
product-skills
16 product management skills for PMs and founders: user interviews, PRD writing, scope cutting, feature prioritization, positioning, strategy, metrics, and more
- GitHub stars
- 74
Install
/plugin install product-skills@productskillsRun it inside Claude Code after adding the marketplace with /plugin marketplace add assimovt/productskills.
Contents
Discovered from sourceSkills 16
Size product bets by separating reversible from irreversible decisions and shaping work to fit an appetite. Most product bets are Type 2 decisions — reversible, low-cost to try, high-cost to deliberate. Move fast on those. Save deliberation for Type 1 decisions that are hard to undo.
Type 1 vs Type 2 Decisions (Bezos)
Type 1 (Irreversible): One-way doors. Hard to undo once committed.
- Choosing a core technology/platform
- Pricing model changes that affect existing customers
- Killing a product line
- Public commitments (partnerships, integrations)
Type 2 (Reversible): Two-way doors. Easy to undo or iterate.
- Most new features (can ship, measure, remove)
- UI/UX changes (can A/B test or revert)
- Internal tooling decisions
- Most API additions (harder to remove, but additive is safer)
Rule: Use lightweight process for Type 2. Use deliberate process for Type 1. Most product teams over-process Type 2 decisions and under-process Type 1 decisions.
Shape Up Pitch Format
When proposing a bet, structure it as a Shape Up pitch:
1. Problem
A specific story showing real pain. Not an abstract need — a concrete situation with a real user.
"When a PM finishes a customer interview, they spend 45 minutes transcribing notes into a PRD. By the time they're done, the emotional context is gone and the PRD reads like a requirements list."
2. Appetite
How much time is this worth? Not how long it will take — how much you're willing to invest.
- Small bet: 1-2 weeks
- Medium bet: 3-4 weeks
- Large bet: 6 weeks (maximum for Shape Up)
If you can't fit the solution in the appetite, reshape or kill it.
3. Solution
Breadboard-level, not pixel-perfect. Show the key interactions and flows without getting into visual design. Fat-marker sketches, flow diagrams, or written walkthroughs.
4. Rabbit Holes
Known risks and unknowns that could blow up the timeline. For each: what's the risk and how will you mitigate it?
5. No-Gos
What's explicitly excluded. This is as important as what's included — it prevents scope creep during execution.
Expected Value Assessment
For larger bets, estimate expected value:
EV = (Upside x P(success)) - (Downside x P(failure)) + Learning Value
- Upside: Best-case outcome (metric improvement, revenue, users)
- P(success): Probability it works (be honest — most features have 30-50% success rate)
- Downside: Cost if it fails (time, opportunity cost, technical debt)
- Learning value: What you'll learn even if it fails. High learning value makes negative-EV bets worthwhile for early-stage products.
Guidelines
- CRITICAL: Classify every decision as Type 1 or Type 2 before deciding how much process to apply.
- NEVER spend 6 weeks deliberating a Type 2 decision. Ship it, measure it, adjust.
- NEVER rush a Type 1 decision because of artificial urgency. These are worth slowing down for.
- ALWAYS include No-Gos in a pitch. Without explicit exclusions, scope will grow.
- ALWAYS include Rabbit Holes. The risks you name are less dangerous than the ones you don't.
- NEVER pitch without a stated appetite. "Build this" without a time budget is an open invitation to over-engineer.
Built on Shape Up (Basecamp) and Jeff Bezos's Type 1/Type 2 decision framework. Skills from productskills.
Analyze competitors to find positioning gaps, not to copy features. The goal is to understand where alternatives fail your ICP so you can win on what matters to them.
Identify All Alternatives
List every alternative your ICP considers. Three categories:
- Direct competitors: Products that solve the same problem for the same audience.
- Indirect competitors: Products that solve the problem differently or serve an adjacent audience.
- Non-consumption: "Do nothing" and "manual workaround" (spreadsheets, email, hiring someone). ALWAYS include these — they're often the biggest competitor.
Feature Matrix
Build a comparison table focused on what YOUR ICP cares about, not every feature that exists.
| Capability | Your Product | Competitor A | Competitor B | Manual Workaround |
|---|---|---|---|---|
| [ICP-relevant capability] | How you do it | How they do it | How they do it | How people hack it |
Rate each: Strong / Adequate / Weak / Missing. Don't use checkmarks — they hide nuance. Include pricing model in the matrix — it constrains product decisions.
Example row: | Real-time collaboration | Strong (live cursors) | Weak (polling-based) | Missing | Google Docs + Slack — Adequate | So what: Their users are duct-taping two tools; our strength is their pain point.
Positioning Map
Plot competitors on two axes that matter to your ICP. Choose axes where you can credibly win at least one.
Common axis pairs:
- Ease of use vs Power/flexibility
- Speed to value vs Depth of solution
- Price vs Completeness
- Self-serve vs White-glove
The empty quadrant is your opportunity. If no quadrant is empty, you need a different framing.
Example (project management tools, ICP = small remote teams):
Powerful
|
Jira ------+------ Linear
|
Easy ----------------+---------------- Hard
|
Trello -----------+------ Asana
|
Simple
Empty quadrant: Powerful + Easy. That's the opportunity.
Strategic Gap Analysis
For each competitor, answer:
- Who do they serve best? (Their ICP vs yours)
- Where are they overserving? (Features their users don't need — signals bloat)
- Where are they underserving? (Complaints, missing features, workarounds their users build)
- What would make their users switch? (The trigger event + unmet need)
Guidelines
- CRITICAL: ALWAYS include "do nothing" and "manual workaround" as competitors. For most startups, inertia is the primary competitor.
- NEVER just list features. Analyze positioning, strategic choices, and gaps.
