Random header image... Refresh for more!

Personas in Action: Reflecting on User-Detachment & Generalization

Personas are talked about plenty that there is probably little else I can contribute to the discussion. However, I want to share a bit about my first official experience using this method in my design work. I say “officially” because were learned to use them in grad school but after getting a week-long training in this area at Cooper earlier this year, I’ve now realized how bad we were at the time. Now, I feel I have a better understanding of the usefulness and limits of personas. (not to mention the sweet Cooper University certificate to prove it :) For awhile I was skeptical of them–namely because I felt they marginalized the user into some wild estimation that gave the designer little insight. After all, I’m married to a cultural anthropologist and this is touchy territory. However, as I went through training I began to gain a different perspective on this method and its practicality. At the time I had also came across Steve Portigal’s Interactions column where he completely assulted this method and I found that this article and many others completely miss the point of Personas.

Personas, “when created correctly and responsibly” (disclaimer), are great ways for designers to stay grounded when designing features/products. Do they generalize users? Definitely. Do they cause you to lose some personal contact with your users? Somewhat. But, these can actually be good things. It’s entirely impossible for a designer to fully empathize with more than maybe 3-5 users. In fact, if you try to, the sheer difficulty in maintaining all of your users’ individual needs, desires and characteristics can overwhelm you to the point of ineffectiveness. Personas help generate patterns from users you interview so that you can actually detach yourself a little and concentrate on the characteristics that matter. In this post, I mostly only discuss the stage in Persona creation where the designer goes from research to characteristics to patterns. I don’t go into actually writing the Persona (I’m still too bad at this to write about).

As I used this method recently, when I got to the point of mapping user patterns based on characteristics I gleaned from 6+ user interviews, I found that it was difficult to really move forward until I was able to begin generalizing. I plotted my users on my characteristic charts and probably stared at them for 2 days before settling on a couple patterns. The reason why I struggled was because I kept looking at each individual users and thinking back on their story and workflow. Often, I opened my notes for them back up for reference. It was debilitating because I was struggling so much to empathetically process everyone I talked to. I finally started crossing out users’ names from my charts and replacing them with colored squares (btw, this is an advantage of plotting users in InDesign or other publishing software). It was almost emotionally difficult for me to really “let go” this way. I felt like I was throwing my beloved users, whom I cared so much for through their stories, out the window. Instead of seeing Steve and Susan, I saw green square and blue square. However, this release cleared everything up. The ultimate goal is to design interfaces that will satisfy most (all is typically an impossibility) of your users. By generalizing them, it freed me from the constraints of too much information and helped the common patterns emerge to the surface. This is the purpose of well-formed Personas, and this stage of the process is crucial. Any caring designer will hate the idea of “abandoning” their user for a characterization but it is the best way to ensure that you stay focused ONLY on the characteristics that matter for what you are designing.

I should mention that if you don’t believe the importance and value of generalizing users, just watch The Office, the “Survivor Man” episode. Basically, Michael Scott leaves to survive in the wilderness and leaves Jim to run the office in his absence. Immediately, the issue of empoyee birthdays comes up. Jim decides it’s stupid to have a bunch of parties so he proposes to group them into one. A great idea except that he lets everyone come and tell him what kind of cake/dessert they want and everyone’s request just piss everyone else off. It’s a disaster and Jim bails on the plan. The problem wasn’t Jim’s idea, it was that he had way too much interaction with his “users” and satisfying them each individually that it became impossible to satisfy any of them. Maybe this is only relevant to me…I digress…

Personas, in the very way that they force you to generalize users, help you sort through the particulars of your users. If you’re not convinced from my experience and an Office reference, then here’s a counter-example. A feature last year failed and had to be cut. It failed for many reasons but from a design perspective I always felt it was grounded in the wrong use cases. In short, several users were interviewed in researching the domain, but in talking to the designer, one customer would always stick out. Ultimately, it appeared that this particular user and their needs drove the design. The problem was that they had very specific and complex needs. As a result, the design accomodated these complex needs and completely alienated more common use cases. In the end, the designer empathized too greatly with a particular user and it was reflected in the overly-complex design. Persona creation helps prevent this. By finding the patterns that exist across users, Personas act as lenses for designers to filter out all the noise. Instead of referring to customers by name, we talk about the Personas because they are representations of commonalities across a broad user base. Steve and Susan may end up being Thomas, the CAD Engineer. But by focusing on Thomas, we are thus focusing on Steve and Susan’s common needs rather than one at the expense of another.

I’ve glossed over the intricacies of the Persona creation process. I simply wanted to focus on the aspect of generalization and how this can be positive for designers. You can read more about the stage I spoke about (from research to Personas) from Kim Goodwin. And of course anything else related to Cooper’s methods can be found on their homepage.

1 comment

1 Craig { 07.10.09 at 2:53 pm }

I really like the evolution of your specific individual personas written in your diagram into a single colored card. I think it well framed the general problem facing the concept of personas. Too often these imaginary individuals are thought of as “real individuals” rather than the more correct “real patterns of behavior”. The case described presented this idea very well.

Leave a Comment