What I Learned from Web 1.0

I want to thank Ken Yarmosh at Technosight who invited me to join in his second Blogoposium on "Lessons Learned from Web 1.0." I learned many lessons but here's the one I felt was most important:

It's hard to change user behavior.

I was at a software company that tried to do this and failed. We identified a big problem (an obsolete hierarchical desktop metaphor) and came up with an ingenious solution (organize all your electronic information into a time-ordered stream, like a supercharged email inbox.) I still think the concept is great, but let's leave that for another conversation.

We were an early pioneer in desktop search and got arrows in our backs. Our problem was that we tried to change user behavior in two ways. (Yes, in hindsight this seems like a really, really bad idea. But to be fair there was a lot of experimenting going on, much like today.)

First, we asked users to abandon files and folders. Most people we talked to knew that the filing system was broken and they were routinely losing documents and messages, but it didn't mean that they were ready to move onto something else. It took several years before people were ready to consider search as a UI metaphor for the PC. And it's still not clear that the Google Desktop will reach critical mass.

And speaking of search.... our second mistake was to ask users to view search results sorted by time instead of relevance. Of course now blog search is based mostly on time, but even back then users were already conditioned to expect results in relevance order.

We were trying to push time order as being more useful than poor relevance, citing the power of chronological recall and narrative storytelling. But user behavior in search was already set and we couldn't change it. In the end, we had a small (but passionate!) group of users that couldn't measure up to the expectations of our investors.

So the lesson learned was that most of the successful companies of Web 1.0 solved focused problems with incremental improvements that didn't upset user behavior. We're now applying those lessons at Zoundry. We're focused on solving a small but meaningful problem (affiliate marketing is too complex for nontechie writers) with a simple product based on UI standards (word processing + search) plus a healthy dose of anticipating user needs. And we're improving our products rapidly with user feedback - like all good Web 1.0 and 2.0 companies.

Technorati : Zoundry, blogoposium2, search
Del.icio.us : Zoundry, blogoposium2, search
Ice Rocket : Zoundry, blogoposium2, search