- ALWAYS focus on your ICP's needs, not the broadest possible comparison.
- NEVER copy competitor features without understanding why they built them and whether your ICP needs them.
- ALWAYS note competitor pricing and business model — these constrain their product decisions.
- NEVER present a feature matrix without strategic implications. "So what?" is the question every row should answer.
Skills from productskills.
Design experiments that actually prove something. Most A/B tests fail because they test vague ideas, run too short, or peek at results. A well-designed experiment has a clear hypothesis, adequate power, and a pre-committed analysis plan.
Hypothesis Template
Every experiment starts with a written hypothesis before any work begins:
"If we [make this specific change] for [this audience], then [this metric] will [change in this direction] by [this amount], because [this reason based on evidence]."
Example:
"If we replace the 5-step onboarding wizard with a single guided first-project flow for new signups, then 7-day activation rate will increase from 23% to 35%, because 4/6 interviewed users said they wanted to 'just start using it' not 'set everything up first.'"
Every part matters:
- Specific change: Not "improve onboarding" — the exact change
- Audience: Who sees this? New users only? Free tier only?
- Metric + direction + amount: A number you'll measure
- Because: The evidence-based reason. No evidence = no experiment.
Experiment Design
1. Primary Metric
One metric the experiment is designed to move. Not three. One. Additional metrics are guardrails.
2. Guardrail Metrics
Metrics that must NOT degrade. These prevent "winning" by breaking something else.
3. Sample Size
Calculate BEFORE running. Use a sample size calculator with:
- Baseline conversion rate (current number)
- Minimum detectable effect (smallest change worth caring about)
- Statistical significance (95% is standard)
- Power (80% minimum)
If you need 50,000 users and you get 500/week, the experiment will take 100 weeks. Either increase the MDE or don't run the experiment.
4. Duration
Run for at least one full business cycle (usually 1-2 weeks minimum) to capture day-of-week effects. NEVER run less than 7 days.
5. Analysis Plan
Write BEFORE launching: what metric, what threshold, what you'll do if it wins/loses/is inconclusive. Pre-commit to avoid post-hoc storytelling.
Common Mistakes
- Peeking: Checking results daily and stopping when you see significance. This inflates false positive rates. Set a duration and don't peek until it's done.
- Underpowered tests: Running a test with too few users to detect a meaningful effect. The result is "inconclusive" which teaches you nothing.
- Testing too many things: Changing 5 things at once means you don't know what worked. One change per experiment.
- No guardrails: Increasing signups by removing the password field "works" but breaks everything else.
- Post-hoc storytelling: The test was inconclusive but "we noticed an interesting trend in the 25-34 age group." That's noise, not signal.
Guidelines
- CRITICAL: Write the hypothesis and analysis plan BEFORE building anything. If you can't state what you're testing and how you'll decide, you're not ready.
- NEVER peek at results before the pre-determined end date. Statistical significance at day 3 of a 14-day test is meaningless.
- NEVER run an experiment without calculating required sample size first.
- ALWAYS include guardrail metrics. Winning the primary metric while tanking retention is not a win.
- NEVER test more than one major change per experiment. Isolate variables.
- ALWAYS document learnings from inconclusive tests. "We couldn't detect an effect with N=5,000 users" is still useful information — it suggests the effect is small or nonexistent.
Built on controlled experimentation methodology (Kohavi, Tang, Xu). Skills from productskills.
Prioritize with math, not opinions. RICE scoring forces explicit tradeoffs. The enabler/blocker lens from Linear ensures you're not just building fun things while adoption barriers remain.
RICE Scoring
Score every candidate feature on four dimensions:
- Reach: How many users/accounts will this affect in a set time period? Use real numbers from analytics, not gut feel. "500 users/quarter" not "a lot."
- Impact: How much will this move the target metric per user? Score 0.25 (minimal), 0.5 (low), 1 (medium), 2 (high), 3 (massive). Be honest — most features are a 1.
- Confidence: How sure are you about Reach and Impact? 100% = hard data. 80% = strong evidence. 50% = gut feel. NEVER score 100% without quantitative data.
- Effort: Person-weeks of work. Include design, engineering, QA, and any cross-team coordination. Round up.
RICE = (Reach x Impact x Confidence) / Effort
Example: SSO — Reach: 500 users/qtr, Impact: 2 (high — unlocks enterprise deals), Confidence: 80%, Effort: 4 person-weeks. RICE = (500 x 2 x 0.8) / 4 = 200. Tag: Blocker.
Rank by score. The math won't be perfect, but it forces you to justify each dimension.
Enablers vs Blockers (Linear)
After RICE scoring, classify each feature:
- Blocker: Removes a barrier to adoption or retention. Users are churning, stuck, or can't even start because this is missing. Examples: missing SSO for enterprise deals, broken mobile experience, no data export.
- Enabler: Delights existing users or deepens engagement. Users are already successful but this makes them more so. Examples: keyboard shortcuts, advanced filters, integrations.
Rule: Prioritize blockers over enablers when growing. Removing friction > adding delight when you're trying to grow. Flip this when retention is strong but engagement is flat.
Running a Prioritization Session
- List all candidates with a one-sentence description
- Score each on R, I, C, E independently — don't anchor on each other
- Calculate RICE scores and rank
- Tag each as Blocker or Enabler
- Check: are any Blockers ranked below Enablers? Justify or re-rank.
- Top 3-5 items = your next cycle
Guidelines
- ALWAYS show the math. A prioritization without visible scores is just an opinion list.
- NEVER let Confidence = 100% without hard data (analytics, A/B test results, or customer commit).
- ALWAYS separate enablers from blockers before finalizing the rank.
- NEVER prioritize more than one cycle at a time. The world changes. Re-score next cycle.
