How to create a user persona: step by step for beginners

Step 1

Decide which product and decisions the persona is for

As Nielsen Norman Group defines it, a persona is a fictional, yet realistic, description of a typical or target user of a product. Its job is to help a team understand the people it designs for and avoid designing "for everyone". So first write down which product or feature the persona is for and which decisions it should help make. For a practice exercise, take a familiar situation: an app for booking sports classes, a home rental service, a library with an online catalog.

• •write the scenario: "people looking for a workout near home for the first time";
• •write down 2–3 decisions the persona will affect: what's on the first screen, which filters, how to send reminders;
• •one exercise, one or two personas. More at the start will only blur the focus.

Step 2

Gather data about real people

NN/g stresses that personas must be based on user research. A persona may be a made-up person, but the information in it should be about real people. Among the methods NN/g lists are interviews, surveys, field studies and longitudinal studies. For a study project, the simplest option is a few interviews with people who have experience of your topic.

• •ask about past experience: "Tell me about the last time you looked for a place to exercise";
• •find out goals, circumstances, habits, workarounds and annoyances;
• •write down exact quotes and behavior, not just answers to direct questions;
• •note where people answer similarly and where differently: that's where different personas come from.

💡 If you have no data and can't collect it yet, don't pass guesses off as research: call such a persona a working hypothesis and write down what needs checking.

Step 3

Group observations and find types of behavior

NN/g advises starting persona creation with user characteristics observed in research and grouping them: separate "characters" gradually emerge from the clusters. A handy tool is an affinity diagram: one observation per sticky note, then grouping similar ones. Another intermediate step is an empathy map, where you lay out what a person says, thinks, does and feels.

• •group by behavior and goals, not by age or gender: "plans everything ahead" versus "decides at the last minute";
• •if two groups are too similar, merge them; if a group doesn't matter for the product, drop it;
• •keep the groups that affect the design in different ways.

Step 4

Describe the persona — only what helps the design

Once the types of behavior are clear, turn each one into a persona. NN/g lists common parts: a name, age, gender and a photo, a short tagline, experience with the product or service, context of use, goals and concerns during key tasks, and a quote that sums up their attitude. The name and photo are there to make the persona memorable. Every other detail should pass NN/g's simple test: if it won't affect the design or help make a decision, remove it.

• •context: which device, how often, by choice or because of work;
• •goals: what the person wants to achieve and what "success" means to them;
• •concerns and barriers: what stops them, what they doubt;
• •a quote from an interview: a real one is better than one you invented;
• •don't add a favorite coffee or star sign if they have nothing to do with the product.

💡 Use a photo you have the rights to, and don't use a photo of an interview participant: a persona is a composite, not a real person.

Step 5

Use the persona and check it

A persona works when people use it. NN/g recommends creating personas as a team, because then people believe in them and use them more. Take the persona into your study project: walk through the key scenarios through its eyes, use it as a guide for recruiting usability test participants, check whether your decisions match its goals.

• •write 2–3 scenarios: "Olena wants to find a workout nearby in five minutes because she's on her lunch break";
• •check every design decision: does it help the persona or get in the way;
• •update the persona when new data contradicts it;
• •for a case study, show the path: data → groups → persona → decisions.

💡 Show the persona in the community and ask: "Is it clear how it affects the design?" If the answer is no, the description has too much decoration and too few goals.