Our Blog
The User Is Always Right: Usability Testing
I’m a third of the way through Steve Mulder and Ziv Yaar’s insightful guide to to creating personas, The User is Always Right. As part of that process the authors provide concrete examples and rules for conducting user research for the initial phase of design for web.
I’m finding that so far, along with some great nuggets of insight, a good portion of Mulder and Yaar’s process refreshingly mirror our own here.
“…Traditional usability testing often isn’t as helpful for creating personas…When you give specific tasks to users, the test becomes about the user goals you have chosen instead of the users’ unique goals.”
One example of that is the caution the authors’ express when using usability tests as a means of qualitative research for persona development. In trying to understand users’ goals to ultimately craft the best experience, traditional usability testing tells us successes or failures of a given task at hand. Rather than uncovering what we don’t know, this type of research usually tests what we suppose is right or wrong with what exists versus informing what needs to exist.
That being said, user testing is effective when the conditions are right. Field, ethnographic and quantitative research are other means of research that provide great insights in our experience and that Mulder and Yaar speak to at length in the book.
If you’re interested, the pitfalls of user testing is something T+L have always been cautious of and something David has spoken about in detail more recently.
I’m looking forward to finishing the book and providing more feedback as to how it compares and contrasts to our process here.
I look forward to reading the book.
Its true that usability testing is not useful for creating personas, but then again i didnt think the point of usability testing was personas?
I recently sat through a 1 hour usability test (conducted in a proper lab with cameras, sound proof wall where we watched and a monitor to record each move), and it was very eye opening to see a user perform basic tasks.
Overall the activity had a great outcome as the designers were able to pin point the sections of their sites that were most challenging for users.
But none of this went into a persona, if anything the tasks set out for the user were taken from personas.
If personas are driving usability tests, that’s certainly not a bad thing. But what the book seems to drive at is that goal-driven personas are the ends to the research; they are the direct inputs to a redesign rather than the research itself.
Our experience is though usability testing has it’s place, often the environment is too unnatural, the variables too unruly to control and the given tasks too specific to have any genuine ‘aha’ moments out of it. More on that in the previous posts I linked to above.
I’m glad to hear the book is useful. I think the important point I’m arguing is that it’s not enough to build personas just out of what people SAY in interviews. Somehow we also have to observe what they actually DO. I believe behaviors are particularly important for creating personas, so the more actual behavioral research we can fold into it, the better. Usability testing, while it has its limitations and while its primary objective isn’t persona building, is one potential tool for researching actual behaviors. So are field studies, which are usually more effective in building personas. And so is web analytics data, to a degree.
Ya put a smile in my face.