Taking Personas Too Far

I don't have to tell you that at Cooper, we love personas—how could we not?—and we're glad to see continued excitement about them. That said, although personas are essential design tools, we think some people may be losing sight of the fact that they're just tools, and tools with a specific purpose, at that. Lately, we've been seeing a lot of gold-plated hammers—unnecessarily elaborate communication about personas—and some fundamental misunderstandings about the relationships among research, personas, and scenarios.

Be only as elaborate as you need to be

Recently we've been seeing people try new and entertaining ways to communicate about their personas. Over the years, we've seen our clients do everything from printing posters and t-shirts to playing persona trivia. This kind of internal marketing is important, and there's no reason it can't be fun. However, it's easy to overdo it to the point that personas become an end in themselves, rather than a favorite tool among several used to guide product and design decisions. Before expending a lot of effort and money on a novel communication method, ask yourself what that method will accomplish that text, photos, and illustrations won't, and whether the gain is really worth it. Also ask yourself whether that method will obscure the most important aspects of the personas: their behaviors and goals.

For example, I recently heard about a Web design agency building "persona living rooms" that are furnished and decorated according to the personas' tastes and filled with magazines the personas read. It's an interesting idea for helping people who don't participate in the research to get a feel for the personas' environments and attitudes. Understanding the personas' experience goals is particularly important for any situation where brand is a central consideration, such as a corporate Web site. Certainly, the novelty of the rooms will generate some attention and excitement, which is always a good thing.

In the long run, though, is this kind of three dimensional mood board really worth the investment? Say you use 600 square feet of office space to build mini living rooms for half a dozen personas. At San Francisco rates, that's about $20,000 a year just for space that could be used by real employees. Even if you manage to furnish them cheaply—and that's a big if—it would be easy to spend $30,000 or more on such an effort. Physical representations of space are helpful for designing physical products, but generally aren't necessary for software or Web sites. The exception is where the personas' workspace is quite unusual; for example, if your persona is a stock trader who keeps a dozen applications and information sources open on half a dozen monitors, you'll want to see how your application fits into that environment as you design it. However, most organizations are far better off with a compelling (but far cheaper) print or digital artifact. Cost aside, portable artifacts are more effective because they can be shared by distributed teams and go to the meetings where the decisions are made. Also, a room is a lot of effort to expend on something that focuses on just one aspect of the personas, without conveying critical information about goals and specific behaviors.

In using personas to design for a wide range of client projects and cultures, we've generally found that three types of persona-related communication are most effective.

  • Detailed descriptions
    First, each persona should have a detailed description that outlines current behavior and frustrations, goals, and so forth in narrative form, because stories are more compelling and memorable than lists. This detailed information helps people get to know the personas and serves as a lasting reference. Depending on the product you're designing, it can be helpful to include compelling research photos, sketches of how physical space is laid out, workflow diagrams, and collages of images that convey visual impressions of attitudes, taste, or lifestyle. For example, on a recent project involving a medical device for seniors, we saw a tremendous difference between the attitude and motivation levels of some of the interviewees. To help our client's team members share that research experience at a gut level, we included photos that emphasized the contrast between two of the personas. One was an active person with a positive outlook, so we included brightly-colored photos of a neat living space, an exercise class, and smiling people. The other was quite depressed and not at all sociable, so we used washed-out photos of a messy living space and no people.

  • Quick reference tools
    The second important item is some form of quick-reference tool that's easy to use in meetings and keep available in offices. This might be a laminated placemat or table tent people can carry with them or a set of posters on the wall. In either case, it's helpful if these items include photos, names, goals, and just a few key facts.

  • Meetings to introduce the personas
    Finally, it's important to have a live meeting when introducing the personas so you can cheerlead, persuade, and address any concerns people have. Whether you make this a fairly serious meeting or have some fun with it depends on your corporate culture. It's helpful to introduce the personas to a small group of stakeholders before sharing them with a larger team.
Beyond these methods, it's a matter of finding simple ways to keep the personas in everyone's minds over time.

Get the basics right before you experiment

We hope designers will continue to experiment with personas because that's how methods evolve and improve. However, it's important to be sure that any experiment—whether in creating, communicating, or using personas—gets the fundamentals right. One type of experiment we've heard about gets pretty far from the fundamentals: having personas take on lives of their own, such as posting photos and blogs in online communities. In other words, someone posing as the persona is interacting with real people and seeing how they respond. The first problem with this is that personas are models that encapsulate research findings, not research tools themselves. You should already have this kind of information before you create your personas. If you want to understand how teenagers interact online, for example, you'll learn far more by observing them doing so and talking with them about it. Also, most responsible researchers would agree that when research participants are identifiable—as opposed to anonymous aggregate data collection in healthcare, for example—participation should be voluntary and researchers should identify themselves as such, even online. For these reasons, having someone pose as a persona to interact with real people seems questionable from an ethical standpoint.

Regardless of what you try with personas, keep these things in mind:

  • To be effective, personas need to be based on sound field research, not derived from the functionality you've already decided to build
  • Personas should incorporate attitudes and experience goals when that's useful, but are first and foremost about behavior and end goals
  • Keep personas and scenarios distinct. The term "personas" is derived from the Latin dramatis personae, the cast of characters in a play. A set of characters is only interesting in the context of a plot, which is why we use scenarios. Personas should focus on current behavior, not speculation about future behavior with a product. Scenarios describe the personas' future behavior, first in an idealized and high-level fashion, and later in a somewhat more pragmatic and detailed way.

Personas are indispensable tools. If you make them well and use them for their intended purpose, they'll continue to be useful for years to come. They're just tools, though, and I've never yet seen a hammer that needed a room—or a life—of its own.

Post a comment

We’re trying to advance the conversation, and we trust that you will, too. We’d rather not moderate, but we will remove any comments that are blatantly inflammatory or inappropriate. Let it fly, but keep it clean. Thanks.

Post this comment