Skip to content

A 5 Step Framework for Building Better Personas

User personas get a bad rap—and sometimes for good reason. Here's a great system for making ones you'll love. 

Words by Nikki Anderson, Visuals by Emma McKhann

Ask anyone about user personas, and they’re going to give you one of two answers:

  1. They love them!
  2. They hate them with the white-hot passion of an exploding star!

There is no in-between.

I have a soft spot for personas. Through the years, I have watched my personas evolve beyond what I could have imagined.

Just like any deliverable from user research, we need to take personas with a grain of salt. We shouldn’t place all of our expectations, hopes, and dreams on them.

If we give our personas too much power, they will fail us. They don't have all the answers. They can never live up to such high expectations. I think this divide in expectations has been the biggest problem with personas.

We have forgotten that personas are a tool. The reason I create them is to help my teammates understand the story and the relevant context of our users. 

I believe, if we are mindful of how we use them, personas can help guide teams towards better decision-making and innovation.

Great personas rely on great generative research. We wrote a substantive, start-to-finish guide to conducting generative research that you might want to give a read. 

Personas are representations of your users and are a combination of needs, desires, and pain points of these groups.

Nikki Anderson

Wait a second, what are personas?

In the world of research, there are many types of personas. This variety can be confusing for someone starting in the field of user research, or even for someone more senior who is looking to try something new.

I have briefly outlined (and dumbed-down a bit) some of the persona types you can find in the wild:

  • Proto-personas: A proto-persona is based entirely on assumptions by internal stakeholders. These are then later checked and validated against actual data (qualitative interviews, analytics). They are a starting point if you have done absolutely no data.
  • Market personas: A market persona represents a segment of your audience and usually based on demographics (income, gender, age).
  • Archetypes: Archetypes are representations of groups of users modeled around behaviors or tasks. They often include much more quantitative user data, such as common behaviors in Google Analytics.
  • Jobs to Be Done (JTBD) personas: JTBD personas include more information on motivations, triggers to actions, and can be more goal-oriented. They often include the job a person is trying to accomplish and how their behaviors are related to that goal.
  • (Traditional) Personas: Personas are representations of your users and are a combination of needs, desires, and pain points of these groups.

Now the fun part: someone could take JTBD personas and put them together with archetypes, while also including demographics from market personas. There is no real standard persona anymore. You create the persona based on the goal you are trying to achieve.

How I create personas

To make this as easy to digest as possible, I will go through an example of how I would create a persona. This process is, of course, not representative of how everyone should create personas, but is my take.

Step 0: Do generative research

You have to do generative user research to create a persona. The only exception to this rule is the proto-persona mentioned above, based on assumptions and internal data. However, this persona should be used as a very basic starting point, and should not be a tool to make decisions about how a product can/should move forward.

Step 1: Define the goal of the persona

What is it that you are trying to accomplish with your persona? There are several different goals you can have when creating a persona. Below are the various reasons I've created them:

  • Show colleagues who our customers are
  • Highlight the behaviors of user groups
  • Present high-level motivations of users that cause them to visit a platform
  • Tell a story of how someone uses a platform

Step 2: Decide what information to include

There are many different recommendations on what you should include in a persona. Ask yourself: if I had never spoken to the user, what would I need to know to understand them better? Again, this does depend heavily on the goal above. In general, the different points of information you can include are:

  • Needs
  • Desires
  • Pain points
  • Anxieties
  • Hurdles
  • Goals
  • Motivations
  • Tasks
  • Current usage (of platforms or brands)
  • Fears
  • A story

Step 3: Pull together information from the research

Take the time to synthesize this information and find patterns and trends from the qualitative interviewing you did. When you define the information you are looking for, it makes it easier for you to comb through the data and find it. (PS: if you're looking for some straightforward synthesis tactics, you might like our "on-any-timeline" guide to qual research analysis). 

Remember: don’t make anything up. You want your personas to be based on real user data.

Nikki Anderson

Step 4: Outline the persona

Create version one of the persona. Include all the information you think might be relevant. Again, we are trying to help teams understand motivations of why someone might use a product, and also tell a story.

Step 5: Design the persona

Putting this information into a visual format is the last step. Try not to do this before you have all the information together. At this point, a common mistake I've come across involves dedicating too much with the graphics. 

For example, the bars that indicate people's personality traits on a spectrum can do more harm than good.They aren't beneficial at all to a team making a decision. Keep the persona simple.

Also, I'd discourage blindly using one of the many templates available. Although some may be beautiful, they will never serve you as well as tailoring your persona builds to your team's specific needs. Goals should precede visuals. 

