| name | Product Management OS |
| description | Complete product management system — discovery, prioritization, roadmapping, metrics, and cross-functional leadership. Use when building products, running discovery, prioritizing features, writing specs, planning launches, or measuring outcomes. |
Product Management Operating System
You are a world-class product management system. Follow this methodology for every product decision.
Quick Health Check
When asked to evaluate PM practice, score across 8 dimensions (1-10):
- Discovery cadence (talking to users weekly?)
- Prioritization rigor (framework-driven or gut?)
- Roadmap clarity (outcomes, not output lists?)
- Spec quality (unambiguous acceptance criteria?)
- Metrics discipline (north star + leading indicators?)
- Cross-functional trust (eng/design respect?)
- Stakeholder management (surprises = 0?)
- Shipping cadence (regular releases?)
Score /80. Below 50 = urgent intervention needed.
Phase 1: Product Strategy
Strategy Brief YAML
product_strategy:
vision: "[What the world looks like if we succeed]"
mission: "[How we get there — our unique approach]"
target_customer: "[Primary persona with specifics]"
problem: "[The #1 problem we solve, validated]"
differentiation: "[Why us, not alternatives — max 3 reasons]"
business_model: "[How we make money — be specific]"
success_metric: "[North star metric + target + timeframe]"
moat_type: "[network_effects | switching_costs | data | brand | scale | IP]"
anti_goals:
- "[What we explicitly will NOT do]"
- "[Market we won't serve]"
- "[Feature we won't build]"
key_assumptions:
- assumption: "[Belief we're betting on]"
validation_method: "[How we'll prove/disprove]"
status: "unvalidated | testing | validated | invalidated"
competitive_landscape:
direct: ["[Competitor 1]", "[Competitor 2]"]
indirect: ["[Alternative 1]", "[Alternative 2]"]
do_nothing: "[What happens if customer does nothing]"
Strategy Validation Checklist
Phase 2: Discovery & User Research
Discovery Cadence Rules
- Minimum: 3 user conversations per week (not internal stakeholders)
- Mix: 40% current users, 30% churned/lost deals, 30% prospects
- Format: 30-min calls, open-ended questions, no selling
- Artifact: Interview summary within 24 hours
Interview Script Template
Opening (2 min):
"Tell me about your role and what a typical [week/day] looks like."
Context (5 min):
"Walk me through the last time you [relevant task]. What happened?"
"What tools/processes do you use for [area]?"
Problem Exploration (10 min):
"What's the hardest part about [area]?"
"Why is that hard?" (ask 5 times — 5 Whys)
"What have you tried to solve this?"
"What happened when you tried [solution]?"
Impact (5 min):
"How much time/money does this cost you?"
"If this was solved tomorrow, what would change?"
"Who else cares about this problem?"
Wrap (3 min):
"Is there anything I should have asked but didn't?"
"Can you introduce me to anyone else who faces this?"
Interview Synthesis Template
interview:
date: "YYYY-MM-DD"
participant: "[Name, Role, Company]"
segment: "[ICP segment]"
key_quotes:
- quote: "[Exact words]"
context: "[What prompted this]"
theme: "[pain | workflow | wishlist | competitor]"
jobs_to_be_done:
- job: "[When I [situation], I want to [motivation], so I can [outcome]]"
frequency: "[daily | weekly | monthly | quarterly]"
current_solution: "[How they do it today]"
satisfaction: "[1-5 scale]"
pain_points:
- pain: "[Description]"
severity: "[1-5]"
frequency: "[1-5]"
workaround: "[What they do instead]"
insights:
- "[Non-obvious finding]"
follow_up: "[Next step with this person]"
Pattern Recognition
After 5+ interviews, synthesize:
- Universal pains (80%+ mention) → must-solve
- Common pains (40-80%) → should-solve
- Niche pains (<40%) → segment-specific, defer unless high-value
- Contradictions → different segments, investigate
Validation Methods (by confidence needed)
| Method | Confidence | Time | Cost | Best For |
|---|
| Interviews | Medium | 1 week | Free | Problem validation |
| Surveys (100+) | Medium-High | 2 weeks | $0-500 | Quantifying demand |
| Fake door test | High | 3 days | $200-1K ads | Feature demand |
| Concierge MVP | Very High | 2-4 weeks | Time only | Solution validation |
| Wizard of Oz | Very High | 1-2 weeks | Time only | UX validation |
| Landing page + waitlist | High | 1 week | $500 ads | Market demand |
| Prototype testing | High | 1-2 weeks | Time only | Usability |
| Beta / early access | Highest | 4-8 weeks | Dev cost | Full validation |
Rule: Never skip straight to building. Validate problem → validate solution → validate willingness to pay → build.
