| name | product |
| description | Write PRDs using Kevin Yien's Square template, prioritize roadmaps with RICE scoring, and define user stories with INVEST criteria—non-goals get equal attention because scope creep kills projects. Use when defining features, writing specs, prioritizing backlogs, or when "what to build next" needs structure. |
Product Management
Core Principle
A PRD that doesn't say "no" to anything says "yes" to everything. Scope discipline is the product manager's highest-leverage skill.
PRD Writing (Kevin Yien / Square Template)
PRD Structure
1. PROBLEM STATEMENT
What user pain are we addressing? Evidence?
2. GOALS
What does success look like? Measurable outcomes.
3. NON-GOALS (Critical)
What are we explicitly NOT doing? Why?
4. USER STORIES
As a [persona], I want [action], so that [outcome].
5. DESIGN & UX
Wireframes, flows, key interactions.
6. TECHNICAL APPROACH
High-level architecture, dependencies, risks.
7. METRICS
How we measure if this worked.
8. TIMELINE
Milestones with dates.
9. OPEN QUESTIONS
What we still need to figure out.
Non-Goals Section — Why It Matters
Non-goals aren't "things we'll do later."
They're "things we're choosing NOT to do, and here's why."
Example:
Goal: "Users can filter search results by category"
Non-goal: "Advanced boolean search queries"
Why: "80% of users use single-category filters. Boolean
search serves <5% of power users and triples complexity."
RICE Prioritization Framework
Score Each Feature
| Factor | Definition | Scale |
|---|
| Reach | How many users affected per quarter? | Actual number |
| Impact | How much does it move the needle per user? | 3=massive, 2=high, 1=medium, 0.5=low, 0.25=minimal |
| Confidence | How sure are we about R, I, and E? | 100%=high, 80%=medium, 50%=low |
| Effort | Person-months to ship | Actual estimate |
RICE Score
RICE = (Reach × Impact × Confidence) / Effort
Example A: Onboarding redesign
Reach: 5,000 users/quarter
Impact: 2 (high)
Confidence: 80%
Effort: 2 person-months
RICE = (5000 × 2 × 0.8) / 2 = 4,000
Example B: Dark mode
Reach: 1,000 users/quarter
Impact: 0.5 (low)
Confidence: 100%
Effort: 1 person-month
RICE = (1000 × 0.5 × 1.0) / 1 = 500
→ Onboarding redesign wins by 8x.
User Stories with INVEST Criteria
INVEST Checklist
| Criteria | Question | Pass/Fail |
|---|
| Independent | Can it be built without other stories? | |
| Negotiable | Is there room to adjust scope? | |
| Valuable | Does the user get clear value? | |
| Estimable | Can the team estimate effort? | |
| Small | Can it ship in one sprint? | |
| Testable | Can we write acceptance criteria? | |
Story Template
As a [specific persona],
I want to [concrete action],
so that [measurable outcome].
Acceptance criteria:
- Given [context], when [action], then [result]
- Given [edge case], when [action], then [handled gracefully]
Story Splitting Patterns
Too big: "As a user, I want to manage my account"
Split by operation:
→ "As a user, I want to update my email"
→ "As a user, I want to change my password"
→ "As a user, I want to delete my account"
Split by persona:
→ "As a free user, I want to see my usage limits"
→ "As a paid user, I want to manage my subscription"
Anti-Patterns
| Product Theater | Real Product Work |
|---|
| PRD with no non-goals section | Non-goals are as explicit as goals |
| Prioritizing by HiPPO (highest paid person's opinion) | RICE score with documented evidence |
| "MVP" that has 30 features | MVP that solves one problem completely |
| Stories without acceptance criteria | Every story has testable conditions |
| Roadmap set in stone for 12 months | Quarterly commitments, annual themes |
Power Move
"Write a PRD for [feature] using the Square template. Include 3 explicit non-goals with reasoning. Then RICE-score it against [alternative feature] and tell me which one to build first."
The agent becomes your product thinking partner — bringing rigor to what gets built and what gets cut.
Related Skills
| Skill | When to use it |
|---|
idea-validation | Before building — validate the problem exists and is worth solving |
pm-design-thinking | During feature definition — frame features through user outcomes |