Tacit knowledge and software usability

I spent some more time last night sorting out my friend Larry's weblog. He's a psychologist who specializes in Asperger's Syndrome. Larry's really interested in open-sourcing his ideas and would be an ideal candidate to run a Radio Community Server that would be a knowledge exchange for the AS community. So I've been watching his experience with Radio very closely.

In a way, some of the problems he has run into were my fault. For example, when I installed Radio for him, I put it in c:\\radio instead of c:\\program files\\radio userland. I do this to avoid ever having to type directory names with spaces in them. But when Larry accidentally double-clicked the Radio installer instead of the Radio app one day, and reinstalled the software, he ended up with copies in both locations. This created all sorts of weirdness that I've only now (I think) untangled.

Left to his own devices, Larry would have probably been perfectly happy with no item titles. But I, having championed that feature, encouraged him to turn it on. After raising his expectations, I then dashed them. We were both mystified not to see those item titles appear on the site. I had completely forgotten that the <%itemTitle%> macro needed to be added to the item template! When I did that for Larry last night, it was clear to both of us he'd never have figured that out.

It was the same for comments. Larry really wanted comments, and thought by enabling the Pref he should have them. Editing an HTML template and adding a macro to it is just not something he'd ever have gotten to on his own.

This is not simply a Radio issue of course. I have a column due up on BYTE.com later today that talks about how to manage email more effectively in Outlook. I felt a little guilty writing that column, because it seemed too elementary. But when I spent a few minutes applying those techniques to Larry's Outlook -- things like filtering, grouping responses, customizing search views -- it really made me think about how much tacit knowledge some of us have, and how little of that knowledge is available to others. Larry has been suffering from the tyranny of the inbox because even the concept of email filtering was unknown to him, never mind the mechanisms available in his software. He's thrilled by the little bit of configuration I did for him.

One final tweak was to write a script that backs up his Radio files, and also his My Documents directory (with several thousand documents that had never been saved elsewhere) to a Jazz drive that had been sitting around unused. Larry then told me about how another professor at the college had lost a 25-page paper that wasn't backed up. Now he can click a desktop icon to safeguard his stuff. Had I not intervened, this clearly wouldn't have happened. And I'm certain that scores of Larry's colleagues are in the same boat he was in.

As technologists, we hold all sorts of knowledge that is tacit. We ourselves don't realize that we possess it, and we don't realize that others (most others) don't. Radio does a remarkable job of delivering an out-of-the-box experience that doesn't depend on too much tacit knowledge. When you try to go further, you're on a slippery slope, but this is true of all software.

For years people have argued that software must relentlessly improve its score on the "mom test" and that is certainly true. But there's another angle here comes back to the KM aspects of blogging. When we narrate, we externalize what we know. We convert tacit knowledge to explicit knowledge. This can help software become more usable for two reasons. First, when technologists narrate what they know, they're more likely to realize how much tacit knowledge they have and expect in others. Second, when non-technologists narrate what they know, technologists can see more clearly that the expected tacit knowledge is missing.

Former URL: http://weblog.infoworld.com/udell/2002/05/13.html#a235