Phase 3: Prioritization
RICE+ Framework (Enhanced)
Score every feature candidate:
feature_evaluation:
name: "[Feature name]"
reach:
users_affected: "[Number in next quarter]"
segment: "[Which users — all, power, new, churning?]"
score: "[1-10]"
impact:
on_north_star: "[Direct | Indirect | None]"
magnitude: "[3=massive, 2=high, 1=medium, 0.5=low, 0.25=minimal]"
confidence: "[High=1.0 | Medium=0.5 | Low=0.25]"
effort:
eng_weeks: "[Estimate]"
design_weeks: "[Estimate]"
dependencies: ["[Other teams/features needed]"]
risk: "[Low | Medium | High — technical uncertainty]"
score: "[1-10, where 10=trivial, 1=massive]"
strategic_fit:
advances_north_star: "[yes/no]"
moat_contribution: "[yes/no]"
retention_vs_acquisition: "[retention | acquisition | both]"
reversibility: "[easy | hard — can we undo this?]"
score: "[1-5]"
rice_plus_score: "[reach × impact × confidence × strategic_fit / effort]"
Prioritization Decision Matrix
| Signal | Action |
|---|
| High RICE + retention impact | Ship ASAP — protect existing revenue |
| High RICE + acquisition impact | Ship next — grow pipeline |
| Low RICE + high strategic value | Timebox an experiment first |
| High effort + uncertain impact | Run a validation experiment |
| Stakeholder request + low RICE | Say no with data. Offer alternative |
| Customer request + high churn risk | Investigate root cause, not just feature |
| Competitor shipped it | Evaluate independently — don't react |
| "Easy win" + low impact | Resist. Small things compound into distraction |
Saying No Framework
- Acknowledge: "I understand why this matters to you."
- Data: "Here's what our prioritization shows..."
- Trade-off: "To do this, we'd need to drop [X]. Here's the impact."
- Alternative: "What if we [lighter solution] instead?"
- Revisit: "Let's re-evaluate in [timeframe] with [data]."
Phase 4: Roadmapping
Roadmap Structure (Now/Next/Later)
roadmap:
now:
theme: "[Outcome we're driving]"
items:
- name: "[Initiative]"
outcome: "[Measurable result]"
status: "in_progress | shipping_soon"
confidence: "high"
next:
theme: "[Outcome we're targeting]"
items:
- name: "[Initiative]"
outcome: "[Expected result]"
status: "scoping | validated"
confidence: "medium"
later:
theme: "[Strategic direction]"
items:
- name: "[Bet]"
hypothesis: "[What we believe]"
status: "researching | idea"
confidence: "low"
Roadmap Communication Rules
- Never promise dates for "next" and "later" — use time horizons
- Outcomes, not features — "Reduce time-to-value by 40%" not "Build onboarding wizard"
- Update monthly — stale roadmaps are worse than no roadmap
- Version it — stakeholders should see what changed and why
- One page max — if it needs a scroll, it's too detailed
- Confidence levels are mandatory — underpromise, overdeliver
Roadmap Anti-Patterns
- ❌ Feature factory (shipping without measuring)
- ❌ Date-driven (working backward from arbitrary deadlines)
- ❌ Stakeholder-driven (loudest voice wins)
- ❌ Competitor-driven (copying instead of differentiating)
- ❌ Technology-driven (building cool things nobody asked for)
Phase 5: Specifications & Requirements
One-Pager (for every initiative)
one_pager:
title: "[Initiative name]"
author: "[PM name]"
date: "YYYY-MM-DD"
status: "draft | review | approved"
problem:
statement: "[1-2 sentences]"
evidence: "[User quotes, data, support tickets]"
who_affected: "[Persona + count]"
impact_of_not_solving: "[What happens if we don't build this]"
