do First
• Big Bets (High Impact, High Effort) → Plan & Resource
• Fill-Ins (Low Impact, Low Effort) → Do When Available
• Money Pit (Low Impact, High Effort) → Avoid/Deprioritize

Output Format:

| Item | Impact | Effort | Quadrant | Priority | Recommendation |
|------|--------|--------|----------|----------|----------------|
| Feature A | High | Low | Quick Win | 1 | Ship this sprint |
| Feature B | High | High | Big Bet | 2 | Plan for Q2 |
| Feature C | Low | Low | Fill-In | 3 | If time permits |
| Feature D | Low | High | Money Pit | 4 | Defer indefinitely |

Visual Matrix:

           LOW EFFORT          HIGH EFFORT
         ┌─────────────────┬─────────────────┐
HIGH     │  QUICK WINS     │   BIG BETS      │
IMPACT   │  • Feature A    │   • Feature B   │
         │  • Feature C    │                 │
         ├─────────────────┼─────────────────┤
LOW      │  FILL-INS       │  MONEY PIT      │
IMPACT   │  • Feature D    │   • Feature E   │
         │                 │                 │
         └─────────────────┴─────────────────┘

---

3. Value vs. Complexity

When to use: When technical complexity is a key factor

Best for: Engineering-heavy decisions, technical debt prioritization

Requires: Business value assessment + technical complexity rating

Considers:

• Value: Business value, user value, strategic alignment
• Complexity: Technical complexity, dependencies, risk, maintenance burden

Output Format:

| Item | Business Value | Technical Complexity | Quadrant | Priority |
|------|----------------|---------------------|----------|----------|
| API Upgrade | High | Low | Quick Win | 1 |
| Platform Migration | High | High | Big Bet | 2 |
| UI Polish | Low | Low | Fill-In | 3 |
| Legacy Refactor | Low | High | Money Pit | 4 |

---

4. Weighted Scoring

When to use: When you have custom criteria important to your context

Best for: Strategic initiatives, complex B2B decisions, custom business contexts

Requires: Defined criteria and weights (must total 100%)

Common Criteria Examples:

• Strategic alignment (30%)
• Customer value (25%)
• Revenue potential (20%)
• Effort (15%)
• Risk (10%)

Output Format:

| Feature | Strategic (30%) | Customer (25%) | Revenue (20%) | Effort (15%) | Risk (10%) | Total Score | Rank |
|---------|-----------------|----------------|---------------|--------------|------------|-------------|------|
| Feature A | 9 (2.7) | 8 (2.0) | 7 (1.4) | 9 (1.35) | 8 (0.8) | 8.25 | 1 |
| Feature B | 7 (2.1) | 9 (2.25) | 6 (1.2) | 7 (1.05) | 7 (0.7) | 7.30 | 2 |

Note: Numbers in parentheses show weighted contribution

---

Framework Selection Guide

| Your Situation | Recommended Framework | Why | |----------------|----------------------|-----| | Have quantitative data | RICE | Most objective, data-driven | | Need quick decision | Impact/Effort Matrix | Visual, fast alignment | | High technical complexity | Value vs. Complexity | Accounts for engineering reality | | Custom business criteria | Weighted Scoring | Flexible to your context | | Early startup, high uncertainty | Impact/Effort Matrix | Works with limited data | | Enterprise, many stakeholders | RICE or Weighted Scoring | Defensible, systematic | | Managing technical debt | Value vs. Complexity | Balances business + technical |

---

How to Use This Skill

Input Format

Provide Claude with:

1. List of items to prioritize (features, initiatives, tasks)
2. Context (your product, stage, constraints)
3. Framework preference (or ask Claude to recommend)
4. Available data (estimates, metrics, constraints)

When You Don't Have Data

This is common and okay! Most PMs lack perfect data. Claude can help in two ways:

Option 1: Claude asks clarifying questions If you say "I don't have reach estimates," Claude will ask:

• How many customers have requested this?
• What % of your user base would use this feature?
• Is this a must-have for a specific segment?

Option 2: Claude provides baseline estimates Based on your product context, Claude can suggest reasonable ranges:

• "For B2B SaaS, SSO typically reaches 60-80% of enterprise customers"
• "Mobile adoption for desk-based products: 20-30%, field-based: 70-90%"
• "Search improvements usually have 'High' impact if search is used daily"

Claude will:

• ✅ Explain reasoning behind estimates
• ✅ Use industry benchmarks and patterns
• ✅ Flag high-uncertainty items for validation
• ✅ Give you ranges (e.g., "1000-2000 users") not false precision
• ✅ Mark estimates clearly so you know what to validate

Just tell Claude: "I don't have data for X, can you help estimate?"

