• A little thing about personas

    For the longest time, I was skeptical of the value of personas in the user-centered design process. They felt too narrow, too constricted, too idiosyncratic for me to design for.

    That’s changed of late. User research is primarily about empathy — getting designers and developers to have empathy for their users, and be able to deliver products and services that really appreciate the users’ needs and goals. And personas are perhaps the best tool in the user-centered design toolbox for communicating empathy — they feel like real people with real concerns, and when crafted well, can transfer insights realized through research to other members of the project team.

    Because, for me, personas are all about empathy, I get riled up when I see persona deliverables that diminish the reader’s capacity for empathy. And perhaps the most pernicious element that I keep seeing cropping up in personas is the “user type.”


    Here’s a persona chart developed for Kivio, a diagramming tool similar to Visio. (I found it here.) There are many qualities I appreciate about this chart, such as just enough information to understand who these folks are (more detail if you click the link), specific characteristics about how they would engage with this product, good images to help us believe in these as individuals. All of these are great at creating empathy for them, and helping designers better understand how to serve them.

    But there’s one thing that keeps them at arm’s length: “The researcher,” “The Sysadmin,” “The OSS Developer,” “The CS Student.” It is common practice for personas to be given these kinds of user type labels but that practice diminishes our ability to empathize.

    It might seem like a small thing, but the moment these folks are typed they become members of a class, and their identities are now of being in those groups — and you start referring to that persona as “the Sysadmin” and not “Donald.” It doesn’t take much to go from typing to stereotyping–“Sysadmins want these kinds of features”–and once that happens, empathy is lost.

    There are 8 thoughts on this idea

    1. Leisa Reichelt

      yes, I agree Peter. For me personas are not so much design tools, they’re more about user research and communicating the UCD process to clients – and most importantly – eliminating ‘elastic user’ syndrome. Giving personas these group labels certainly doesn’t help achieve these goals.

      I think empathy is also a good argument for leaving *in* the more ‘fluffy’ details (pet’s name, type of car driven etc) rather than simply focussing on the tasks or ‘actions’ or motivations of each persona. I’ve noticed a prevailing force towards more ‘scientific’ personas these days, and I don’t think it’s particularly helpful.

    2. Ted Boren

      I totally agree that personas are about empathy, but not ONLY about that. When you water down a persona with too much made up stuff (the pet’s name example above) you dilute the informative power of the persona. Is the fact the person is a pet owner relevant? Maybe for your market it is, and you have data to back it up. If not, you leave your developer scratching their head wonder why this particular piece matters. There should be PLENTY of relevant detail available (or discoverable) to make personas interesting, personal, and “real.” We shouldn’t just make it up. As an example, at Microsoft, I was lucky enough to have some very specific psychographic data on a particular market – genre of books they liked to read, typical family size, economic data, transportation preferences, etc. I had also done a lot of contextual inquiry visits to really flesh out the real experiences of these people. By combining these data sources I could say stuff like, “Mikael gets up a 6:00 most days because he needs to catch an early train. He relies heavily on public transport because he doesn’t own his own car, as is not uncommon in downtown Stockholm. On the way to work he catches up on his e-mail and voicemail, and occasionally reads a science fiction novel (his favorite genre).” None of this is just invented; it was what we observed and what the psychographic data suggest. Does that make it less empathetic? I don’t think so — it just takes some effort.

    3. Amar

      I would agree with all of the above comments and would like to reiterate the point that Personas should avoid defining people as ‘users.’

      People lead multifaceted, contradictory, complex lives and defining their role as ‘users’ diminishes their motivations and goals for interacting with a product, service or environment.

      A good persona according to Carrol should read like a movie script or your favorite character in a well written book (based on primary and secondary research of course). In short, it matters how they are composed. (Nielsen,L., 2002)

    4. Slav

      The traditional principles of UCD do inspire, even if no clear questions are answered. The purpose is to get a better idea of the content, persuasion processes and not focus on the personas of primary objective. The primary personas important only after the structure of architecture along with messaging are complete.

    Comments are closed.

  • Close
    Team Profile