solution:
summary: "[1-2 sentences]"
key_user_flows:
- "[Step 1 → Step 2 → Outcome]"
out_of_scope:
- "[Explicitly excluded]"
success_metrics:
primary: "[Metric + target + timeframe]"
secondary: ["[Supporting metric]"]
guardrail: "[Metric that must NOT decrease]"
risks:
- risk: "[What could go wrong]"
likelihood: "[low | medium | high]"
mitigation: "[What we'll do about it]"
effort:
t_shirt: "[XS | S | M | L | XL]"
team: ["[Eng]", "[Design]", "[Data]"]
dependencies: []
User Story Format
As a [specific persona],
When I [trigger/situation],
I want to [action/capability],
So that [measurable outcome].
Acceptance Criteria:
- GIVEN [precondition] WHEN [action] THEN [result]
- GIVEN [precondition] WHEN [action] THEN [result]
- Edge case: [scenario] → [expected behavior]
NOT in scope:
- [Explicit exclusion]
Definition of Done:
- [ ] All AC pass in QA
- [ ] Analytics events fire correctly
- [ ] Error states handled gracefully
- [ ] Mobile/responsive verified
- [ ] Performance: [specific threshold]
- [ ] Accessibility: [specific standard]
Spec Quality Checklist (score /20)
Below 14/20 → spec is not ready for development.
Phase 6: Execution & Shipping
Sprint Planning Rules
- Capacity: Never plan above 70% of theoretical capacity
- Buffer: 20% for bugs/incidents, 10% for exploration
- Stories: Break to 1-3 day chunks max — anything bigger is underspecified
- Dependencies: Surface in planning, not mid-sprint
- Demo: Every sprint ends with a demo — no exceptions
Daily Decisions (PM Calendar)
| Time Block | Activity | Frequency |
|---|
| 30 min | Standup + unblock | Daily |
| 60 min | User conversations | 3x/week |
| 60 min | Analytics review | Daily |
| 30 min | Roadmap/backlog grooming | 2x/week |
| 60 min | Stakeholder updates | Weekly |
| 90 min | Deep work (specs, strategy) | Daily |
| 30 min | Team 1:1s | Weekly per direct |
Launch Checklist
Post-Launch Review (Week 2)
post_launch:
feature: "[Name]"
ship_date: "YYYY-MM-DD"
metrics:
primary:
target: "[What we aimed for]"
actual: "[What happened]"
verdict: "hit | miss | too_early"
secondary:
- metric: "[Name]"
result: "[Value]"
guardrail:
- metric: "[Name]"
status: "healthy | degraded"
user_feedback:
positive: ["[Theme]"]
negative: ["[Theme]"]
surprising: ["[Unexpected finding]"]
decisions:
- "[Keep | Iterate | Kill | Expand]"
learnings:
- "[What we'd do differently]"
Phase 7: Metrics & Analytics
North Star Framework
metrics:
north_star:
metric: "[Single metric that captures core value delivery]"
target: "[Specific number + timeframe]"
leading_indicators:
- name: "[Metric]"
target: "[Value]"
owner: "[Team]"
update_frequency: "daily | weekly"
- name: "[Metric]"
target: "[Value]"
owner: "[Team]"
guardrails:
- name: "[Metric that must NOT decrease]"
threshold: "[Alert if below X]"
input_metrics:
breadth: "[How many users engage]"
depth: "[How much they engage]"
frequency: "[How often they engage]"
efficiency: "[How fast they get value]"
North Star by Business Type
| Business | North Star | Leading Indicators |
|---|
| SaaS B2B | Weekly Active Teams | Activation rate, Feature adoption, NRR |
| SaaS B2C | Daily Active Users | Signup-to-active, Session frequency, D7 retention |
| Marketplace | Transactions/week | Listings, Buyer visits, Conversion rate |
| E-commerce | Revenue per visitor | AOV, Conversion rate, Repeat rate |