- CRITICAL: If two items are within 20% RICE score of each other, they're effectively tied. Use the blocker/enabler lens to break ties, or pick the one with higher Confidence.
Built on Intercom's RICE framework and the Linear Method. Skills from productskills.
Understand why customers "hire" and "fire" products. JTBD (Christensen/Moesta) reframes competition: you're not competing against similar products, you're competing against whatever the customer currently uses to get the job done — including doing nothing.
Job Statement Format
Structure every job as: "When [situation], I want to [motivation], so I can [expected outcome]."
- "When I finish a customer call, I want to capture key insights quickly, so I can share them with my team before I forget the context."
- NOT: "Users want better note-taking." (Too vague, no situation, no outcome.)
The situation is critical — the same person has different jobs in different contexts.
Forces of Progress
Four forces determine whether someone switches to your product. Map all four:
DRIVING ADOPTION RESISTING ADOPTION
================ ==================
Push: Pain with current solution Anxiety: Fear of the new solution
Pull: Attraction of new solution Habit: Comfort with current way
Push (away from current solution)
What's broken or frustrating about how they do it today? Stronger push = more urgency.
- "I spend 3 hours every week manually copying data between tools."
Pull (toward your solution)
What's attractive about the new way? This is your value proposition.
- "Automatic sync means I never copy data again."
Anxiety (about the new solution)
What are they afraid of? Switching costs, learning curves, data migration, reliability.
- "What if the sync breaks and I lose data?"
Habit (of the current solution)
What keeps them doing it the old way? Muscle memory, sunk costs, "good enough."
- "I've built all my templates in the current tool over 2 years."
To win: Push + Pull must be stronger than Anxiety + Habit.
If anxiety is high, reduce it (free trials, migration tools, guarantees). If habit is strong, find the trigger moment when the current solution fails hard enough to break inertia.
Conducting JTBD Analysis
- Interview recent switchers — people who adopted your product (or a competitor) in the last 90 days. Their memory is fresh.
- Map the timeline: First thought → Passive looking → Active looking → Decision → Purchase → First use → Ongoing use.
- Identify the trigger event: What specific moment made them start looking? This is your marketing gold.
- Map all four forces for each switcher.
- Find patterns across 5+ interviews.
Guidelines
- CRITICAL: NEVER define jobs as features. "I want to export to PDF" is a feature request. The job is "I want to share a polished document with my investors."
- ALWAYS include the situation/context in job statements. The same person has different jobs at different times.
- ALWAYS map all four forces — teams tend to focus on Push and Pull while ignoring Anxiety and Habit (the reasons people DON'T switch).
- NEVER interview only happy customers. Interview people who evaluated you and chose a competitor — they reveal your weaknesses.
- ALWAYS identify the trigger event. Without a trigger, there's no urgency to switch.
Built on Jobs-to-be-Done (Clayton Christensen) and the Forces of Progress (Bob Moesta). Skills from productskills.
Launch based on the size of the bet, not the size of your ambition. Most features deserve a silent or soft launch. Save the big-bang for genuinely new products or major pivots. Linear's principle: launch early, launch often, launch multiple times.
Launch Tiers
Tier 1: Silent Launch
Ship it, turn it on, see what happens. No announcement.
Use for: Bug fixes, minor improvements, internal tools, infrastructure changes, experiments.
Checklist:
- Feature is behind a flag or gradually rolled out
- Monitoring/alerting is in place
- Rollback plan exists
- One person is on point to watch metrics for 24 hours
Tier 2: Soft Launch
Ship it, tell existing users, collect feedback before going wide.
Use for: New features for existing users, significant UX changes, pricing changes, beta programs.
Checklist:
- Changelog entry or in-app notification written
- Help docs updated (or created)
- Support team briefed with FAQ
- Feedback collection mechanism in place (survey, feedback button, Slack channel)
- Known limitations documented
- Rollback plan exists
- 1-2 week feedback period before going wide
Example: Launching real-time collaboration — Changelog: "Edit together, see cursors live." Support FAQ covers latency expectations and max concurrent editors. Roll out to 20% of teams for 3 days via feature flag, then 100%. Rollback: disable WebSocket sync, fall back to polling. Feedback widget live for 2 weeks. Success metric: % of teams with 2+ simultaneous editors.
Tier 3: Big-Bang Launch
Coordinated, public, designed for maximum reach. Expensive in time and attention.
Use for: New product launch, major pivot, rebrand, entering a new market. Maximum 2-3 per year.
Checklist:
Product
- Feature complete and tested at scale
- Onboarding flow tested with 5+ new users
- Performance benchmarked under expected load
Content
- Landing page or product page live
- Blog post / announcement written
- Demo video or screenshots prepared
- Social posts drafted for launch day and day +1, +3, +7
Distribution
- Email to existing users scheduled
- Social media posts queued
- Community posts drafted (relevant subreddits, Slack groups, Discord)
- Product Hunt launch prepared (if applicable)
- Influencer/partner outreach done 1 week before
Support
- Support team briefed and staffed for launch day
- FAQ and known issues documented
- Escalation path defined
Measurement
- Success metrics defined before launch (not after)
- Analytics tracking verified
- Daily review scheduled for first week
Post-Launch
Regardless of tier:
- Day 1: Watch metrics. Fix anything broken. Respond to all feedback.
- Day 3: First retro — what's working, what's not, what surprised you.
- Week 1: Decide whether to iterate, expand, or pull back.
- Week 2-4: Write up learnings. Update the roadmap based on real data.
Guidelines
- CRITICAL: Default to Tier 1 or 2. Upgrade to Tier 3 only with strong justification. Over-launching creates announcement fatigue.
- NEVER launch without a rollback plan. If you can't revert, you're not ready.
