May 10, 2007: There's been lots of heat generated lately (though it's been going on for a while) about the role of ethnography in the design process. I want to add my own couple of cents without, I hope, repeating the usual turf-protecting, inside-knowledge-demonstrating sort of posturing of which the debate generally consists.
I believe that good design ethnography (ethnography in the service of design as contrasted with the academic variety) is not a matter of methodology, of knowing some special trick or fieldwork technique (as some HCI folks seem to think). Nor is the magic to be found in finding the perfect high-theoretical frame (PDF) (as seems to be the academic anthropology fixation, though that fixation makes a lot more sense in a non-applied context). Even direct experience, as important as it is, cannot guarantee success. All that stuff matters, but I believe the key is in the makeup of the individual practitioner (yet another human node in the landscape of human "centered" design). This may go against a lot of the academic bluster (e.g. "let's credential ethnographers" or the more gutteral: "you call that ethnography?") and it probably also goes against the currently-rampant design-shop posing (our "proprietary ethno-magical process"). But I think it's worth saying. Like most (all?) activities, some people are just plain good at it (see Chipchase, Jan who, whatever you think of his level of methodological or theoretical rigor, seems to have an uncanny knack for uncovering, noticing and then theorizing about the things people do).
Also underestimated and of great importance (as Jared pointed out to me), but perhaps the topic for another post, is the question of what you do with the results of the research. This, too, is a matter of specific actors and situations, but this part of the process seems much more susceptible to methodological intervention.