Founder debugger

Purpose

Prevent symptom treatment. Locate the earliest broken link in the startup system, state the evidence, and route to one focused test and one retained specialist skill.

Give the founder a decision, the evidence behind it, the strongest contrary case, and the smallest executable next step. Do not produce generic encouragement or a long menu of tactics.

When to Trigger

Use this skill according to the frontmatter description. If the stated issue is a symptom of an earlier broken link, say so and route to founder-debugger or the owning retained skill.

Inputs Needed

• Stage, target user, product, and business model
• Runway, gross burn, net burn, revenue, and growth
• Acquisition by channel and segment
• Activation and time-to-value
• Core-action retention cohorts
• Pricing, sales funnel, and loss reasons
• Current weekly priority, shipping cadence, and founder time
• Recent user evidence and major assumptions
• What changed immediately before the problem appeared

Ask only for missing inputs capable of changing the decision. For reversible actions, make assumptions explicit and propose a bounded test instead of blocking on perfect data.

Questions to Ask

• What exact outcome is stuck?
• When did it change, and for which segment?
• Is the problem real and urgent without our solution?
• Can a target user receive the promised value now?
• Where does the first-value path fail?
• Do activated users repeat the core action at the natural cadence?
• Do good-fit users understand and buy the value?
• Can we repeatedly reach more users like retained customers?
• Is execution failing despite a correct diagnosis?
• Does the current trajectory reach breakeven before cash runs out?
• What evidence would disprove the team's favorite explanation?

Mental Models

• Earliest broken link.
• Symptom versus mechanism.
• Stage before tactic.
• Cohort before aggregate.
• Behavior before narrative.
• One constraint before backlog.
• Fundraising is downstream of company quality.
• Advice is a hypothesis, not authority.

Use these models as competing lenses. Select the one that best explains the observed behavior, state what evidence would falsify it, and convert it into a decision rather than repeating it as a slogan.

YC Principles

• Before PMF, the central loop is build product, talk to users, and iterate.
• Do not scale acquisition into weak activation or retention.
• Investors cannot repair a company whose founders do not understand the problem.
• Slow development, investor-driven strategy, copying peers, and cofounder conflict can masquerade as market problems.
• Use measurement to run hypothesis-test-conclude-repeat rather than arguing from intuition alone.

Paul Graham Principles

• Most startup deaths trace to a small set of founder, user, focus, spending, launch, and financing errors.
• Be stubborn about creating value but flexible about the method.
• Default-dead math changes every other recommendation.
• When users do not want the product enough, hiring and scaling usually make repair harder.

Garry Tan Principles

• Demand high agency: face the facts, get close to the work, and run the next test.
• Preserve craft and user truth while using AI to accelerate analysis.
• Do not let strategy theater replace building, selling, or direct observation.

Decision Frameworks

Ordered diagnostic gates

1. Problem: a narrow segment has a frequent, severe problem and costly alternative.
2. MVP value: a real user can complete the core job.
3. Activation: good-fit users reach first value reliably.
4. Retention: cohorts repeat the core action at the natural cadence.
5. Pricing: value capture and bill structure fit the value and buyer.
6. Positioning/sales: the right buyer understands, trusts, and can purchase.
7. Distribution: one route repeatedly reaches retained users.
8. Execution: the team finishes the right work quickly.
9. Default alive: cash reaches breakeven or a credible financing milestone.
10. Fundraising: only now decide whether capital changes the mechanism.

Contradiction test

1. If signups are high but activation is low, acquisition is not the bottleneck.
2. If activation is high but retention falls continuously, onboarding is not sufficient.
3. If retained users refer but prospects do not convert, inspect targeting, positioning, price, or sales.
4. If every segment retains poorly, revisit problem or product value.
5. If one segment retains strongly, concentrate before broadening.
6. If the plan is correct but nothing ships, diagnose priorities, ownership, scope, or founder time.

Evidence report

1. State the symptom.
2. Name the earliest failed gate.
3. Cite the strongest supporting and contrary evidence.
4. Name what is not yet diagnosed.
5. Choose one two-week test with owner, metric, threshold, and stop rule.
6. Route to exactly one primary skill and optionally one supporting skill.

Step-by-Step Process

1. Write facts without explanations.
2. Segment every funnel and cohort by user quality and acquisition source.
3. Run the gates in order; do not skip to the founder's proposed remedy.
4. At the first failed gate, test whether the evidence is instrumentation error, selection error, or a real mechanism.
5. Identify the smallest experiment capable of changing the diagnosis.
6. Freeze downstream scaling until the gate passes.
7. Assign a founder owner and short deadline.
8. Precommit metric, threshold, and stop rule.
9. Run the test and inspect behavior plus qualitative mechanism.
10. Route the next decision to the retained skill that owns the failed gate.

Checklists

Diagnosis

• Company stage and segment are explicit.
• Metrics are cohort-based and defined.
• The earliest failed gate is named.
• Contrary evidence is included.
• Fundraising is not used as a generic answer.