- ALWAYS define success metrics BEFORE launch, not after. Post-hoc metrics are just storytelling.
- NEVER launch on a Friday. Launch Monday-Wednesday to have the week for monitoring and iteration.
- ALWAYS have one person on point for the first 24 hours. Launches without owners drift.
- NEVER do a big-bang launch for features with known bugs. Soft launch first, fix, then go wide.
Built on Linear's "launch multiple times" philosophy. Skills from productskills.
Define metrics that drive decisions, not dashboards that collect dust. Every metric must answer: "If this number changes, what will we do differently?" If nothing changes, it's a vanity metric.
North Star Metric
One metric that captures the core value your product delivers to customers. Not revenue (that's an output). Not signups (that's vanity). The North Star reflects the moment customers get value.
Examples:
- Slack: Messages sent in channels with 3+ members
- Spotify: Time spent listening
- Airbnb: Nights booked
Test: If this metric goes up, does the company sustainably grow? If yes, it's your North Star.
Input/Output Tree
Build a tree connecting your North Star to actionable inputs:
North Star: Weekly active teams with 3+ collaborators
|
+-- Acquisition: New signups/week
| +-- Website visitors
| +-- Signup conversion rate
|
+-- Activation: % who invite a teammate in 7 days
| +-- Onboarding completion rate
| +-- Time to first shared project
|
+-- Engagement: Documents edited per active team/week
| +-- Feature adoption (comments, mentions, sharing)
| +-- Return frequency
|
+-- Retention: % active at day 30
+-- Weekly return rate
+-- Feature breadth used
Each leaf is a metric a team can directly influence. The tree shows HOW inputs drive the North Star.
Counter-Metrics
EVERY metric gets a counter-metric. This prevents gaming and ensures you're not optimizing one thing at the expense of another.
| Metric | Counter-Metric |
|---|---|
| Signup conversion rate | Activation rate (don't lower the bar to get signups) |
| Time to first value | 30-day retention (don't rush users past learning) |
| Feature adoption | Task completion rate (don't push features that confuse) |
| Revenue per user | Churn rate (don't squeeze users into leaving) |
If the metric improves but the counter-metric degrades, you've optimized the wrong thing.
Guidelines
- CRITICAL: ALWAYS pair every metric with a counter-metric. No exceptions.
- NEVER use vanity metrics: total signups, page views, total users, app downloads. These only go up and tell you nothing.
- NEVER track more than 5-7 metrics actively. If your dashboard has 20 charts, nobody looks at any of them.
- ALWAYS define metrics with exact formulas. "Activation rate" means nothing. "Percentage of new signups who invite at least one teammate within 7 days of account creation" is a metric.
- ALWAYS set a baseline before setting a target. You can't improve what you haven't measured.
- NEVER set targets without understanding the current number and why it's where it is.
- ALWAYS measure outcomes (user behavior) over outputs (features shipped).
Built on Amplitude's North Star framework and the Linear Method. Skills from productskills.
Map opportunities by connecting business outcomes to customer needs to testable solutions. Teresa Torres' Opportunity Solution Trees (OSTs) prevent the two biggest PM mistakes: building solutions without clear problems, and chasing problems disconnected from business goals.
Opportunity Solution Tree Structure
Build the tree top-down, but fill it bottom-up with evidence:
Desired Outcome (business metric you're trying to move)
|
+-- Opportunity 1 (customer need/pain/desire)
| +-- Solution A
| | +-- Experiment 1
| | +-- Experiment 2
| +-- Solution B
| +-- Experiment 3
|
+-- Opportunity 2
+-- Solution C
+-- Solution D
+-- Experiment 4
Level 1: Desired Outcome
One measurable business outcome. Not a feature, not a project — a metric.
- "Increase 7-day activation rate from 23% to 40%"
- NOT: "Improve onboarding" (not measurable)
Level 2: Opportunities
Customer needs, pain points, or desires that, if addressed, would move the outcome. These come from research — interviews, data, support tickets — not brainstorming.
Rules for good opportunities:
- Framed as customer needs, not product features
- "New users don't understand what to do first" (opportunity)
- NOT "Add an onboarding wizard" (solution masquerading as opportunity)
- Each opportunity is independent — addressing one doesn't depend on another
Level 3: Solutions
Multiple possible solutions for each opportunity. Generate at least 3 before evaluating. The goal is to explore the solution space, not commit to the first idea.
Level 4: Experiments
Small, fast tests to validate whether a solution addresses the opportunity. Experiments should answer: "Does this solution actually solve this opportunity?"
Building the Tree
- Start with the outcome. Align with your team or stakeholders on exactly one outcome to focus on.
- Map opportunities from research. Review interview notes, support tickets, analytics. Cluster evidence into distinct opportunities. Each opportunity needs evidence from 3+ sources.
- Generate solutions per opportunity. Brainstorm at least 3 solutions per opportunity. Include wild ideas — they often reveal assumptions.
- Design experiments per solution. What's the smallest test? Prototype, concierge, Wizard of Oz, fake door, A/B test.
- Prioritize which branch to explore. You can't test everything. Pick the opportunity with strongest evidence and the solution with lowest experiment cost.
Guidelines
- CRITICAL: NEVER skip from outcome directly to solutions. The opportunity layer is where the insight lives.
- ALWAYS frame opportunities as customer needs, not features. If it sounds like a feature, push back to the underlying need.
- NEVER have only one solution per opportunity. If you can't think of 3 solutions, you haven't explored the space.
- ALWAYS ground opportunities in evidence from research, not assumptions.
- NEVER pursue more than 2-3 opportunities simultaneously. Focus beats breadth.
- ALWAYS design experiments that could DISPROVE your solution, not just confirm it.
