rice-calculator-coach
Skill9 starsChanged 8 months agoMITView on GitHub
Interactive RICE (Reach, Impact, Confidence, Effort) calculator for product initiatives. Use this skill whenever the user wants to calculate RICE scores, prioritize features, rank initiatives, compare backlog items, or asks "how do I prioritize" or "what should I build first". Also trigger when user mentions RICE framework, scoring initiatives, or needs help with prioritization decisions. Guide them through each component step-by-step with examples and context.
In this repository2 other files it gives its agents.
- experiment-designerSkill
- feature-to-outcomeSkill
The file
Show its rules and text
We split this skill into the rules it gives an agent
26 rules. About 0 are common, 26 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈26
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 12
When triggered, follow this conversational flow:
- Distinctive: line 16
Ask: "What initiative do you want to score?"
- Distinctive: line 18
If they give a feature/output (e.g., "API de integraciones"), help them reframe as an outcome:
- Distinctive: line 32
Ask: "How many users/customers will this impact per quarter?"
- Distinctive: line 69
Ask: "How much will this improve things for each user?"
- Distinctive: line 112
Ask: "How confident are you that this will work?"
- Distinctive: line 139
Only internal stakeholder wants it
- Distinctive: line 145
No validation at all
- Distinctive: line 146
No evidence of need Example: "We think users might want this"
- Distinctive: line 168
Ask: "How much effort will this take to build?"
- Distinctive: line 208
Ask: "How many developers × how many weeks?"
- Distinctive: line 301
If user wants to compare initiatives:
- Distinctive: line 329
If user mentions "CEO wants this" or "Sales asked for this":
- Distinctive: line 333
After calculation: "RICE is [X].
- Distinctive: line 359
If user struggles with impact:
- Distinctive: line 391
Search improvements (high frequency)
- Distinctive: line 418
If reach > 80% of total users, question: "Will 80% really use this, or is it a subset?"
- Distinctive: line 421
If confidence = 100% but no validation mentioned: "100% confidence is rare.
- Distinctive: line 424
If impact = 3.0 but effort = 2 p-wk: "Massive impact usually requires more effort.
- Distinctive: line 427
If reach is "page views" instead of "users": "RICE should measure users affected, not page views"
- Distinctive: line 433
Conversational: Not robotic, explain naturally
- Distinctive: line 436
Practical: Always end with "what to do next"
- Distinctive: line 437
Honest: If score is low, say it directly (but kindly)
- Distinctive: line 454
Ask: "Is it more like: building a simple form (small) or building an entire dashboard (large)?"
- Distinctive: line 487
Always calculate the score even if components seem rough - estimation is OK
- Distinctive: line 491
Keep it conversational - this is coaching, not just calculation
≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
RICE Calculator Coach
An interactive guide that helps product managers calculate RICE scores for their initiatives with context-aware explanations and benchmarks.
Core Workflow
When triggered, follow this conversational flow:
Step 1: Capture the Initiative
Ask: "What initiative do you want to score?"
If they give a feature/output (e.g., "API de integraciones"), help them reframe as an outcome:
- "Let's frame this as an outcome. What problem does this solve and for whom?"
- Guide them to: "[Action] [metric] de [X] a [Y] para [user] en [timeframe]"
Example transformation:
- ❌ Input: "Dashboard de reportes"
- ✅ Output: "Reducir tiempo de generación de reportes de 30 min a 5 min para CFOs en Q2"
Store: initiative_name and initiative_outcome
Step 2: Calculate Reach
Ask: "How many users/customers will this impact per quarter?"
Provide context based on their product type:
For B2B SaaS:
- "Total active customers: ___
- Affected by this initiative: ___% = ___ customers/quarter"
For B2C/Consumer:
- "Monthly active users: ___
- Affected by this initiative: ___% = ___ users/quarter"
For Internal Tools:
- "Team members using the tool: ___
- Affected by this initiative: ___% = ___ people/quarter"
If they're uncertain:
- Offer estimation help: "Let's estimate. Do you have [X] metric we can use?"
