5 tips for building better user personas

I think his name was John.

John was a busy executive. He had no time to waste, and didn't suffer fools gladly. He worked long hours. He wanted to make the best decisions for his company. This meant that he needed the right information at the right time. He wanted his people to get right to the point.

John was my first persona. I think I wrote a good three paragraphs about him—around 500 words. I added all kinds of details that I thought were essential. I wanted our designers and developers to really get John. When they were sketching out a new feature, I wanted them to smell the coffee on his breath while he leaned over their shoulder. I fantasized about going into a sprint planning meeting and hearing one of our devs say, "But what will John think?"

When I shared John with my manager, he highlighted the phrase "get right to the point" and wrote "LOL" beside it. He was absolutely right.

Personas are one of the most commonly used deliverables in design and development these days. Unfortunately, they're also one of the most misused.

Here are some tips on how to make sure you are using personas effectively.


Are personas necessary?

Before I get into discussing how to get the most out of your personas, I want to suggest that they might not even be needed.

A lot of development methodologies don't use personas at all. That's fine. Often these approaches are more experiment based. You come up with a hypothesis, create a prototype, and test it with real users. You collect feedback and make adjustments.

The key is that there's a constant research loop. Proponents cite personas as a means to understand your users. If you are doing that work elsewhere in your process, personas might not be necessary. Don't create personas for the sake of creating personas.

The important thing is that you do the work to understand your users. That brings us to my first bit of advice.


Tip #1: Stick to the facts

Personas are tools for disseminating research.

All too often, personas are sketches of users based on pre-conceived notions of who users are. They might come out of a workshop where stakeholders sit and talk about what they think they know about their users. These get translated into some kind of template and then trickle down to the product or design team

In other cases, organizations create personas based on who they wish their users were. They indulge in a bit of a fantasy and create the ideal user or customer they are trying to reach, whether such a person actually exists or not.

Either approach tends to affirm preconceived notions about customers and their needs. This completely misses out the real value of personas.

Personas should explain qualitative and quantitative research available about your users. If your persona is not based on actual research, it's at best a proto-persona. That means it documents assumptions or hypotheses about your user that your research team needs to test.

If you do not do the research, you are accepting the risk that you're going to be researching in production with live users. That's not necessarily bad, assuming you are able to respond to the feedback. But, if you're trying to be efficient with your resources up front, you might find it's cheaper to do that research before you start writing code.


Tip #2: Look past demographics

In other cases, the personas are demographic stereotypes. If you look at enough personas, they start to look the same. You encounter the millennial who lives on their phone and has a lot of disposable income despite not being able to find a permanent job. Sometimes you'll see a single mom who is trying to make ends meet and lives for her kids. There's usually an older guy in a suit who has a lot of money to spend. Then there's an elderly person who isn't very tech-savvy.

This approach is dangerous. For one thing, it perpetuates stereotypes. "Millennials" are a diverse bunch, believe it or not. So are women, parents, the elderly, and every other bucket that these stock personas might fall into. These personas are based on uninformed assumptions about these groups and propagate those assumptions in technology. It's lazy. It's bad for design. It's bad for society.

Not only that, it's based on a weak understanding of demographics. Demographics are basically just buckets: they describe the make-up of a population based on very simple factors. They do not describe mindset, beliefs, attitudes, or motivations.

And, to complicate matters, we're living in an age of post-demographic consumerism. People can and do fashion themselves and identify with any number of groups or tribes or whatever else you want to call them. They have the power to choose who and what they affiliate themselves with. If you build your product roadmap based on some pre-conceived notion of a demographic category, you're limiting your reach as well as doing your users a disservice.


Tip #3: Focus on function, not form

Personas are not an output. They're an input.

Still, many organizations focus on the form of their personas rather than the function. They'll carefully select a nice stock image of a photogenic person and a few biographical details. They might add some pictures of what's in the user's bag or backpack, along with info about their job, their kids, and their hobbies. Sometimes personas have little graphs indicating things like technical skill.

These things all make for a nice deliverable. Stakeholders love them. It's visually appealing, and it tells a good story. They also offer stakeholders a sense of certainty—a sense that they "know" their users and their needs. Hence the bar graphs: it lends an air of science for instant credibility.

Unfortunately, a lot of the information included in personas is useless. They get cluttered with extraneous information that looks cool doesn't provide value. It's just noise. And worse, people might mistake design elements for meaningful data. That can lead to bad decisions based on guesswork instead of facts.


Tip #4: Keep your audience in mind

What information should you include in a persona? And how many personas do you need? It depends on who will use them and how. Think about what your audience actually needs.

A persona created for a digital marketing strategy is different from the way they're used for a UX project. A marketer might have several audiences they need to appeal to and need to develop a campaign for each. In that case, it is valuable to have a few personas to build for, and a information about how to reach them. Similarly, a content strategist may want an inkling of the topics a persona wants to read about.

When it comes to a UX project, the needs are different. For one, you probably don't need many personas. I'm an adherent of David Travis' belief that you should focus on a small number. Many—maybe even most—apps and websites can get away with one primary persona. Sometimes you might need two, such as if you have buyer and seller roles. Rarely do you need to go beyond that many.

Then, the persona should share only information that's actually relevant. Most of the time I would want to see a list of top tasks or goals. (When creating a large website, I don't think there's anything better than Gerry McGovern's top task analysis methodology for uncovering what your users' priorities are. But if you're using that method, I don't think you need personas at all.) On a smaller-scale project, I want to know what most of the users want to do most of the time. That helps me decide what to prioritize.

I know that some designers do value those kinds of extraneous details, and that's fine. If that's what they need, then they should have access to it. But aside from that, cut the clutter.


Tip #5: Let them live

Finally, it's critical to treat your personas as living documents. As you learn more about your users—and you should always be learning more about your users—update your personas to reflect the new information. Revisit your personas regularly. Do research to see if they still represent your users. Update them accordingly.

It's especially important to review your personas if you are planning any significant change to your business, product, or service, as well as if you are engaging in any planning or roadmapping activities. If your personas are based on good research, they can help inform those kinds of discussions. But if they're outdated, they might lead you down the wrong path.


Get what you give

You get out of personas what you put into them. If you create personas based on stock characters and guesswork, focus on appearances instead of quality, and then stuff them in a drawer until your next project, they're not going to do a lot for you. But if you invest in their creation, care, and feeding, they can be a valuable tool to help your team make better decisions for your products.