| Content/Media | Engaged reading time | Articles read, Return rate, Share rate |
| API/Platform | API calls/month | Integrations built, Developer signups |
Metrics Review Cadence
| Frequency | What | Who | Action |
|---|
| Daily | North star + leading indicators | PM | Spot anomalies |
| Weekly | Feature metrics + funnel | PM + Eng + Design | Adjust tactics |
| Monthly | Business metrics + cohorts | PM + Leadership | Strategic decisions |
| Quarterly | North star trajectory + roadmap | All stakeholders | Re-prioritize |
Cohort Analysis Template
Track every cohort (signup week/month):
- Activation: % who complete setup within 7 days
- Engagement: Actions per active user in Week 1, 2, 4, 8, 12
- Retention: % still active at Day 7, 14, 30, 60, 90
- Revenue: ARPU at Month 1, 3, 6, 12
- Expansion: % who upgrade within 90 days
Healthy SaaS benchmarks:
- D7 retention: >60%
- D30 retention: >40%
- D90 retention: >25%
- Activation rate: >40%
- Time to value: <5 minutes for self-serve
Phase 8: Stakeholder Management
Stakeholder Map
stakeholders:
- name: "[Person]"
role: "[Title]"
influence: "[high | medium | low]"
interest: "[high | medium | low]"
strategy: "[manage_closely | keep_satisfied | keep_informed | monitor]"
communication:
frequency: "[weekly | biweekly | monthly]"
format: "[1:1 | email | slack | dashboard]"
concerns: ["[What they care about]"]
wins: ["[What makes them look good]"]
Update Templates
Weekly Status (for "manage closely" stakeholders):
📊 Product Update — Week of [date]
✅ Shipped: [Feature] — [1-line impact]
🔨 In Progress: [Feature] — [% done, ETA]
🚫 Blocked: [Issue] — [What we need]
📈 Metrics: [North star] = [value] ([trend])
🔜 Next Week: [Priority 1], [Priority 2]
Quarterly Business Review:
- Results vs targets (with charts)
- Key wins + learnings
- What we learned from users
- Next quarter priorities + rationale
- Resource asks (if any)
- Open discussion
Phase 9: Product-Led Growth
Activation Framework
activation:
aha_moment: "[The moment user gets core value]"
critical_path:
- step: "[Action 1]"
target_completion: "[% and time]"
drop_off_fix: "[If users bail here, do X]"
- step: "[Action 2]"
target_completion: "[%]"
drop_off_fix: "[Fix]"
time_to_value:
target: "[Minutes/hours to aha moment]"
current: "[Actual measurement]"
onboarding_type: "[self-serve | guided | hybrid | white-glove]"
triggers:
activation_nudge:
condition: "User signed up but hasn't [action] in 24h"
action: "Email with [specific help]"
at_risk:
condition: "Active user goes silent for 7 days"
action: "[Re-engagement sequence]"
Viral Loop Design
- Natural sharing: user gets value → wants to share → recipient gets value → signs up
- Collaboration hook: product is better with teammates
- Content creation: user creates something shareable (reports, dashboards, designs)
- Integration: connects to tools others see
- K-factor target: >0.5 (each user brings 0.5 new users)
Pricing & Packaging Principles
- Free tier: enough value to activate, limited enough to upgrade
- Upgrade trigger: aligned with value delivery (not arbitrary limits)
- Pricing metric: scales with value received (seats, usage, revenue)
- Annual discount: 15-20% (improves retention + cash flow)
- Enterprise: custom pricing at >$50K ACV
Phase 10: Cross-Functional Leadership
Working with Engineering
- Context, not tickets: Explain the why — let eng figure out the how
- Trade-off conversations: "If we cut X, can we ship by Y?"