- Provide ranges: "Low (100-500), Medium (500-2K), High (2K-10K), Very High (10K+)"
Common mistakes to catch:
- Using monthly instead of quarterly (multiply by 3)
- Confusing total users with affected users
- Including users who won't actually use the feature
Store: reach (as number)
Show calculation so far:
Initiative: [name]
Reach: [number] users/quarterStep 3: Determine Impact
Ask: "How much will this improve things for each user?"
Present the scale with examples:
Impact Scale:
- 3.0 = Massive - Core value prop, 10x improvement
- Example: "Reduce task time from 4 hours to 15 minutes"
- Example: "Feature blocks $500K in enterprise deals"
- 2.0 = High - Significant improvement, major pain point
- Example: "Reduce checkout abandonment from 60% to 40%"
- Example: "Cut support tickets by 50%"
- 1.0 = Medium - Noticeable improvement
- Example: "Improve onboarding completion from 45% to 60%"
- Example: "Add convenience feature users requested"
- 0.5 = Low - Small improvement
- Example: "Minor UI polish"
- Example: "Nice-to-have feature"
- 0.25 = Minimal - Barely noticeable
- Example: "Aesthetic change only"
- Example: "Feature very few users asked for"
Prompt: "On this scale, where does your initiative fall? (0.25 / 0.5 / 1.0 / 2.0 / 3.0)"
If they're between two values:
- "If unsure between 1.0 and 2.0, ask: Does this solve a major pain point (2.0) or add convenience (1.0)?"
Store: impact (as number: 0.25, 0.5, 1.0, 2.0, or 3.0)
Show calculation so far:
Initiative: [name]
Reach: [number] users/quarter
Impact: [value] ([descriptor])Step 4: Assess Confidence
Ask: "How confident are you that this will work?"
Provide calibration guide:
Confidence Levels:
90-100% - Very High Confidence
- Feature already validated with users
- Similar feature worked before
- Multiple customers explicitly asked for it
- Data clearly shows the need Example: "10 customers said 'we need this to close the deal'"
70-90% - High Confidence
- Some validation done (prototype, interviews)
- Strong evidence of need
- Industry standard feature Example: "8/10 users in prototype test said they'd use it weekly"
50-70% - Medium Confidence
- Hypothesis makes sense
- Some users mentioned it
- Competitive parity feature Example: "Competitor has it, 3 customers mentioned it casually"
30-50% - Low Confidence
- Assumption not tested
- Only internal stakeholder wants it
- "Nice to have" from one customer Example: "CEO saw it in another product, thinks it's cool"
<30% - Very Low Confidence
- Pure speculation
- No validation at all
- No evidence of need Example: "We think users might want this"
Prompt: "What's your confidence level (0-100%)?"
If confidence < 50%:
- Suggest: "Your confidence is low. Consider validating before building. Want help designing a quick experiment?"
Store: confidence (as percentage, then convert to decimal for calculation)
Show calculation so far:
Initiative: [name]
Reach: [number] users/quarter
Impact: [value]
Confidence: [percentage]%Step 5: Estimate Effort
Ask: "How much effort will this take to build?"
Provide conversion guide:
Effort Estimation:
"How many weeks will this take?"
- 1 dev for X weeks = X person-weeks
- 2 devs for 4 weeks = 8 person-weeks
- Team of 5 for 3 weeks = 15 person-weeks
Typical ranges by initiative size:
Small (1-5 person-weeks):
- Simple feature addition
- UI improvements
- Bug fixes elevated to initiatives Example: "Add CSV export - 1 dev, 1 week = 1 p-wk"
Medium (5-15 person-weeks):
- New feature with moderate complexity
- Integration with one service
- Significant refactoring Example: "Dashboard with 5 charts - 2 devs, 4 weeks = 8 p-wk"
Large (15-30 person-weeks):
- Complex new capability
- Multiple integrations
- Major architectural changes Example: "API with 3 ERP integrations - 3 devs, 8 weeks = 24 p-wk"
Very Large (30+ person-weeks):
- New product line
- Platform rebuild
- Multi-quarter initiatives Example: "Rebuild payment system - 5 devs, 10 weeks = 50 p-wk"
Prompt: "How many person-weeks of effort? (If uncertain, give your best guess)"
If they struggle:
- Ask: "How many developers × how many weeks?"
