--- title: "Persona" description: "A research-based representation of a group of people with shared goals, behaviour or needs." category: "Research and evidence" canonical: "https://contentlexicon.com/persona" updated: "2026-10-15" --- # Persona _Research and evidence_ > A research-based representation of a group of people with shared goals, behaviour or needs. A persona turns research findings into a character: one named, described person who stands in for many real ones. It usually sets out their context and motivation, the tasks they're trying to complete and what gets in their way. The fiction gives a team someone specific to design and write for, instead of an abstract "user" who can be bent to fit any argument. A persona isn't one research participant, a marketing segment or a demographic sketch. Names, photographs and implausibly specific coffee orders add memorability, not validity; a persona's value comes from the evidence underneath and how directly it informs decisions. Personas entered software design through Alan Cooper, who role-played a user drawn from his interviews while designing a project management tool in the early 1980s and later set out the method in The Inmates Are Running the Asylum, though marketers had been describing archetypal customers for decades before. ## Why it matters for content Personas help [content teams](https://contentlexicon.com/content-team) see that different groups need different things from the same page. A first-time applicant needs reassurance and plain steps; a returning professional wants the reference table and nothing else. Named, evidence-based characters make those differences easier to discuss and prioritise than "our audience" or the preferences of whoever is in the meeting. The details that earn their place are the ones that change content decisions: what someone already knows, what they're trying to get done, what's at stake for them, what device and setting they're in and what might stop them. Age, job title or family situation belong only when they affect those things. Personas also need maintaining. Built once and pinned to a wall, they harden into polite stereotypes and hide the people who don't fit the dominant pattern. Some researchers argue for dropping the character altogether. Indi Young's thinking styles, for example, group people by how they reason towards one goal, with no demographics, on the grounds that demographics rarely explain behaviour. For content work that lens is often more useful, because it ties directly to the decision someone is making on the page. ## How AI changes the picture Generative AI can produce a persona, complete with backstory and stock photo, in seconds, and that ease is the risk. A model asked for "a typical small business owner" returns a plausible average of everything it has read, which is fiction dressed up as insight. Synthetic users like this can help brainstorm questions to ask in research, but they aren't evidence, and treating them as evidence lets invented needs steer real decisions. AI is more useful with the research itself: clustering interview notes, tagging support tickets by goal or pulling recurring frustrations out of hundreds of survey responses. A researcher can then build personas from patterns they've checked against the source. Personas can also become test [prompts](https://contentlexicon.com/prompt), such as asking a model to read a page as a nervous first-time applicant. That can surface gaps, provided everyone remembers it's simulating the persona, not a person. ## Questions people ask **How many personas do we need?** Enough to cover differences that change a content decision, and usually fewer than teams expect. A small service might need only two if the main contrast is between occasional applicants and the staff processing their applications. Add another when research reveals a need the existing ones miss. **Should personas include demographics?** Only where they affect the task or experience. Someone's age might matter for a pension page and be irrelevant to a parking permit. Behaviour, context and existing knowledge usually tell a [writer](https://contentlexicon.com/writer) more than a biography. **What's the difference between a persona and an audience segment?** A segment is a grouping, often defined by shared characteristics for targeting or analysis. A persona takes a pattern from research and gives it a face and a story, so it's easier to keep in mind while writing. The two can coexist, with one segment represented by one or more personas. **Are personas still worth making?** They are when they're grounded in research, used in real decisions and updated as behaviour changes. They're much less useful when they exist mainly as a deliverable. If nobody can point to a decision a persona has changed, it isn't doing its job. ## Learn more - [The origin of personas](https://web.archive.org/web/2010/http://www.cooper.com/journal/2008/05/the_origin_of_personas.html) – Alan Cooper's account, via the Internet Archive, of improvising his first persona while designing software in 1983 and later formalising the method in client projects. - [7 Core Ideas about Personas and The User Experience](https://measuringu.com/personas-ux/) – Jeff Sauro's overview of what personas are for, including their earlier roots in marketing and the work of Angus Jenkinson alongside Cooper's. - [Explanations – Thinking Styles](https://indiyoung.com/explanations-thinking-styles/) – Indi Young on grouping people by how they think about one goal, free of demographics, as an alternative to conventional personas. Related terms: [Audience segmentation](https://contentlexicon.com/audience-segmentation), [Content brief](https://contentlexicon.com/content-brief), [Core content strategy](https://contentlexicon.com/core-content-strategy), [Tone](https://contentlexicon.com/tone), [User needs](https://contentlexicon.com/user-needs) This page was last reviewed in October 2026.