Built on Continuous Discovery Habits by Teresa Torres. Skills from productskills.
Write PRDs that are evidence-first, concise, and scannable. A PRD is a thinking tool — if it's longer than 1200 words, the author hasn't thought hard enough. Brief specs are more likely to be read, more likely to be followed, and more likely to ship on time.
Structure
Follow this structure. Every section is mandatory unless marked optional.
1. Problem
Lead with customer evidence. Quotes, data, support tickets, observed behavior. If you have no evidence, say so explicitly — that's an opinion-driven PRD and should be flagged.
Example:
"3 of 5 interviewed customers abandon onboarding at the team invite step. Average time spent: 47 seconds before closing the tab."
NOT: "Users find onboarding confusing."
2. Goals
Measurable outcomes, not outputs. Each goal has a number attached.
- "Reduce onboarding drop-off from 60% to under 30%"
- NOT "Improve the onboarding experience"
3. Target Users
Who specifically benefits. Be narrow. "Series A startup founders doing PM work before their first PM hire" — not "product managers."
4. Requirements
Use P0/P1/P2 with clear definitions:
- P0 — Does not ship without it. The feature is broken or useless without this.
- P1 — Ships without it, but significantly degraded. Do in same cycle if possible.
- P2 — Nice to have. Candidate for fast follow.
Each requirement is one sentence. If a requirement needs a paragraph to explain, it's too big — break it down.
5. Success Metrics
Numbers. Always numbers. "Reduce X from Y to Z within W weeks."
ALWAYS pair each metric with a counter-metric to prevent gaming:
- Metric: "Increase activation rate to 70%"
- Counter-metric: "Without decreasing 30-day retention below 45%"
6. Out of Scope
What you are deliberately NOT building and why. This section prevents scope creep and signals that you've thought about boundaries.
7. Open Questions (optional)
Unresolved decisions that need input. Each question names who should answer it and by when.
Guidelines
- CRITICAL: 800-1200 words maximum. Longer PRDs mean the thinking isn't done.
- NEVER start from a solution. Start from a problem with evidence.
- NEVER include implementation details — architecture, APIs, database schemas. That's spec territory, not PRD territory.
- ALWAYS include Out of Scope. A PRD without boundaries invites scope creep.
- ALWAYS trace requirements to customer evidence. No evidence = opinion.
- NEVER use weasel metrics: "improve performance," "enhance experience," "increase engagement." Put a number on it or don't include it.
- P0 requirements should number 3-5. If you have more than 5 P0s, you haven't prioritized.
Example: Good vs Bad
Bad P0 requirement:
"The system should handle authentication in a secure and user-friendly manner, supporting multiple providers and edge cases."
Good P0 requirement:
"P0: User can sign up and log in with email + password. Google OAuth is P1."
One sentence. Clear scope. Priority forces a decision.
Built on Linear's "short specs > exhaustive docs" philosophy. Skills from productskills.
Validate problems with evidence, not opinions. A problem worth solving is one that people encounter frequently, feel intensely, and have already tried to solve. If nobody has attempted a workaround, the problem isn't painful enough to build for.
Scoring Framework
Rate each dimension 1-5 based on evidence:
Frequency (How often does this happen?)
- 5: Daily or multiple times per day
- 4: Weekly
- 3: Monthly
- 2: Quarterly
- 1: Rarely or once
Intensity (How painful is it when it happens?)
- 5: Blocks critical work, causes real losses (money, time, reputation)
- 4: Major friction, significant workarounds needed
- 3: Annoying but manageable
- 2: Minor inconvenience
- 1: Barely noticeable
Existing Workarounds (Are people already trying to solve it?)
- 5: Paying for imperfect solutions or built custom tools
- 4: Cobbled together multi-tool workflows (spreadsheets + email + manual steps)
- 3: Have a basic process but it's tedious
- 2: Occasionally Google for solutions
- 1: Haven't tried to solve it
Willingness to Pay (Would they pay to make this go away?)
- 5: Already spending money on partial solutions
- 4: Have explicitly said they'd pay (and named a number)
- 3: Would "probably" pay (hypothetical — discount this)
- 2: Want it free
- 1: Haven't considered paying
Validation Score = Frequency x Intensity x Workarounds x WTP
- 250+: Strong signal. Build.
- 100-249: Promising. Needs more evidence on weak dimensions.
- Under 100: Weak. Either pivot the problem framing or walk away.
Example: "PDF Export" — Frequency: 3 (monthly board reports), Intensity: 2 (screenshot workaround exists), Workarounds: 4 (3 users built browser-print-to-PDF workflows), WTP: 1 (nobody has paid for an export tool). Score = 3 x 2 x 4 x 1 = 24. Verdict: Kill — not painful enough.
Evidence Requirements
Every score MUST cite evidence. Acceptable evidence types, strongest first:
- Observed behavior: You watched someone struggle with this
- Spending: They pay for alternatives or workarounds
- Time invested: They built custom solutions (scripts, spreadsheets, processes)
- Quotes from interviews: Direct quotes about past behavior (Mom Test compliant)
- Support tickets / forum posts: Public complaints with specifics
- Survey responses: Weakest — people say one thing and do another
If your only evidence is survey responses or "my friend said," your validation is weak.
Go / No-Go Decision
Go when:
- Score 250+ with evidence across all four dimensions
- At least 5 people independently describe the same problem
- Someone has spent money or significant time on workarounds
Investigate more when:
- Score 100-249 with strong evidence on 2-3 dimensions
- You've talked to fewer than 5 people
- Workaround evidence is weak
Kill when:
- Score under 100
- Nobody has tried to solve the problem themselves
- "Willingness to pay" is entirely hypothetical
- You've talked to 5+ people and stories don't converge
Guidelines
- CRITICAL: NEVER validate based on opinions or surveys alone. Require evidence of past behavior.