Example Prompts

For RICE:

Using the prioritization skill, score these 5 features with RICE:

1. User dashboard redesign
2. Mobile app notifications  
3. API rate limiting
4. Advanced search filters
5. Export to CSV

Context: B2B SaaS, 10K users, 3-person eng team

For Impact/Effort:

Help me prioritize these 8 backlog items using Impact/Effort matrix:
[list items]

I need to decide what to tackle in next sprint.

For Framework Recommendation:

I have 12 items to prioritize. Not sure which framework to use.

Context: Early-stage startup, limited data, need to move fast,
high uncertainty. What framework should I use and why?

---

Claude's Output Structure

For every prioritization request, Claude will provide:

1. Recommended Framework (with reasoning)
2. Prioritized Table (clean, shareable format)
3. Key Insights (patterns, trade-offs, gaps)
4. Recommendation (what to do next)
5. Assumptions to Validate (what Claude assumed, what you should verify)

---

Output Requirements for Claude

When using this skill, Claude must:

✅ Always use markdown tables for prioritization results ✅ Show calculations for scoring frameworks (RICE, Weighted) ✅ Provide visual matrix for Impact/Effort and Value/Complexity ✅ Include priority ranking (1, 2, 3...) in output tables ✅ List assumptions made that should be validated ✅ Give specific next steps based on the prioritization ✅ Help estimate missing data - When the PM lacks estimates, Claude should:

• Ask 2-3 clarifying questions to gather context, OR
• Provide reasonable baseline estimates based on product type and industry patterns
• Explain reasoning (e.g., "Based on typical B2B SaaS, SSO reaches 70% of enterprise customers")
• Use ranges for uncertain estimates (e.g., "1000-2000 users" not "1500")
• Clearly mark which numbers are estimates vs. provided data
• Flag high-uncertainty items that need validation before decisions

❌ Never use verbose paragraphs for the main prioritization output ❌ Never skip the reasoning for framework selection ❌ Never provide scores without context or explaining the methodology ❌ Never refuse to help due to missing data - provide estimated ranges instead ❌ Never give false precision on estimates (e.g., don't say "exactly 1,247 users")

---

Tips for Better Results

Do:

• Provide context about your product and constraints
• Share what data you have available
• Mention upcoming deadlines or strategic goals
• Ask Claude to explain framework selection
• Request specific output format if you have preferences

Don't:

• Expect perfect precision with limited data
• Skip validating assumptions with your team
• Use complex frameworks when simple ones work
• Forget to communicate the "why" behind priorities
• Treat any framework as the final answer (they're decision aids)

---

Example: RICE Scoring Output

Input:

Use RICE to prioritize these 3 features for our project management tool:

1. Gantt chart view
2. Slack integration
3. Custom fields

We have 5K active teams, 4 engineers, targeting enterprise customers.

Claude Output:

Framework Recommendation

RICE Scoring - You have user base data (5K teams) and can estimate impact. This gives you defensible, quantitative prioritization for stakeholder discussions.

RICE Analysis

| Feature | Reach | Impact | Confidence | Effort (PM) | RICE Score | Rank | |---------|-------|--------|------------|-------------|------------|------| | Slack Integration | 4,000 | 2.0 | 100% | 2.0 | 4,000 | 1 | | Custom Fields | 3,500 | 1.0 | 80% | 2.8 | 1,000 | 2 | | Gantt Chart | 1,000 | 1.0 | 80% | 4.0 | 200 | 3 |

Calculation Details

Slack Integration:

• Reach: 4,000 teams (80% would use integrations based on enterprise SaaS benchmarks)
• Impact: High (2.0) - Reduces context switching, major workflow improvement
• Confidence: 100% - Clear demand signal from customer requests
• Effort: 2 person-months (standard OAuth + webhook implementation)
• Score: (4000 × 2.0 × 1.0) / 2.0 = 4,000

Custom Fields:

• Reach: 3,500 teams (70% of enterprise users need customization)
• Impact: Medium (1.0) - Enables flexibility but doesn't change core workflow
• Confidence: 80% - Some validation needed on field types required
• Effort: 2.8 person-months (schema changes + UI work)
• Score: (3500 × 1.0 × 0.8) / 2.8 = 1,000

Gantt Chart:

• Reach: 1,000 teams (20% - mainly traditional PM users)
• Impact: Medium (1.0) - Alternative timeline views exist
• Confidence: 80% - Nice-to-have feedback, not urgent requests
• Effort: 4 person-months (complex visualization + interactions)
• Score: (1000 × 1.0 × 0.8) / 4.0 = 200

Key Insights

• Integrations >> Views: Your customers value connecting tools over specialized views
• Effort/Value trade-off: Gantt chart takes 2× the effort of Slack integration for 1/4 the reach
• Enterprise focus: Both top priorities serve enterprise customization needs

Recommendation

Ship Slack integration first (Q1), then custom fields (Q2). Consider pausing Gantt chart until you see demand from a specific high-value segment.

