# Ideal customer profile
Use this method to answer which customers to pursue first and why. It may produce an internal strategy memo or a client-ready research artifact, but it keeps evidence, assumptions, and confidential inputs appropriate to the selected audience.
Scope
Use this method for target-segment decisions. Route broad market size, competitive structure, or pricing power questions to their respective methods; use vendor methods for a technology or supplier decision. Do not convert public research into a named-account list or claim that public evidence alone validates demand.
Method
- State the offer, decision, candidate segments, geography, and audience for the artifact. Use supplied first-party evidence as private context unless its use in the output is authorized.
- Build an evidence plan covering population, organizations, buyer/workflow evidence, buying ecosystem, alternatives, and timing signals. Gather first-party material and dated public primary sources where available.
- Compare two or three candidate segments against the same qualitative criteria: problem intensity, ability to pay, reachability, fit with the offer and proof, buying complexity, and timing. Mark each `Strong`, `Mixed`, `Weak`, or `Unknown`; do not manufacture a weighted score.
- Recommend a primary ICP only when the evidence clears the stated gates. Otherwise return a partial result and the smallest validation action that can resolve the limiting gap.
- For the leading segment, map the economic buyer, champion, users, blockers, job to be done, alternatives, decision stages, and observable timing signals.
- Define exclusions and anti-ICP conditions as clearly as inclusions.
- Turn every material uncertainty into a practical validation: first-party data cut, win/loss review, customer interview, buyer conversation, or bounded message test.
Evidence rules
- Capture source URLs while researching, cite non-trivial external claims inline, list sources with short descriptors, and re-check claim-to-source pairing before finalizing.
- Keep first-party provenance at source-type and date level unless the audience has approved more detail.
- Distinguish `Hypothesis`, `Supported`, `Validated`, and `Partial`. Public research alone is not `Validated`. Mark the result `Partial` when an applicable evidence lane is missing or too thin.
- Track source independence: repeated summaries of one interview remain one source. Label evidence as observed, reported, inferred, or unknown and retain conflicts about buyer, budget, urgency, or fit.
- Base an ICP recommendation on three independent applicable lanes: authoritative population evidence where available, primary organization evidence, and observed buyer, workflow, or purchase evidence. A vendor, analyst, or consultancy source cannot be the only support.
Decision and handoff discipline
Compare candidates on a common basis: job intensity, evidence of pain, ability to buy, reachable channel, offer fit, constraints, and expected value. Describe user, champion, economic buyer, and approver separately when known. A profile is a decision hypothesis, not proof of total market demand; distinguish a pilot target from a scaled segment. Hand population and revenue-pool questions to market sizing and offer or price questions to their owning methods.
Output
```markdown # [Offer or market] ideal customer profile ## Decision summary ## Scope and evidence coverage ## Candidate comparison ## Primary ICP card ## Secondary hypothesis ## Anti-ICP and exclusions ## Jobs to be done and buying group ## Timing signals ## Validation plan ## Evidence ledger ## Sources ```
Before a durable or client-facing write, present the proposed recommendation or reason no segment clears the gate, candidate comparison, limitations, exclusions, and validation plan for confirmation. Do not write to CRM, outreach, or another system of record; return a proposed handoff where a host workflow owns it.