Skip to main content

Ideal Customer Profile

Which customer situation can you recognise, reach, and help better than the plausible alternatives?

An ideal customer profile is a falsifiable archetype, not a flattering persona or a list of demographic labels. It connects a struggling moment, desired progress, buying context, switching forces, and evidence.

Changes my mind: Real conversations and choices repeatedly show that the profile does not predict shared progress, urgency, access, or buying behaviour.

Model​

situation → struggle → desired progress → switching forces → evidence → profile decision

Ideal Customer Profile​

  • Valuable outcome: a specific, reachable customer archetype tied to a struggling moment and valuable progress.
  • Trigger and inputs: a validated opportunity, offer definition, customer evidence, loss reasons, and market observations.
  • Required business skills: interviewing, segmentation, behavioural interpretation, evidence grading, and strategic exclusion.
  • Tool categories and selection: choose research, CRM, survey, and analysis tools that preserve source context and expose uncertainty.
  • Concrete outputs: segment boundary, struggling moment, buying roles, switching forces, exclusions, evidence register, and refresh date.
  • Human edge and safe AI: humans choose whom to serve and interpret context. AI may find patterns and draft hypotheses; it must not invent interviews or use sensitive traits as a shortcut for desirability.
  • Proof, cadence, and kill signals: test whether the profile predicts real conversations or choices. Review after material market evidence and at least twice yearly. Kill or split it when members share labels but not a buying situation, or evidence repeatedly contradicts it.

Practice​

Write one sentence: “When [situation] causes [struggle], [role] seeks [progress] but hesitates because [switching force].” Attach at least three source observations, label assumptions, and list who is explicitly outside the profile. Test it in five relevant conversations before increasing precision.

Run It​

Copy this prompt and replace the bracketed fields. Use only evidence you can inspect or source.

Help me define one falsifiable ideal customer profile.

Offer or opportunity: [what may create value]
Decision this profile must support: [the decision]
Observed customer evidence: [interviews, choices, losses, support records, or
other supplied observations]
Reachable channels: [where this customer can actually be reached]
Constraints and exclusions: [who or what must not be targeted]
Review date: [date]

Separate observations, inferences, hypotheses, and unknowns. Do not invent
interviews, market data, customer quotes, urgency, or buying behaviour.

Produce:
1. one reachable ICP hypothesis;
2. its struggling moment and desired progress;
3. buying roles and switching forces;
4. an evidence register with source and confidence;
5. explicit exclusions and two plausible alternative profiles;
6. the unknowns and interview questions that matter next;
7. a five-conversation proof plan;
8. kill or split signals;
9. the final human decision: approve, revise, or reject, with reasons.

The human owner chooses whom to serve. Treat the generated profile as a hypothesis until observed conversations and choices support it.

Learning Return​

Return evidence that changes the struggling moment, segment boundary, reachability, exclusions, or switching forces to the next profile review. Preserve rejected hypotheses so later evidence can distinguish a bad segment from a weak offer or channel.

Failure modes​

A demographic label without a shared situation predicts little. A profile built only from successful customers hides losses and exclusions. False precision turns assumptions into facts.

Context​

Questions​

Next question: Which returned observation should retain, split, or end this profile before the next distribution bet?

  • Which source is observed rather than inferred?
  • Which rejected profile should remain available for comparison?