Hold Your Horses
Steve Portigal

Portigal Consulting | [e-mail]

David Gartner

In the documentary “Keep the River on Your Right,” anthropologist Tobias Schneebaum is the cultural expert aboard a cruise ship traveling the waters near the Asmat region of New Guinea, where Schneebaum once lived. He brings fellow passengers ashore to witness tribal dances and circumcision rituals. Schneebaum characterizes these passengers as tourists who are interested visitors. In contrast, he presents himself as an observer who stayed.

It’s an interesting continuum: from passive interest to active examination, from temporarily visiting to semi-permanently staying.

One of the criticisms (and there are several) that academic anthropology and its adherents have voiced over the contextual approach taken by user researchers concerns the brevity of field time; methods and theory that assume the researcher will live with subjects for months or years are applied to projects lasting only weeks. What depth, critics ask, can be learned in such a short time? Contextual researchers engaged in design and usability activities are more than visitors and less than cohabitants; we are more than interested but less than exhaustive documenters. The key (as any methodolo-gist will tell you) is to match the time invested to the level of insight required.

In product development there’s enormous pressure to produce

results in reduced time. This is why there are practitioners advocating for sexy-sounding approaches such as “extreme user research,” “guerilla ethnography,” or “rapid ethnographic assessment.”

Insights in 33 minutes, or your money back?

Although I’m concerned over the mistaken belief that the time invested can be squeezed and squeezed again while still producing the same value, there’s a more important attribute of time to consider. Rather than the calendar time in the field, let’s take time as a mind-set and consider the pace at which we work.

Last summer I sat in on a focus-group-like session. We were at the end of a long table of people whom we had met in various observations and interviews throughout the previous week. One of the clients who had commissioned the work was sitting at our end of the table and operating the video camera—no small task, with about 12 people engaged in conversation. At one point she turned to me and asked: “We don’t need to get this stuff right now, do we? Nothing’s happening, so I can stop recording?” Surprised, I encouraged her to keep the video rolling. Editing in-camera may have worked for Hitchcock, but it’s absolutely not the way to go for any sort of user-research process. You don’t necessarily know the value of what’s happening in the moment that it’s happening.

As a corollary, although I rarely immediately discard something that happens in these settings as not valuable, I do sometimes notice things that really excite me in the moment, things that are clearly quite valuable. But that’s usually a moment of discovering a pattern across multiple interviews. I would encourage our client with the video camera to simply slow down and let things unfold. It’s not that we aren’t structured or planful in our work, but much of the structuring happens in the preparation: finding the right people to talk to and figuring out what you want to talk with them about. Successfully executing the plan requires that we trust our process, and in many cases that’s about slowing down and building a space for the work to happen. We need to slow ourselves down, and we need to slow down our inner critics. We must be prepared to be surprised when we encounter something we weren’t even looking for. Our video camera (both literally and figuratively) must be on “record.”

Our clients shouldn’t approach contextual research expecting insights to magically appear, nor should they expect them to appear within any individual session. In order to recognize a pattern, we must encounter an individual instance multiple times. I recently taught a design research class to undergraduate industrial designers. For their final project, one team presented a series of