Skip to content

User Persona

A research-based profile of a target user type — their goals, context, and constraints — used to keep design decisions grounded in real people.

Deliverables & Artifacts

A user persona is a research-based profile of a target user: what they are trying to do, the context they work in, the constraints they are under, and what they currently use instead of you. Its purpose is to make design decisions about a specific person rather than about an average.

The format has earned a poor reputation, and mostly it deserves it. A card with a stock photograph, an invented name and a note about the person's commute settles no arguments and survives on a wall until somebody takes it down. The discipline underneath, of naming who you are designing for and being able to defend it, is what actually does the work.

What belongs on one, and what does not

IncludeWhyTypical fillerWhy not
The job they are doingEvery design decision serves a jobAge and locationRarely changes a decision
What success looks like to themLets you judge a design against itA stock photographImplies a precision you do not have
Constraints: time, permissions, skillsExplains why the obvious answer failsFavourite appsInteresting, inert
Expertise with this kind of toolDecides how much explanation is neededA personality descriptionUnfalsifiable
What they use todayThe real competition, usually a spreadsheetAn invented quoteFabricated evidence reads as evidence

The test for any line is whether it could change a decision. If two designs are indistinguishable under it, it is decoration.

B2B personas are roles, not people

Consumer personas describe types of person. B2B products are used by people occupying roles inside an organisation, and the role predicts behaviour far better than the individual does.

A useful B2B set usually separates three or four:

The economic buyer approves the spend, may never log in, and needs to justify the decision to somebody else. They are a persona for the marketing site and the sales conversation, not for the product.

The champion brings the product in and stakes credibility on it. Their goal is partly professional advancement, and their worst moment is defending a tool that has embarrassed them.

The admin configures it: permissions, integrations, data. They meet the longest unaided stretch of the product and the most technical parts of it.

The daily user was assigned this tool. They did not choose it, may resent it, and judge it against how long the task used to take. Products that delight buyers and irritate daily users churn at renewal.

Designing for the champion and forgetting the daily user is the most common failure, because the champion is the one your team talks to.

Building one that is defensible

Start from interviews, not from a workshop. Six to eight conversations per role is enough to see the pattern. Personas assembled in a room describe the room.

Look for behavioural clusters, not demographic ones. People who evaluate before committing and people who dive in behave differently and need different onboarding. Their job titles may be identical.

Write what you do not know. A persona with an explicit gap is more useful than one that quietly guesses, because the gap can be filled later and the guess cannot be spotted.

Date it and name the evidence. A persona built on interviews from two years ago describes a market that has moved.

Alternatives worth knowing

Several teams have abandoned personas for formats that carry the same information with less invitation to invent.

Jobs to be done describes the situation and the outcome rather than the person: when I finish a month-end close, I want to know nothing was missed, so I can sign off without rechecking. It resists demographic filler by construction, and it travels well between teams.

Behavioural segments group people by what they do in the product rather than by who they are. Accounts that invite a second user in week one behave differently from accounts that do not, and that split is measurable rather than asserted.

A written paragraph with the evidence attached. Unglamorous, hard to argue with, and considerably more likely to be read than a card.

The format matters less than two properties: it must come from evidence, and it must be able to tell you that a design is wrong. Anything meeting both is doing the job, whatever it is called.

In practice

A team has three personas on the wall and a roadmap that serves none of them in particular.

Rebuilding them from interviews changes one thing decisively. The team had been designing for an experienced operations lead who wanted control and configurability. Interviews find that most accounts are set up by that person and then used daily by a junior colleague who was given no training and no explanation of why the tool arrived.

Nothing about the buyer persona was wrong. It was simply not the persona that determined whether the account renewed, and the roadmap had been built entirely from it.

The consequence is a shift in what gets prioritised: defaults that work without configuration, explanation at the point of use, and an onboarding path for somebody who did not choose to be there.

Where teams get it wrong

  • Inventing them. A persona without research is the team's assumptions with a name attached.
  • Too many. Seven personas means designing for whichever is convenient.
  • Demographics doing the work. Age and location rarely change an interface.
  • Designing for the buyer. The buyer decides you get in; the daily user decides you stay.
  • Never revisiting them. Markets move, and a persona has no way of telling you it has gone stale.

Frequently Asked Questions