Alternative to consider: After Slack, explore general API access - might serve both custom fields AND specialized view needs.

Assumptions to Validate

• [ ] Slack integration effort estimate - confirm with engineering
• [ ] 80% integration usage rate - check against similar features
• [ ] Custom field requirements - interview 5-10 enterprise customers
• [ ] Whether Gantt is must-have for specific verticals (construction, agencies)

---

Example: Impact/Effort Matrix Output

Input:

Quick prioritization for sprint. Impact/Effort matrix:

1. Fix login bug (affects 5% of users)
2. Add dark mode
3. Improve search performance  
4. Build admin analytics dashboard
5. Set up automated testing
6. Add export to PDF
7. Update documentation
8. Refactor payment code

Context: 2-week sprint, 3 developers

Claude Output:

Framework Recommendation

Impact/Effort Matrix - You need sprint-level decisions with a fixed team. This gives you quick visual prioritization without needing detailed estimates.

Prioritization Matrix

| Item | Impact | Effort | Quadrant | Priority | Sprint Plan | |------|--------|--------|----------|----------|-------------| | Fix login bug | High | Low | 🎯 Quick Win | 1 | Week 1 (2 days) | | Improve search | High | Medium | 🎯 Quick Win | 2 | Week 1-2 (4 days) | | Analytics dashboard | High | High | 📋 Big Bet | 3 | Consider for next sprint | | Automated testing | High | High | 📋 Big Bet | 4 | Consider for next sprint | | Update docs | Low | Low | ⏰ Fill-In | 5 | Squeeze in if time | | Export to PDF | Medium | Medium | ⏰ Fill-In | 6 | Backlog | | Dark mode | Low | Medium | ❌ Money Pit | 7 | Defer | | Refactor payment | Low | High | ❌ Money Pit | 8 | Defer (unless buggy) |

Visual Matrix

           LOW EFFORT              MEDIUM EFFORT           HIGH EFFORT
         ┌──────────────────┬──────────────────────┬─────────────────────┐
         │                  │                      │                     │
HIGH     │  🎯 QUICK WINS   │  🎯 QUICK WINS       │  📋 BIG BETS        │
IMPACT   │                  │                      │                     │
         │  • Login bug     │  • Search perf       │  • Analytics        │
         │                  │                      │  • Auto testing     │
         │                  │                      │                     │
         ├──────────────────┼──────────────────────┼─────────────────────┤
         │                  │                      │                     │
MEDIUM   │                  │  ⏰ FILL-INS         │                     │
IMPACT   │                  │                      │                     │
         │                  │  • Export PDF        │                     │
         │                  │                      │                     │
         ├──────────────────┼──────────────────────┼─────────────────────┤
         │                  │                      │                     │
LOW      │  ⏰ FILL-INS     │  ❌ MONEY PIT        │  ❌ MONEY PIT       │
IMPACT   │                  │                      │                     │
         │  • Update docs   │  • Dark mode         │  • Refactor payment │
         │                  │                      │                     │
         └──────────────────┴──────────────────────┴─────────────────────┘

Sprint Recommendation

Week 1:

• Days 1-2: Fix login bug (Priority 1)
• Days 3-5: Start search performance improvements

Week 2:

• Days 1-3: Complete search performance
• Days 4-5: Update documentation (Fill-In)

Defer to Next Sprint: Choose ONE big bet based on strategic priority:

• Analytics dashboard if you need revenue/usage insights NOW
• Automated testing if you need to improve long-term velocity

Key Insights

• You have 2 quick wins that fit in one sprint - do both
• Both big bets require full sprint commitment - don't split focus
• Dark mode is tempting but medium effort for low impact
• Payment refactor: only prioritize if it's causing bugs

Critical Question to Answer

Login bug affects only 5% of users - Are these high-value customers? If yes, this jumps to absol

…

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: sruthir28
• Source: sruthir28/enterprise-ai-skills
• License: MIT

Install and usage instructions live in the source repository linked above.