It’s Time to Transition to Inclusive Personas

I recently shared a meme on LinkedIn with a message about that idea being a key reason why we redesigned our persona templates. It got more traction than almost anything else I’ve shared, which inspired me to write about inclusive personas.

“Inclusive personas” has been a popular phrase in the space recently, but one might wonder: “What exactly are they?”, “Why are they important?” or, “How do we effectively update our persona templates?”

Outdated personas can be laden with details that have the potential to cause bias and exclusion. We can make them more inclusive by focusing on user thoughts, needs, wants, and behavior, rather than demographics.

While I realize some peers in the industry advocate doing away with personas altogether for various reasons, I have found personas to be effective ways for large teams with frequent turnover to maintain focus and empathy for an archetype(s) of the users of our products and to prevent the problem of trying to solve for either everyone or just for ourselves.

A brief history of personas

It’s hard to imagine people designing products for people without personas, but actually, they’ve only been ubiquitous for a couple of decades—when all things usability became a normal part of the product development cycle. That’s not to say that no one ever thought of who they were designing for before, but the “persona” as an artifact was not a thing.

I looked back in one of the earlier books I bought on user experience—User and Task Analysis for Interface Design by JoAnn Hackos and Janice Redish from 1998—and the only reference to defining users was to think about primary, secondary, and surrogate users and to plot them on a matrix that looked like this, only going so far as to describe them with one or two nouns:

Tasks likely to be performed

Software engineer-turned-usability guru, Alan Cooper, championed the use of personas in the late 1990s, publishing a book called The Inmates Are Running the Asylum in which he wrote for the first time about his invention, personas.

In 2006, John Pruitt and Tamara Adlin published an excellent tome called The Persona Lifecyle, whose guidelines I’ve followed in numerous career engagements.

With few exceptions, persona artifacts have included a photo of the archetype user they intend to represent. Some have leaned more heavily toward providing demographic data points, especially characteristics such as:

• Age
• Gender
• Education
• Race (not typically explicitly stated, but implied via image)
• Income
• Occupation
• Location
• Marital status

Not only that, but quite often the picture takes up more space on the artifact than anything else. If you do a Google image search on “personas,” you’ll see countless examples of this pattern.

This was not a bad instinct—the idea was to try to avoid the natural urge to think about ourselves when designing (particularly in formerly development-driven companies where creating hard-to-use software that made sense to developers but not anyone else was the norm), and instead putting the focus on the users of our products.

We also know that people’s eyes are drawn to faces, particularly to those they consider attractive, and that adding human photos to all kinds of content has been a standard since we started taking photographs.

This background is intrinsic to how we make relationships with people in real life—we meet them, we see what they look like, and we learn their names and other details about them. Once we spend enough time with them, we know their names and faces by memory.

These qualities of human images have been helpful in cementing personas as part of the design process in companies worldwide. I’ve worked at several places where we start every project talking about Jill or Ken or Noah or Sara and what that persona needs, augmented by the personas being plastered around our workplaces. One company even created full-size cardboard images of their target persona and brought the artifact to project meetings to highlight the intention of being laser-focused on the user.

So, what happened to cause us to rethink the way we created personas?

Enter: Inclusive personas

In 2020, most people in the U.S. knew that racism and the various ways it has shaped people’s perspectives was a hot topic. Where I worked at the time, the User Experience Research team played a major role in developing personas, and one of my fellow researchers was working on some new persona requests.

As she began to explore how best to create these personas, she found articles and books explaining how some of the content typically included in personas can lead to assumptions and implicit biases that really don’t need to be there for personas to serve their purpose.

This enterprising and engaged researcher compiled a list of references for resources focused on the representation of age, race, gender, and ability in personas, as well as materials about design justice and diverse representation in design.

Using that to formulate a plan for how to update our persona approach, she brought the proposal to management, who shared the idea with the rest of the team and got others interested in working as a team to bring the plan to fruition.

The team’s approach was heavily influenced by the work of Indi Young, freelance researcher and co-founder of Adaptive Path. She recommends basing persona-type work on “thinking styles,” which focus more on behavioral audience segments that represent patterns of thinking, emotional reactions, and guiding principles for different archetypes.

As in many initiatives, timing is critical to success, and it was the right time and place for the message to be heard and well-received. The working team ran workshops and iterated on updating the persona template. They quickly evaluated the company’s four core personas, removed the content they deemed to have potential for introducing bias, and began sharing the updates in key circles.

The concept resonated well with leadership, who provided a venue to share at a monthly meeting attended by all staff involved in User Experience and many in digital product development. They also provided an invitation to regularly scheduled diversity and inclusion events.

What’s wrong with pictures and other demographics?

Some may object to this approach, wondering why the change is needed when considering the impact of re-thinking established processes and socializing new content. Many people have put a lot of effort into integrating personas into their team’s work efforts, to the point that everyone on the team knows the personas’ names and can describe them.

So, do we really need to do this? My answer is yes, because of two primary pillars of user experience and related research: empathy and attempting to remove bias.

A primary reason why we create personas is so that we have empathy for the users of our products. Getting to know details about a person allows us to relate to them and to understand more about their wants and needs. However, when we provide information like race, gender, and marital status, our brains tend to draw conclusions based on prior experiences we've had that may cause us to not think of users in objective terms.

One example that I particularly liked was provided by Indi Young in the article above. In Twitter conversations about these kinds of biases, she referenced someone who used a persona with a photo of a young blonde-haired woman, and it was reported that the persona would get dismissed as “The Blonde” (replete with eye-rolls in my imagination).