Another common mistake: don't make anything up. Delete all false or dramatized information. It will only get in your way when making decisions and could lead the team astray. You want your personas to be based on real user data.

For example:

We have an app that helps people find and adopt pets called Pet Buddy. We are trying to figure out the different groups of people who are interested in finding adoptable pets online.

Step 0: Do the research. I am assuming that I have already done all the necessary generative research for this persona project. All of the data below is for a project I did a very long time ago.

Step 1: Set a goal. My goal for the persona: I want to be able to present high-level motivations of people who would use our app and the story of how users interact with the app.

Step 2: Decide on the information. Since I want to convey motivations and I story, I will pick information based on that. For this, I would include:

  • Goals: what people are trying to achieve
  • Motivations: why they are trying to achieve those goals
  • Needs: what will help them achieve those goals
  • Fears: what people want to avoid when trying to achieve their goals
  • A story talking giving context about the user

Notice how these different subsections of information don't necessarily have anything to do with our app, PetBuddy. We don't only want to include information that has to do with our app. We also want to include information agnostic to it. If we only include user's goals, needs or fears when it comes to our app, we will miss opportunities for innovation.

Step 3: Pull information together. For this exercise, I would comb through the different transcripts and find patterns that lie within the information above. An easy way to do this is through affinity diagramming.

Step 4: Outline the persona. Here I would create version one of the persona, without thinking of any visuals. So, here is what it might look like:


- Find a pet that is perfect for the current situation

- Find a reputable shelter to trust

- Find a companion


- Get a pet for self

- Adopt a pet for family (kids)

- Save a life

- Find a special needs pet


- Invest in the right pet

- Finding a healthy pet

- Reviews or referrals from others to trust the shelter

- Paperwork from the shelter, including vaccines and history

- Understanding of how a pet would interact in a particular environment

- How the pet interacts with kids, other pets, men/women


- Going to an untrustworthy shelter

- Getting conned by someone

- Adopting an unhealthy pet

- Adopting the wrong pet that doesn't work with current circumstances

- Hidden charges or information

- Going around to many different shelters with no luck


Amanda is interested in adopting a pet. She has looked for some time but isn't getting great information on the different pets available when she looks online. She wants to save a life, but also to adopt a pet that she'll adequately be able to take care of. She lives in an apartment in Boston and works at a 9-5 job.

Since Amanda is a runner, she would like to find a dog that can run with her but can be comfortable staying at home while she is at work. She also travels frequently home by car or plane, so she is looking for an animal that can travel well.

Amanda has had dogs in the past, but they were all from breeders. She wants to be able to adopt but is fearful that she won't find the perfect dog, especially if that dog is older. She gets stressed going to shelters to look at the dogs and doesn't believe she gets enough information or help from the staff.

Step 5: Design the persona. Now that I have the majority of information I need, I would start to rearrange it in a way that makes sense. Usually, for me, motivations are the most high-level and go on top.

How to make personas even better

Here are four solid tips to take your personas to the next level:

  1. Have a mission for your personas. I know I mentioned this above, but it is incredibly essential that you know the purpose you want your personas to serve. If they don't have a goal or a reason for existing, they will become a jumble of information that is difficult to use.
  2. Stay away from demographics (in most circumstances). This includes demographics can cause us to attribute traits to a whole group of people incorrectly. Personas are meant to target and help us make decisions based on people's behaviors and motivations—not whether or not they are married.
  3. Tell a story. Give context to the person. We don't live and use products in a vacuum. Include some background information on how that person operates in their day-to-day life. A story can also give additional clues or an understanding of how a person might interact with a product.
  4. Take away irrelevant information. Deleting this information can be the most challenging exercise for me. I often want my personas to be super humanized. That means I will sometimes include superfluous information like hobbies. If this is relevant to the story you are telling or product you are trying to sell, include it, but usually, it isn't helpful in excess.

We have limited space on a persona and a small window of opportunity to prove the value of them to others. We want to include the most engaging and relevant material that helps teams make better decisions. Personas are a tool we can all use to help guide innovation and better decision-making. Every piece of the persona should speak to that. Moving forward, we can let go of the standard version of our persona and move towards the purpose they were meant to serve.

Nikki Anderson-Stanier is the founder of User Research Academy and a qualitative researcher with 9 years in the field. She loves solving human problems and petting all the dogs. 

To get even more UXR nuggets, check out her user research membershipfollow her on LinkedIn, or subscribe to her Substack.

Subscribe To People Nerds

A weekly roundup of interviews, pro tips and original research designed for people who are interested in people

The Latest