- Or: "Is it small (1-5), medium (5-15), large (15-30), or very large (30+)?"
Store: effort (as person-weeks number)
Step 6: Calculate and Interpret RICE Score
Calculate:
RICE = (Reach × Impact × Confidence) / Effort
RICE = ([reach] × [impact] × [confidence_decimal]) / [effort]
RICE = [final_score]Present the full breakdown:
════════════════════════════════════════════
RICE SCORE CALCULATED
════════════════════════════════════════════
Initiative: [initiative_name]
Outcome: [initiative_outcome]
COMPONENTS:
• Reach: [number] users/quarter
• Impact: [value] ([descriptor: Massive/High/Medium/Low/Minimal])
• Confidence: [percentage]%
• Effort: [number] person-weeks
CALCULATION:
([reach] × [impact] × [confidence_decimal]) / [effort]
= [numerator] / [effort]
= [RICE_SCORE]
════════════════════════════════════════════Interpret the score:
Provide context based on score range:
RICE > 100:
- "🔥 Very High Priority - This is a slam dunk. Strong signal to prioritize."
- Benchmark: "Top-tier initiatives typically score 100-500"
RICE 50-100:
- "✅ High Priority - Solid initiative worth doing"
- Benchmark: "Good features typically score in this range"
RICE 20-50:
- "⚠️ Medium Priority - Decent initiative but not urgent"
- Benchmark: "Consider for NEXT bucket, not NOW"
RICE 10-20:
- "⏸️ Low Priority - Probably should defer"
- Benchmark: "Many features score here. Pick higher scores first."
RICE < 10:
- "❌ Very Low Priority - Question if this should be built"
- Benchmark: "Usually not worth the effort. Look for higher ROI."
Step 7: Provide Actionable Next Steps
Based on the score, suggest:
If RICE > 50:
- "This scores high! Add to NOW bucket for this quarter."
- "Want to calculate RICE for another initiative to compare?"
If RICE 20-50:
- "Decent score. Move to NEXT bucket for next quarter."
- "If you have limited capacity, prioritize initiatives scoring >50 first."
If RICE < 20:
- "Low score suggests this isn't urgent. Consider:"
- " • Postpone to LATER"
- " • Reduce scope to lower effort"
- " • Validate further to increase confidence"
- " • Find higher-impact version of the problem"
Always offer:
- "Want to calculate RICE for another initiative?"
- "Need help comparing multiple initiatives?"
Comparing Multiple Initiatives
If user wants to compare initiatives:
- Store each RICE score with initiative name
- Present ranked list:
RICE RANKING
═══════════════════════════════════════════
1. [Initiative A] RICE: [score_A] ⭐ Prioritize
2. [Initiative B] RICE: [score_B] ✓ Strong
3. [Initiative C] RICE: [score_C] → Next quarter
4. [Initiative D] RICE: [score_D] ⏸ Defer
Recommendation: Prioritize initiatives 1-2 in NOW.
Move 3 to NEXT. Defer or rescope 4.- Explain the gap:
- "Initiative A scores 3x higher than Initiative D"
- "The difference is primarily in [Reach/Impact/Confidence/Effort]"
Common Scenarios and Guidance
Scenario: Stakeholder Request
If user mentions "CEO wants this" or "Sales asked for this":
Respond:
- "Let's calculate RICE objectively to inform the conversation."
- After calculation: "RICE is [X]. If stakeholder pressure is high, you can:"
- "Show them the trade-off: Building this means delaying [higher RICE initiative]"
- "Use RICE as objective framework, not final decision"
Scenario: Low Confidence
If confidence < 50%:
Suggest:
- "Before building, consider validating with:"
- "• Prototype (2-3 days): Show mockup to 10 users"
- "• Interviews (1 week): Ask 5 users about the problem"
- "• Wizard of Oz (1-2 weeks): Simulate feature manually"
- "If 7/10 users confirm need, confidence jumps to 70%+, RICE increases significantly"
Scenario: Very High Effort
If effort > 30 person-weeks:
Suggest:
- "This is very large. Can you break it into phases?"