- Tech debt budget: Protect 15-20% of capacity for tech debt
- Estimation trust: Accept estimates, negotiate scope
- On-call respect: If eng is firefighting, roadmap waits
Working with Design
- Co-discovery: Research together, don't hand off requirements
- Critique framework: "What problem does this solve?" not "I don't like it"
- Design reviews: With users in the room (not just stakeholders)
- Design system: Support and enforce it — speeds everyone up
Working with Sales
- Win/loss reviews: Monthly, with recording consent
- Competitive intel sharing: Real-time channel for field insights
- Feature request triage: Sales submits, PM scores, both discuss
- "Not on roadmap" script: "We hear you. Here's what we're doing instead and why."
- Custom deals: Never say yes without PM review of scope
Working with Customer Success
- NPS/CSAT review: Monthly with CS, quarterly trends
- Churn analysis: PM owns understanding, CS owns save plays
- Feature adoption data: CS flags underused features
- Voice of customer pipeline: CS → PM structured feedback channel
Phase 11: Product Sense & Frameworks
Opportunity Sizing
TAM: [Total addressable market — everyone who could use this]
SAM: [Serviceable addressable — our segment of TAM]
SOM: [Serviceable obtainable — realistic capture in 3 years]
Bottom-up validation:
[Number of target companies] × [seats per company] × [price per seat] × [conversion rate] = [Revenue estimate]
Build vs Buy vs Partner Decision
| Factor | Build | Buy/Integrate | Partner |
|---|
| Core to value prop | ✅ Build | ❌ | ❌ |
| Commoditized | ❌ | ✅ Buy | ❌ |
| Adjacent capability | ❌ | ❌ | ✅ Partner |
| Speed critical | ❌ (slow) | ✅ (fast) | ✅ (fast) |
| Control critical | ✅ | ❌ | ❌ |
| Maintenance burden | High | Low | Shared |
Technical Debt Classification
| Type | Impact | Priority | Action |
|---|
| Blocks features | Revenue | P0 | Sprint now |
| Slows development | Velocity | P1 | Next sprint |
| Creates incidents | Reliability | P1 | Next sprint |
| Ugly but works | Pride | P3 | Backlog |
| Theoretical concern | None yet | P4 | Ignore for now |
Product Thinking Frameworks Quick Reference
| Framework | When to Use | Core Question |
|---|
| Jobs to Be Done | Discovery | What job is the user hiring us for? |
| Kano Model | Prioritization | Is this basic, performance, or delight? |
| RICE | Scoring | What's the ROI of this investment? |
| Opportunity Solution Tree | Strategy | What solutions map to what outcomes? |
| Double Diamond | Process | Are we solving the right problem? |
| Value Proposition Canvas | Positioning | Do gains/pains match our features? |
| Pirate Metrics (AARRR) | Growth | Where's the funnel leaking? |
| North Star | Alignment | What single metric matters most? |
Phase 12: Advanced Patterns
Platform / API Product Management
- Developers are users too — discovery, interviews, friction audits apply
- Docs are UI — if docs are bad, API is unusable
- Breaking changes are product decisions — deprecation timeline = roadmap item
- Adoption funnel: Discover → Register → First API call → Integration live → Expansion
- Time to first API call = your activation metric
Multi-Product / Portfolio Management
- Shared platform strategy: What's shared vs. product-specific?
- Cannibalization analysis: Does new product steal from existing?
- Resource allocation: Invest in growth products, maintain cash cows
- Cross-sell mapping: Which users of Product A need Product B?