- ALWAYS require workaround evidence. No workarounds = not painful enough.
- NEVER count "I would definitely use that!" as validation. Enthusiasm is not evidence.
- ALWAYS talk to at least 5 people in the ICP before making a go/no-go call.
- NEVER confuse "interesting problem" with "problem worth solving." Interesting doesn't pay the bills.
Built on YC's "talk to users" philosophy and The Mom Test (Rob Fitzpatrick). Skills from productskills.
Position products by starting from what customers would do without you — not from your features or aspirations. April Dunford's Obviously Awesome framework works because it's grounded in reality: what alternatives exist, what you do differently, and who cares most about that difference.
The Five Steps (In Order)
These steps are sequential. Do NOT skip ahead or rearrange.
1. Competitive Alternatives
What would your best customers do if you didn't exist? List real alternatives:
- Direct competitors
- Indirect solutions (spreadsheets, hiring someone, different tools)
- Do nothing
Be honest. If "do nothing" is the primary alternative, that tells you a lot about urgency.
2. Unique Attributes
What do you have that the alternatives don't? List concrete, verifiable capabilities — not marketing spin.
- "Real-time collaboration on specs" (verifiable)
- NOT "best-in-class experience" (unverifiable)
3. Value
What do those unique attributes ENABLE for the customer? Translate features into outcomes.
- Attribute: "Real-time cursors and inline comments on any document"
- Value: "Get feedback from your team in minutes instead of waiting for a meeting"
4. Best-Fit Customers
Who cares MOST about the value you deliver? Define them tightly:
- "Remote teams of 5-15 who collaborate daily but drown in Slack threads and lost Google Docs"
- NOT "product managers" (too broad)
Best-fit customers have: the problem acutely, tried alternatives, have budget/authority, and get the most value from your unique attributes.
5. Market Category
What market do you position in so that your value is obvious? This is the LAST decision, not the first.
Options:
- Existing category: "We're a [known category] that [key differentiator]." Works when the category is well-understood and your differentiator is strong.
- Subcategory: "We're [modifier] + [existing category]." Works when you serve a specific segment better. E.g., "async-first collaboration tool."
- New category: Create a new name. Only works if you have the budget and patience to educate the market. Almost always the wrong choice for startups.
Guidelines
- CRITICAL: NEVER position in a vacuum. Always start from competitive alternatives.
- NEVER lead with the market category. It's the output of positioning, not the input.
- ALWAYS position for your best-fit customers, not for everyone. Trying to appeal to everyone means resonating with no one.
- NEVER confuse features with value. Features are what you built. Value is what the customer gets.
- ALWAYS pressure-test by asking: "If I told my best-fit customer this positioning in one sentence, would they immediately get why they should care?"
- NEVER use superlatives you can't prove ("best," "most powerful," "leading"). Use specifics.
Built on Obviously Awesome by April Dunford. Skills from productskills.
Turn raw research into atomic insights that drive decisions. Good synthesis surfaces patterns, bad synthesis creates narrative fiction. The goal is structured evidence, not a compelling story that cherry-picks quotes.
Atomic Research Method
Break research into four levels, bottom-up:
1. Nuggets (Raw Evidence)
Individual observations from a single source. Each nugget is:
- One observation per nugget (not a paragraph)
- Tagged with source (participant ID, date, method)
- Direct quotes preferred over your interpretation
Example: "[P3, Jan 12] 'I spend 30 minutes after every customer call just trying to remember what they said.'"
2. Patterns (Recurring Themes)
Group nuggets that point to the same phenomenon. A pattern requires evidence from 3+ independent sources.
Example: "5 of 7 PMs report spending 20-45 minutes on post-call documentation. All describe it as tedious and low-value."
3. Insights (Implications)
What the pattern means for the product. An insight connects a pattern to a product opportunity or risk.
Example: "Post-call documentation is a high-frequency pain point (daily for active PMs) with no satisfying solution. Current workarounds (voice memos, bullet lists) lose context and emotional nuance."
4. Recommendations (Actions)
Specific product actions justified by insights. Each recommendation traces back through the chain: recommendation ← insight ← pattern ← nuggets.
Evidence Strength
Rate every pattern and insight:
| Strength | Criteria |
|---|---|
| Strong | 5+ sources, consistent behavior observed, corroborated by data |
| Moderate | 3-4 sources, mostly consistent, some data support |
| Emerging | 2 sources, needs more evidence before acting |
| Weak | Single source or contradictory evidence |
NEVER make product recommendations based on Weak or Emerging evidence. Flag them for further research.
Synthesis Process
- Extract all nuggets from raw notes — one per line, tagged with source
- Affinity map: group nuggets by theme (not by interview)
- Name each group as a pattern with a count ("5/7 participants...")
- Derive insights from strong and moderate patterns only
- Write recommendations that trace back to specific insights
- Flag contradictions explicitly — don't smooth them over
Guidelines
- CRITICAL: NEVER synthesize from fewer than 3 data points. Two people saying the same thing is a coincidence, not a pattern.
- ALWAYS cite specific evidence. "Users struggle with X" is worthless. "5 of 7 participants spent 20+ minutes on X, with P2 calling it 'the worst part of my week'" is evidence.
- ALWAYS weight actions over opinions. What people DO matters more than what they SAY.
- NEVER cherry-pick quotes to support a narrative. Include contradictory evidence.
- ALWAYS flag your confidence level. Distinguish "we know" from "we think."
- NEVER present synthesis without the evidence chain. Anyone reading should be able to trace a recommendation back to raw nuggets.