Ms. Young provided another enlightening example from a misguided attempt to get away from people’s names, but still causing bias by replacing a name with an adjective. In a team she worked with that focused on frequent flyers, one persona was originally titled “The Grumbler.” However, she found that the name caused a team member to feel negatively toward the persona, calling them “a person who has a maladjusted attitude toward life.”

Young led them to change the name to something that group members would be proud to use to describe themselves in the context of a travel day, and they landed on “The Frustrated.”

In the financial services industry, it can feel natural to segment existing and prospective customers based on the amount of money they have invested or are able to invest because that is a minimum requirement for many investment products.

I even once thought that I cleverly named a secondary persona “The One-Percenters,” since that was a very common phrase that year, but when I think about the biases that can cause, I want to erase that version from document history. Similarly, introducing a phrase that one person might think would encourage empathy, like “low income,” allows people to infer different meanings from the words.

As researchers, we always strive to reduce the amount of bias that can affect our studies, from cognitive to experimentation to exclusionary biases, so anything that can help us remove any potential bias effects will make our research that much better.

That’s why I prioritized updating our team’s persona sets away from the names and faces I had grown to love (and, ahem, spent hours printing out, spray-gluing to poster boards, and shipping to all team members telling them to hang them up in their offices) and trying to commit to memory the new adjective-based descriptions.

In other writings on this subject, some have argued that demographic details and especially pictures can be used to represent people with non-majority characteristics or who aren’t usually stereotypically associated with a certain role to encourage others to think more broadly.

While I understand this point, any picture can contribute to bias one way or the other, so I believe the attempt to be as removed from bias as possible is the more critical goal. Moreover, this attempt to represent a group can often be seen as inauthentic representation and potentially as tokenism, which is yet another entrée for bias introduction.

How to create inclusive personas

Redesigning our personas turned out to be relatively straightforward for us. This evolution does not require throwing everything you have out the window—hopefully, your original personas had much more than just demographics and included content focusing on needs, wants, thoughts and behaviors. In our case, it boiled down to these changes:

• Removed the human name and replaced it with an explanatory adjective or two—e.g. from “Peter” to “The Decision Maker.”
  
  
• Changed the point-of-view from third person to first person. This turned out to be effective in a couple of ways—one, it removed the need for pronouns, but two, it brought more life and engagement to the persona by having them describe themselves directly.
  
  
• Replaced the facial image with a custom illustration created by one of our generous designers, who volunteered to create images for us. Note that while an image is not required, it can provide a nice balance for a template, but remember to consider any biases you might introduce within that image (like using a smiley face emoticon with a skin color).
  
  
• Removed references to other demographic information like gender, race, location, education, or occupation, unless the data was very important to the persona.
  
  
• Replaced any references to “wife” or “husband” with “spouse” or “partner.”
  
  
• Edited any other terms, phrases, or descriptions that seemed likely to introduce bias. For example, rather than saying this person just got promoted and has more money to invest—(maybe a team member reading the persona just got passed over for a well-deserved promotion and it creates negative feelings)—instead, explain simply that the person is looking for an investment opportunity for money they do not need to spend for a while.

This still left us with plenty of content, as well as some much-needed space on the page to include more details on goals, concerns, behaviors, expectations, and experience with various platforms. And all of us involved in the design discipline understand that a little extra white space never hurt anybody.

As I mentioned, sometimes a particular demographic point is important, and it should be retained. For example, if you work on a product that only serves a certain age group, such as children, then age should be a key characteristic of your persona. Or, if you are providing support to a certain people group, you would probably want to include that detail. Several other good exceptions are explained in Indi Young’s article.

Impact and adoption

The User Experience Research team already had a working group called “Persona Pros” that provided consulting services to groups who were trying to create or identify existing personas for their projects, so that was an easy venue in which to cement the use of the new inclusive template.

Now, whenever any of the researchers see one of the old personas, they immediately present the inclusive templates and concept, and several existing groups have already modified their old personas.

The timing was right to make this change at our organization, one that has always been supportive of any plan that promotes bringing more people to the table and not excluding people.

We did have some slight resistance from the groups and projects where a lot of persona development had been done and where significant effort was placed on knowing the personas and using their names regularly in design thinking activities. However, all understood the larger purpose of the change and ultimately adopted the new template.

An objection voiced more than once was that one cannot have empathy without some personal details; without it, the persona is just a faceless, nameless random person that they don’t really know. I see some validity to this point, but then I think about how my brain fills in the visual gaps for fictional characters when I read even if they aren’t described physically.

I conclude that we are still capable of empathizing with these personas, it just might require a bit more cognitive work for those that are accustomed to having photos.

Ms. Young also rebuts this objection: “Cognitive empathy requires not a face, not preferences and demographics, but the underlying reasoning, reactions, and guiding principles. Without these, you cannot develop empathy. And if you cannot develop empathy, you cannot wield it—you cannot walk in someone’s shoes.”

Besides, we should be empathizing with the needs, wants, and concerns of the users as a group of people, not necessarily as individuals, and that doesn’t require a name and a face. User experience practitioners are often told to, “Fall in love with the problem, not the solution,” (attributed to Kaaren Hanson, Intuit) so the focus should be on the problems we are trying to solve for those we design for.

For the most part, the impact has been impressive. I see the new persona template being used extensively with positive reception.

For those that simply do not want to use the new template, we didn’t put ourselves in the role of the “Persona Police”, but only that we recommend a particular approach and we provide reasons to adopt it. We are banking on the wisdom of the now clichéd, but very applicable, movie quote from Field of Dreams: “If you build it, they will come.”