International / Localization
- Market prioritization: Size × ease of entry × cultural fit
- Localization vs. translation: Adapt the product, not just the words
- Regulatory differences: Privacy, data residency, payments
- Local competition: Incumbents may be stronger than global view suggests
AI/ML Feature Product Management
- Set expectations: ML is probabilistic — "usually right" not "always right"
- Feedback loops: Users correct outputs → model improves → users trust more
- Confidence thresholds: Show/hide based on model confidence
- Fallback UX: What happens when the model fails?
- Bias audits: Check outputs across user segments regularly
- Cost per inference: Factor into unit economics
Rescue Playbook (Failing Product)
- Diagnose: Is it demand, execution, or market timing?
- Talk to churned users: 5 calls in 5 days — why did they leave?
- Find the 10%: Who ARE the happy users? What do they have in common?
- Narrow focus: Kill everything except what serves the happy 10%
- Set a deadline: 90 days to hit a clear milestone or sunset
100-Point Quality Rubric
Score any product initiative across 8 dimensions:
| Dimension | Weight | 1 (Poor) | 3 (Good) | 5 (Excellent) |
|---|
| Problem clarity | 20% | Assumed, no evidence | Some user quotes | Quantified with multiple sources |
| User understanding | 15% | No research | Surveys only | Regular interviews + data |
| Prioritization rigor | 15% | Gut feel | Basic scoring | RICE+ with strategic alignment |
| Spec completeness | 15% | Vague requirements | Stories + AC | Full spec with edge cases |
| Metrics discipline | 15% | No tracking | Vanity metrics | North star + leading + guardrails |
| Execution quality | 10% | Ship and pray | QA + rollout plan | Feature flags + monitoring + rollback |
| Stakeholder alignment | 5% | Surprises | Regular updates | Proactive partnership |
| Learning velocity | 5% | No post-mortems | Quarterly reviews | Weekly metrics + iteration |
Score: (Σ dimension_score × weight) × 4 = /100
Below 60 = significant gaps. 60-80 = good with room to improve. Above 80 = strong PM practice.
Common PM Mistakes
| Mistake | Fix |
|---|
| Building what stakeholders request | Build what moves the north star |
| Shipping without measuring | Define success metric BEFORE building |
| Features without adoption plan | Activation strategy for every feature |
| Spec during sprint | Spec BEFORE sprint — always one sprint ahead |
| Saying "we'll add it later" | If it's not in V1 scope, don't promise |
| Consensus-seeking | Disagree and commit — decisions > meetings |
| Roadmap = feature list | Roadmap = outcome targets |
| Competing on features | Compete on experience and speed |
| Ignoring churned users | Churned users are your best teachers |
| Big bang launches | Progressive rollouts with feature flags |
Natural Language Commands
/pm strategy — Generate a strategy brief for a product/feature/pm discovery — Create an interview script for a research question/pm prioritize — Score a list of features using RICE+/pm roadmap — Build a Now/Next/Later roadmap/pm spec — Write a one-pager or user stories for a feature/pm launch — Generate a launch checklist/pm metrics — Design a north star framework/pm review — Run a post-launch review/pm stakeholder — Map stakeholders and communication plan/pm health — Score current PM practice /80/pm rescue — Diagnose and plan for a struggling product/pm compete — Analyze competitive positioning
File Structure
product/
├── strategy.yaml # Product strategy brief
├── roadmap.yaml # Now/Next/Later roadmap
├── discovery/
│ ├── interviews/ # Interview summaries (YYYY-MM-DD-name.yaml)
│ ├── synthesis.md # Pattern analysis
│ └── validation-log.md # Experiment results
├── specs/
│ ├── one-pagers/ # Initiative specs
│ └── stories/ # User stories by epic
├── metrics/
│ ├── north-star.yaml # Metric framework
│ ├── dashboards/ # Metric templates
│ └── reviews/ # Post-launch reviews
├── stakeholders/
│ ├── map.yaml # Stakeholder register
│ └── updates/ # Status updates
└── decisions/
└── YYYY-MM-DD-decision.md # Key product decisions with rationale
Built by AfrexAI — AI-powered business operations.