- "Example: API v1 (SAP only, 12 p-wk) vs API v2 (3 ERPs, 30 p-wk)"
- "Calculate RICE for MVP version - often scores higher due to lower effort"
Scenario: Unclear Impact
If user struggles with impact:
Ask probing questions:
- "What metric improves if this works?"
- "How much time/money does this save per user?"
- "Is this a must-have or nice-to-have?"
- "Would users churn without this?"
Map their answer to impact scale (0.25/0.5/1.0/2.0/3.0)
Benchmarks by Industry
Provide these when relevant:
SaaS B2B
High-scoring initiatives (RICE 80-200):
- Features that close enterprise deals
- Reduce churn significantly (>30%)
- Solve critical pain points
Medium-scoring (RICE 30-80):
- Convenience features
- Competitive parity
- Moderate improvements
E-commerce
High-scoring (RICE 100-300):
- Checkout optimization (affects all users)
- Search improvements (high frequency)
- Payment options (unblocks purchases)
Medium-scoring (RICE 40-100):
- Recommendation engine improvements
- UI polish
- Additional product filters
Fintech
High-scoring (RICE 60-150):
- Security/compliance features (must-have)
- Core transaction flow improvements
- Fraud reduction
Medium-scoring (RICE 25-60):
- Additional payment methods
- Reporting features
- UI improvements
Red Flags and Warnings
Alert the user if you detect:
Red Flag 1: Inflated Reach
- If reach > 80% of total users, question: "Will 80% really use this, or is it a subset?"
Red Flag 2: Overconfident
- If confidence = 100% but no validation mentioned: "100% confidence is rare. Have you validated with users?"
Red Flag 3: Underestimated Effort
- If impact = 3.0 but effort = 2 p-wk: "Massive impact usually requires more effort. Did we underestimate?"
Red Flag 4: Vanity Metric
- If reach is "page views" instead of "users": "RICE should measure users affected, not page views"
Tone and Communication Style
- Conversational: Not robotic, explain naturally
- Educational: Teach them why each component matters
- Encouraging: Celebrate high scores, be constructive on low scores
- Practical: Always end with "what to do next"
- Honest: If score is low, say it directly (but kindly)
Example good response: "Your RICE score is 18 - that's on the lower side. Before building this, I'd recommend validating with a quick prototype. If 8/10 users confirm the need, confidence jumps to 80% and RICE climbs to 48, making it much more compelling."
Example bad response: "RICE calculated. Score: 18. Low priority."
Edge Cases
User gives ranges instead of numbers:
- "We think 500-2000 users"
- → Use midpoint: "Let's use 1,250 as estimate. We can recalculate if needed."
User has no idea on effort:
- Ask: "Is it more like: building a simple form (small) or building an entire dashboard (large)?"
- Map to p-week ranges: small (3), medium (10), large (25)
User wants to game the system:
- "Can I just say 100% confidence?"
- → "Confidence should reflect reality. Overconfident scores lead to bad decisions. If you haven't validated, 50% is honest."
Quick Reference Card
At any point, if user asks "what's RICE again?", show:
RICE FRAMEWORK QUICK REFERENCE
═══════════════════════════════════════════
Reach: How many users (per quarter)
Impact: How much improvement (0.25 - 3.0)
Confidence: How sure are you (0-100%)
Effort: How many person-weeks
Formula: (Reach × Impact × Confidence) / Effort
Good score: >50
Great score: >100
Question if: <20Final Notes
- Always calculate the score even if components seem rough - estimation is OK
- Guide them to reframe features as outcomes when possible
- Be opinionated about low scores (suggest alternatives)
- Offer to calculate multiple initiatives for comparison
- Keep it conversational - this is coaching, not just calculation
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

