How to create a user persona
A practical, research-first walkthrough for making a user persona your team will trust, with a free tool and an example to copy.
A user persona is a one-page portrait of a person your product has to work for. It has a name, a face, a few facts, and the things that matter most: what this person is trying to do, what gets in their way, and when your product shows up in their life. Teams make personas so that "the user" stops being an abstraction in meetings and becomes someone specific you can argue about.
The persona itself is easy to make. You can fill one in with the free persona generator in ten minutes. The hard part, and the part that decides whether anyone trusts it, is where the content comes from. A persona is only as good as the research under it. This guide walks through both: the research, and then the page.
Start with the decision, not the template
Before you open a template, write down the decision the persona needs to help with. "Should the onboarding ask for a phone number?" "Which of the three dashboard ideas do we build first?" "Who are we writing the help pages for?"
A persona that exists to answer a question is sharp. A persona made because the project plan has a box labeled "personas" is a poster. If you cannot name a decision, you may not need a persona yet, and that is a fine thing to discover early.
Talk to real people
The content of a persona comes from research. The cheapest research that still counts is a handful of interviews: three to five conversations of about thirty minutes for each type of customer or visitor, with people who do the thing your product supports. Watch them work if you can. Ask what they did last time, not what they would do in theory. (Our guide to persona interview questions has a script you can take into the room.)
If you have other sources, use them. Support tickets, sales call notes, survey free-text answers and analytics all count, as long as you treat them as evidence about specific people rather than averages. An average user does not exist; a persona describes someone who could.
Keep the raw material. Quotes, screenshots, the surprising moment in interview three. You will need them for the "Based on" line at the bottom of the page.
Find the pattern
After the interviews, lay out what you heard. Index cards, a whiteboard, a spreadsheet: anything that lets you move things around. Look for goals that repeat, frustrations that repeat, and situations that repeat. Usually two or three clusters appear. Each cluster that would make you design differently is a candidate persona.
Resist the urge to make one persona per interviewee, and resist the urge to make one persona that covers everyone. If two groups want the same things and are blocked by the same things, they are one persona even if their job titles differ. If one group works on a phone in a truck and the other at a desk with two monitors, they are probably two.
Fill in the page
Now open the tool and pick a type. The user persona template starts you with a summary, goals, frustrations, scenarios and a comfort-with-technology scale. Change anything that does not fit; the sections are a starting point, not a form.
A few rules that keep the page honest:
- Name and role first. A name people will remember and a role in plain words: "Night shift lead, fulfillment warehouse," not "Operations stakeholder."
- One quote, in their words. Take it from an interview. A real sentence does more than a paragraph of description.
- Goals are outcomes, not features. "Finish charting before the end of the shift" is a goal. "Wants a mobile app" is a feature request wearing a goal's clothes.
- Frustrations are specific. "Session timeouts that log her out mid-note" tells a designer what to fix. "Frustrated by technology" tells them nothing.
- Scenarios are moments. When, where and under what pressure does this person meet your product? Three scenarios will shape more design decisions than any other section.
- Leave out what you did not learn. An empty field is honest. A guessed field is a liability, because someone will design for it.
Look at Denise Walker, the emergency room nurse in our gallery for a sense of the texture: the scenarios are concrete, the frustrations name the software's actual behavior, and nothing on the page is there for decoration.

Denise Walker
Emergency room nurse
“If it takes more than three taps while I'm standing, I'll do it later. Later is where mistakes happen.”
- Age
- 41
- Location
- Baltimore, Maryland
- Occupation
- Charge nurse, Level II trauma center
- Education
- BSN, 17 years in emergency care
Summary
Denise runs the floor on three twelve-hour shifts a week and covers charge duties on two of them. She charts on a workstation on wheels between rooms, usually with a phone ringing and a patient asking for water. She has outlasted four electronic record systems and trusts none of them.
Goals
- Finish charting before the end of the shift, not after it
- See a patient's allergies and last meds without opening three tabs
- Hand off to the next shift with nothing lost in the gap
Frustrations
- Alerts that fire for everything, so the real ones get ignored
- Session timeouts that log her out mid-note
- Training videos recorded for office staff, not for someone standing at a bedside
Scenarios
- A trauma arrives while she is two notes behind; she needs to park her work and come back to it exactly where it was
- A patient's family asks what medication was given; she wants the answer on one screen
- Shift change at 7 am: she reviews fourteen patients in ten minutes
Comfort with technology
Based on
Illustrative composite. Drawn from published studies of clinician documentation burden and from patterns Mobility Labs has seen in healthcare projects. Not a research persona: nobody was interviewed for this specific profile.
Write the "Based on" line
Every persona template in the tool ends with a section called "Based on." Fill it in before you share the page. "Four interviews with floor nurses at two hospitals, March 2026, plus 140 support tickets tagged 'charting'." That one sentence does three things: it tells readers how much to trust the persona, it tells you when it needs refreshing, and it stops the persona from being quoted as fact two years later by someone who never saw the research.
If the persona is an assumption, say so. "Based on: the team's current assumptions, to be tested in interviews next month" is a legitimate thing to write. It makes a proto-persona, which is useful, as long as nobody mistakes it for the researched kind.
Share it where decisions happen
A persona in a slide deck gets looked at once. A persona in the design review, the ticket, the pull request description, the kickoff doc gets used. The tool gives every persona a link you can paste anywhere and a PNG or SVG you can drop into Figma, Miro or a slide. Put it where the arguments are.
Then use it in sentences. "Would Denise do this between patients?" is a design critique. "Does this meet user needs?" is not.
Keep it alive
Personas go stale. Products change, users change, and the person you interviewed in March has a new job by October. Put a date on the page, and when the "Based on" line is more than a year old, go talk to people again. Updating a persona after a round of research is a sign the team is learning. A persona that has not changed in three years is a sign nobody is reading it.
Make one now
Pick a type, fill in what you know, leave blank what you do not, and write the "Based on" line last. Make a persona, free, no sign-up. If the research is the part you need help with, that is the work Mobility Labs does.