On March 10, 2007 STC Austin hosted a seminar by STC Associate Fellow Whitney Quesenbery (www.WQusability.com). The seminar, "Using Personas", had us write scenarios, walk through sites using personas, and derive design/documentation ideas. While I was happy to learn about persona usage, I was discouraged by the requirement to amass user research data so that it can be distilled down into useful personas (with what time and resources will this happen?). When she mentioned the shortcut of "surrogates", I immediately connected with the conclusions of a book I just finished: Stumbling on Happiness.
In Stumbling on Happiness, Gilbert teases out of brain and psychology research conclusions about why we are so bad about making choices to ensure our future happiness. Much of it came down to the problem of forecasting, of responding to the future as we imagine it. He concludes that, because imagination involves so many perception flaws (extrapolating on the present, focusing on past highs/lows, filtering through current beliefs), we'd do far better to rely on surrogates: find someone who's already experienced what we're imagining and ask what they think/feel. He details why we're so resistant to doing this, not least of which is our proud belief in our uniqueness -- when, in fact, we all are amazingly (or depressingly) similar.
So, apart from well-researched personas, I see ways to leverage surrogates for accuracy. First, in documentation/training, is to use our own first exposure to the product to itemize problems in the interface (we being surrogates for the user) -- we have to do this early, before we become accustomed to the UI, and we have to actually struggle through complete user tasks, rather than imagine how it might be to do it all. Next, ask end-user trainers to be surrogates for their students; have them itemize problems and issues they know they'd have to address in class. Also, ask support reps to be our surrogates in communicating with the users, delivering specific questions and sending up feedback from users. Don't assume; find out.
I think surrogacy can be a useful concept for nudging us out of our navel-gazing and guestimations and into reality-based UX evaluation -- which, after all, is the purpose of using personas.
Why Personas?
The purpose of personas in UX design is to summarize what we know about users to make it accessible to UX designers, developers, etc. PM has a wealth of information, but it's too much for many people to understand, relate to, and identify with.
Are you familiar with how public schools are rated exemplary, recognized, etc.? The purpose it to summarize the mountains of data that are collected, so the public can understand and relate to them.
So it is with personas. They usually have names and a few irrelevant details so that they will be seem more real – people who can be the personification of the customers and whom we can strive to serve as well as possible. (By customer, I mean the person who uses the software, not the purchasing agent.)
Cheryl Dane suggested that we have a naming contest for personas, so they would have more visible to the community. It’s a great idea.
I believe personas are worth the effort, and UX is still planning to develop a few personas for iMIS, but it is not our highest priority right now.