The first 3 interviews are free · Start free or see the prices

Persona Synthesis Template: Build Personas from Interviews

Use this template to turn interview data into personas that are based on behavior. You place participants on behavioral variables, find the clusters, then fill in 9 elements for each persona.

Analyze a transcript free

Use it for

  • Transcripts or notes from 10 to 25 interviews in one domain, such as travel planning or team budgets.
  • Basic facts about each participant: role, company size or life stage, and the tools they use.
  • The decisions the personas must support, for example which features to build first or which market to enter.

Steps

  1. List the behavioral variables

    Read 5 transcripts and list the ways in which participants behave differently. Write each variable as a range with two ends, such as "plans ahead" to "decides on the day". Aim for 6 to 10 variables.

  2. Place each participant on each variable

    Make a grid with variables as rows and participants as columns. Mark where each participant sits on each range, based on what they did, not what they said they prefer.

  3. Find the clusters

    Look for participants who sit together on several variables at once. A cluster across 3 or more variables is a candidate persona. Most studies give 2 to 5 clusters.

  4. Check that each cluster is distinct

    Two clusters must differ on at least one variable that changes a design decision. If they differ only in demographics, merge them.

  5. Fill in the persona elements

    For each cluster, fill in the elements in this template with data from the participants in it. Every statement must trace back to at least two participants.

  6. Name and test the personas

    Name each persona by its main behavior, such as "The weekend planner". Then ask the team to sort 5 new transcripts into the personas. If the team cannot agree, the personas are not distinct.

Codebook

Copy the codebook as a table. It pastes into a spreadsheet with one row for each code.

Defining behavior

The pattern of behavior that separates this persona from the others.

Include
Write the 2 or 3 positions on the behavioral variables that all members of the cluster share.
Exclude
Do not write age, gender, job title or income here. Put those in Context only if they explain the behavior.
Example
“I do the whole month's shopping in one go on the first Saturday, with a list, and I do not go back.”

Goal

What the persona wants to achieve in the domain, in their own terms.

Include
Write 1 to 3 outcomes that members of the cluster stated or showed through their actions.
Exclude
Do not write a product feature as a goal. "Use an app with reminders" is a solution, not a goal.
Example
“I want to never run out of the basics in the middle of the week.”

Context

The situation in which the persona does the task: place, time, devices, people around them and constraints.

Include
Write the conditions that shape the behavior, such as a shared household, a commute or a fixed budget.
Exclude
Do not write events that cause a change. Put those in Trigger.
Example
“I shop with two kids in the car, so I have about 40 minutes and no patience for queues.”

Trigger

The events that make the persona start the task or change how they do it.

Include
Write the specific moments that start the behavior, such as payday, an empty fridge or a guest visit.
Exclude
Do not write ongoing conditions. Put those in Context.
Example
“As soon as my salary comes in on the 1st, I open the list and start.”

Current tools and methods

The products, services and manual methods the persona uses for the task today.

Include
Write the tools that cluster members use and what each one does for them.
Exclude
Do not write what is wrong with the tools. Put that in Frustrations.
Example
“A paper list on the fridge, then the store's app for the delivery slot.”

Frustrations

The recurring problems the persona has with the task or with current tools.

Include
Write problems that two or more cluster members described, with the cost to them.
Exclude
Do not write single complaints from one participant. Keep those in the raw notes.
Example
“Half the time something is out of stock and the substitute is useless.”

Decision drivers

The factors the persona uses when they choose between options.

Include
Write what the persona compares and what makes them pick one option, in order of weight.
Exclude
Do not write factors that never changed a choice. Only include drivers linked to a real decision.
Example
“I pick the slot first, then the store. Price matters less than getting it on Saturday morning.”

Expertise

How much the persona knows about the domain and the tools.

Include
Write the level of skill the cluster showed, such as workarounds they built or terms they use.
Exclude
Do not rate expertise from the participant's own claim alone. Use what they did and how they explained it.
Example
“I have a spreadsheet that tells me which store is cheapest for each item this month.”

Representative quote

One real quote that captures the persona's defining behavior in their own words.

Include
Choose a quote from a cluster member that a reader would remember and that matches the Defining behavior.
Exclude
Do not write a quote yourself or merge two quotes. Do not pick a quote only because it is funny.
Example
“If it is not on the list, it does not exist.”

Report outline

  1. The method: number of interviews, the behavioral variables and how participants were clustered.
  2. The grid of participants on variables, so readers can see the clusters.
  3. One page per persona with all 9 elements and the participant IDs behind them.
  4. The differences between personas that change a design decision.
  5. Which persona to design for first, and why.
  6. Limits: groups that were missing from the sample and variables that need more data.

Mistakes to avoid

  • Do not start from demographics. Personas built on age or job title do not tell designers what to build.
  • Do not invent details to make the persona feel real. A stock photo and a hobby that no participant mentioned make the persona less trusted.
  • Do not make one persona per participant. A persona is a pattern across several people.
  • Do not keep personas after the market changes. Check them against new interviews at least once a year.

Guides that use it

Questions

Most persona studies need 12 to 20 interviews. You need enough people to see clusters. If a cluster has only one person, interview more people like them.

A segment groups people by facts you can measure, such as company size. A persona groups people by behavior and goals. A persona can cross several segments.

Most products need 2 to 5. Each persona must lead to a different design decision. If two personas would get the same product, merge them.

Yes. The Board tab can group a study's clips by persona, with one lane per person. You can rename and merge lanes as you build clusters.

Ask real people your question

The AI moderator interviews screened people by voice or text. Each claim in the study report links to its clip. The first 3 interviews are free.

Start free