Built on the Atomic Research method (Daniel Pidcock) and Continuous Discovery Habits (Teresa Torres). Skills from productskills.
Build roadmaps organized by outcomes, not feature lists. A roadmap answers "what problems are we solving and in what order?" — not "what features will we ship and when?" Dates on a roadmap beyond 6 weeks are fiction. Treat them that way.
Now / Next / Later
Three time horizons, decreasing in certainty:
Now (Committed — this cycle)
- Currently in progress or about to start
- Fully shaped: problem defined, solution scoped, appetite set
- Team assigned, expected to ship this cycle
- 1-3 items maximum
Next (Shaped — next 1-2 cycles)
- Problem validated, solution partially shaped
- Not yet assigned to a team
- May change based on what we learn from "Now" items
- 3-5 items maximum
Later (Raw — ideas worth exploring)
- Problems we believe are real but haven't validated
- No solution shaped yet
- Will be promoted to "Next" when evidence supports it, or killed
- No limit, but prune quarterly
Roadmap Items Are Outcomes, Not Features
Each roadmap item is framed as a problem to solve or an outcome to achieve:
Bad (feature-list roadmap):
- Build team collaboration features
- Add CSV export
- Redesign settings page
Good (outcome-based roadmap):
- Reduce time-to-first-value for team signups from 15 min to under 5 min
- Enable users to get their data out without contacting support
- Reduce settings-related support tickets by 50%
The solution emerges during shaping, not during roadmap planning.
Shape Up Cycle Planning
For teams using Shape Up cycles:
- Betting table: Leadership reviews shaped pitches and bets on what to build next cycle. Not everything gets picked.
- Cool-down: 1-2 weeks between cycles for bug fixes, exploration, and shaping future work.
- No carry-over: Work that doesn't ship in a cycle is not automatically carried over. It must be re-pitched.
Presenting the Roadmap
Keep it to one page or one screen. If your roadmap needs scrolling, it's too detailed.
Format as a simple table:
| Horizon | Outcome | Evidence | Status |
|---|---|---|---|
| Now | Reduce onboarding drop-off to under 30% | 5 interviews, 60% current drop-off | In progress, week 2/3 |
| Next | Enable self-serve data export | 12 support tickets/month | Shaped, needs team |
| Later | Multi-workspace support | 3 enterprise prospects requested | Unvalidated |
Guidelines
- CRITICAL: NEVER put dates on roadmap items beyond 6 weeks. You don't know and pretending you do erodes trust.
- NEVER build a feature-list roadmap. Organize by outcomes and problems.
- ALWAYS limit "Now" to 1-3 items. If you're working on 8 things, you're finishing none of them.
- NEVER let "Later" items skip to "Now" without shaping and evidence.
- ALWAYS prune "Later" quarterly. Kill items that haven't gained evidence in 3 months.
- ALWAYS review and update the roadmap at the start of every cycle, not just quarterly.
- NEVER present a roadmap without evidence for each item. A roadmap without evidence is a wish list.
Built on Shape Up (Basecamp) cycle planning and the Now/Next/Later framework. Skills from productskills.
Cut scope by flipping the question. Instead of "How long will this take?" ask "How much time is this problem WORTH?" Set the appetite first, then shape the solution to fit. This is Shape Up's core insight: fixed time, variable scope.
Set the Appetite
Before touching scope, declare the time budget:
- Small batch: 1-2 weeks for 1-2 people
- Big batch: 6 weeks max for a small team
If a project takes more than 6 weeks (or 3 weeks for 1-3 people per Linear's method), it's too big. Break it down or walk away.
The appetite is NOT an estimate. An estimate says "this will take X." An appetite says "this is WORTH X." If the shaped solution doesn't fit the appetite, cut scope until it does — or kill the project.
Scope Hammering
For every item in the spec, ask these questions in order:
- Must-have or nice-to-have? If the feature ships without it and users still get value, it's nice-to-have. Cut it.
- Shippable without it? Can you deploy a working version that doesn't include this? If yes, cut it from v1.
- Edge case or core path? If fewer than 20% of users will encounter this scenario, defer it.
- Compare to baseline. The baseline is NO feature — not the ideal version. "Is this better than nothing?" always wins over "Is this as good as it could be?"
Shape the Solution Down
When scope is still too big after hammering:
- Cut breadth, not quality. Support 1 auth provider well instead of 3 badly. Handle the happy path completely instead of covering every error state.
- Remove entire surfaces. Kill the settings page. Remove the dashboard. Ship with sensible defaults instead of configurability.
- Time-bound the edge cases. "We'll handle CSV import of up to 1,000 rows. Above that, they email us." Solve it manually until scale demands automation.
- Ask: "What's the simplest version a customer would thank us for?" That's your scope.
Circuit Breaker (Shape Up)
If it doesn't ship in one cycle, it's cancelled by default. No automatic rollover. The team must re-pitch for another cycle. This prevents zombie projects that drag on forever.
Guidelines
- CRITICAL: NEVER cut quality — cut BREADTH. A polished feature that does one thing > a buggy feature that does five.
- NEVER estimate. Set appetite. "This is worth 2 weeks" not "This will take 2 weeks."
- NEVER let v1 have more than 5 must-haves. If it does, you haven't scoped hard enough.
- ALWAYS compare to baseline (no feature at all), not to the ideal implementation.
- ALWAYS ask "What can we ship in [appetite]?" not "How long until we ship everything?"
- NEVER carry work over automatically. Unfinished work must be re-pitched.
Built on Shape Up (Basecamp) and the Linear Method. Skills from productskills.
Write strategy that forces hard choices. A strategy that doesn't say "no" to something isn't a strategy — it's a wish list. Strategy is not goals, not aspirations, and not a list of things you want to do. It's a coherent set of choices about where to play and how to win.