Prescription

• One primary skill owns the next step.
• One test can settle the key uncertainty.
• Owner, deadline, metric, threshold, and stop rule exist.
• Downstream work to pause is named.
• Safety, legal, personnel, or insolvency issues are escalated appropriately.

Red Flags

• The founder asks for channels before showing retention.
• The team blames price when users never receive value.
• Averages combine unrelated segments or cohorts.
• The diagnosis is a list of ten possible problems with no ordering.
• Hiring is proposed before the work is understood.
• Fundraising is proposed because runway is low, without default-alive math.
• Advice follows a famous company analogy instead of current evidence.
• No one owns the proposed test.

Common Mistakes

• Accepting the founder's framing.
• Treating correlation as mechanism.
• Optimizing the easiest metric instead of the earliest broken link.
• Confusing revenue with PMF or signups with distribution.
• Recommending a full strategy before running a reversible test.
• Ignoring founder conflict, slow shipping, or poor instrumentation.
• Routing to multiple skills and recreating the original ambiguity.

Metrics

• Time from symptom to named failed gate
• Share of diagnoses with cohort evidence
• Routed test completion
• Diagnosis reversal rate
• Bottleneck movement after test
• Downstream work paused
• Founder hours to decisive evidence
• Unresolved instrumentation gaps
• Runway consumed per diagnostic cycle
• Recurrence of the same mechanism

For every metric, define the unit, numerator, denominator, cohort, segment, cadence, source event, and owner. Prefer decision thresholds and cohort movement over universal benchmarks.

Example Scenarios

Scenario 1

Many demos, few sales, weak activation: diagnose product value or onboarding before outbound, price, or sales hiring.

Scenario 2

Strong retention in agencies, weak aggregate retention: diagnose segment dilution; concentrate distribution and pricing around agencies.

Scenario 3

Good retention and sales but four months runway: route first to default alive; capital may be appropriate only after conservative runway and milestone analysis.

AI Prompt Templates

Template 1

Diagnose this startup using the ordered gates. Do not accept my proposed cause. State the earliest failed gate, evidence, strongest contrary case, one test, and one primary skill route.
End with a decision, owner, deadline, metric, threshold, stop rule, and strongest contrary case.

Template 2

Turn this funnel, cohort, pricing, runway, and interview evidence into a symptom-mechanism map. Separate measurement gaps from company problems.
End with a decision, owner, deadline, metric, threshold, stop rule, and strongest contrary case.

Template 3

Run a pre-mortem on this proposed fix. Which earlier gate could make it useless, and what cheap test should precede it?
End with a decision, owner, deadline, metric, threshold, stop rule, and strongest contrary case.

Related Skills

• founder-talking-to-users
• founder-mvp
• founder-retention
• founder-pricing
• founder-distribution
• founder-prioritization
• founder-default-alive

Further Reading

1. YC's Essential Startup Advice — The early-stage order of operations: launch, talk to users, iterate, and delay scaling until people want the product.
2. The 5 Things That Kill Startups Post-Seed — A compact diagnostic list: missing PMF, investor-driven strategy, copying peers, slow development, and cofounder conflict.
3. The Scientific Method for Startups — A disciplined hypothesize-test-conclude-repeat loop supported by instrumentation.
4. The Real Product-Market Fit — Distinguishes a decent business or product from unmistakable market pull.
5. Office Hours with Michael Seibel — Concrete founder diagnosis across PMF, financing, advice, and operating stage.
6. Why Investors Can't Fix Your Company — Why founders cannot outsource product, strategy, or company repair to investors.

Source Links

• YC's Essential Startup Advice — Y Combinator. The early-stage order of operations: launch, talk to users, iterate, and delay scaling until people want the product.
• The 5 Things That Kill Startups Post-Seed — Y Combinator. A compact diagnostic list: missing PMF, investor-driven strategy, copying peers, slow development, and cofounder conflict.
• The Scientific Method for Startups — Michael Seibel. A disciplined hypothesize-test-conclude-repeat loop supported by instrumentation.
• The Real Product-Market Fit — Michael Seibel. Distinguishes a decent business or product from unmistakable market pull.
• Office Hours with Michael Seibel — Michael Seibel. Concrete founder diagnosis across PMF, financing, advice, and operating stage.
• Why Investors Can't Fix Your Company — Dalton Caldwell and Michael Seibel. Why founders cannot outsource product, strategy, or company repair to investors.
• The 18 Mistakes That Kill Startups — Paul Graham. Failure patterns spanning founders, users, focus, launch, spending, and fundraising.
• Relentlessly Resourceful — Paul Graham. Resourcefulness as the ability to find another path rather than remain attached to one method.
• Default Alive or Default Dead? — Paul Graham. The defining test: under conservative constant-expense and recent-growth assumptions, does cash reach breakeven?
• Startup Playbook — Sam Altman. A broad operating reference for product, growth, focus, hiring, competition, and execution.