Skills on AI

Active theme: Light

Customer Persona

Use when asked to build a customer persona — a composite profile of a customer segment, covering goals, pain points, and behavior, used to guide product and marketing decisions — distinct from a [[customer-journey-map]], which maps what that customer actually goes through rather than who they are.

A customer persona is a composite profile representing a real segment of customers, built to give a team a concrete, shared reference for who they're building or marketing for. It's not a biography of one individual and not a demographic label — it's a synthesis of patterns seen across many real customers, used to keep product and marketing decisions grounded in an actual audience instead of whoever's in the room.

Key components

  • Goals and motivations — what the customer is actually trying to accomplish, and why it matters to them, not just what feature they ask for.
  • Pain points — the specific frictions, frustrations, or constraints that get in the way of that goal today.
  • Current workaround — how they solve the problem now, including "nothing" or "a spreadsheet" — this shows what a new solution is actually competing against.
  • Where they can be reached — the channels, communities, or moments where this segment can realistically be found and messaged, useful to both product and marketing.

Grounding in real data, not assumptions

A persona is only useful if it's built from actual research: customer interviews, support tickets, sales call notes, usage data, surveys. The difference between a persona and a demographic stereotype is evidence — "Maria, 34, marketing manager" with no source behind the details is a guess wearing a persona's format. A persona built from real data can cite where each trait came from, and gets revised when new data contradicts it. If nobody can point to where a claim on the persona came from, treat that claim as an assumption to go verify, not a fact to design around.

Common pitfalls

  • Built from internal assumptions, not customer data — a persona assembled from what the team already believes about its customers confirms existing bias instead of testing it, and tends to describe an idealized customer rather than a real one.
  • Too many personas to actually use — a dozen finely differentiated personas sounds thorough but nobody can hold them all in mind when making a decision; a handful of clearly distinct personas beats many overlapping ones.
  • Created once and never revisited — the customer base shifts as the product, market, and pricing change; a persona frozen at launch stops matching who's actually buying or using the product a year later.
  • Treated as a demographic checklist — age, job title, and income without goals or pain points gives a team a label, not the insight needed to make a product or messaging decision.
  • Used to justify a decision already made — cherry-picking whichever persona detail supports a preferred outcome defeats the point of grounding decisions in real customer evidence.

Learn more

View customer-persona/SKILL.md on GitHub