The Strategy Kernel (Richard Rumelt)
Every strategy document must contain these three elements:
1. Diagnosis
What's actually going on? Name the challenge clearly. A good diagnosis simplifies complexity by identifying the critical factors.
- "Our activation rate is 23% because new users don't understand the product's value in their first session."
- NOT: "We need to grow faster." (That's an aspiration, not a diagnosis.)
2. Guiding Policy
The overall approach for dealing with the challenge. This is the big directional bet — it rules things in AND rules things out.
- "Focus entirely on time-to-first-value for solo users before expanding to teams."
- NOT: "Improve the product across all dimensions." (That's not a choice.)
3. Coherent Actions
Specific, coordinated actions that execute the guiding policy. Actions should reinforce each other.
- "Rebuild onboarding as a guided first-project flow. Remove the team invite step from signup. Add inline tooltips on the three core features. Measure activation at 'first project completed' not 'account created.'"
Playing to Win Framework (Lafley/Martin)
Structure the strategy as five cascading choices:
- Winning Aspiration: What does winning look like? (Not revenue targets — the change you want to create)
- Where to Play: Which customers, segments, geographies, channels? Be specific about what you're NOT pursuing.
- How to Win: What's your competitive advantage in your chosen space?
- Capabilities: What must you be great at to win this way?
- Management Systems: How will you track whether the strategy is working?
Each choice constrains the next. If "Where to Play" doesn't narrow the field, it's not a choice.
Guidelines
- CRITICAL: Every strategy doc MUST include what you will NOT do. A strategy without tradeoffs is not a strategy.
- NEVER confuse strategy with goals. "Reach $10M ARR" is a goal. "Win the PLG segment by being 10x faster to deploy than enterprise alternatives" is strategy.
- NEVER list more than 3 strategic priorities. If everything is a priority, nothing is.
- ALWAYS include a diagnosis that names the core challenge. If you can't name the problem, you can't solve it.
- ALWAYS make the guiding policy falsifiable. Someone should be able to argue against it.
- NEVER write strategy in a vacuum. Ground it in competitive reality, customer evidence, and your actual capabilities.
Example: Good vs Bad
Bad guiding policy:
"We will build the best product in the market by focusing on quality, speed, and customer satisfaction."
Good guiding policy:
"We will win technical PMs at Series A-B startups by being the fastest path from customer insight to shipped feature — sacrificing enterprise compliance features and multi-team coordination to stay opinionated and fast."
The good version names who, names what you're sacrificing, and could be argued against.
Built on Good Strategy Bad Strategy (Richard Rumelt) and Playing to Win (Lafley/Martin). Skills from productskills.
Run user interviews that surface real problems, not polite lies. Most interviews fail because the interviewer pitches their idea and asks "would you use this?" The Mom Test (Rob Fitzpatrick) fixes this: talk about their life, not your idea.
The Five Questions (Eric Migicovsky, YC)
Use these as the backbone of every interview. They work in order — each question digs deeper.
-
What's the hardest part about [doing the thing you're exploring]? Opens with their pain, not your solution.
-
Tell me about the last time you encountered that problem. Forces specifics. Past behavior > hypothetical opinions.
-
Why was that hard? Uncovers root cause. Their first answer is usually surface-level.
-
What, if anything, have you done to try to solve this? Existing workarounds = real pain. No workarounds = not painful enough.
-
What don't you love about the solutions you've tried? Reveals gaps in existing alternatives. This is your opportunity space.
Mom Test Rules
These rules override all default interviewing instincts:
- NEVER ask "Would you use this?" or "What would you pay?" — hypothetical questions get hypothetical answers.
- NEVER pitch your solution during the interview. The moment you pitch, you stop learning.
- ALWAYS ask about past behavior. "Tell me about the last time..." not "Would you ever..."
- Deflect compliments immediately. "That sounds great!" → "Thanks — but tell me more about how you currently handle this."
- Anchor fluff to specifics. "I always struggle with..." → "When was the last time? Walk me through it."
Follow-up Framework (Gustaf Alstromer, YC)
After the Five Questions, probe for signal strength:
- Frequency: "How often does this come up?" (Daily = strong signal)
- Workarounds: "Show me how you do it today." (Screenshots, spreadsheets, hacks = gold)
- Willingness to pay: "Have you looked for or paid for a solution?" (Spending = strongest signal)
- Emotional weight: Watch for energy shifts — frustration, sighing, storytelling. That's where the real pain lives.
Conducting the Interview
- 80% listening, 20% questioning. If you're talking more than them, you're doing it wrong.
- Take notes on exact quotes, not your interpretation.
- Seek advancement, not enthusiasm. Advancement = they give you time, reputation, or money (intro to a colleague, agree to a pilot, share their data). Enthusiasm = "sounds cool!" (worthless).
- Stop interviewing when you stop hearing new things. Usually 5-8 conversations for a focused problem space.
Guidelines
- CRITICAL: The interview is about THEIR life, not YOUR idea. If you mention your product at all, save it for the last 2 minutes.
- NEVER ask leading questions. "Don't you think X would be better?" is a pitch disguised as a question.
- ALWAYS get specific. Generalities are worthless. "Some people" → "Who?" "Sometimes" → "When was the last time?"
- NEVER end without asking: "Who else should I talk to about this?" — your best interviewees come from referrals.
- ALWAYS include a reference to
references/interview-guide-template.mdfor the full question template when building an interview guide.
Built on The Mom Test (Rob Fitzpatrick), YC's Five Questions (Eric Migicovsky), and Gustaf Alstromer's follow-up framework. Skills from productskills.
Sponsors
Details
- Marketplace
- assimovt/productskills
- Category
- Community