We did a lot of things right in the first 2 years of Judy's Book business (FYI - July 2004 was the start of the business), but I think we made two reasonably big mistakes. Both mistakes prevented us as a company from achieving critical mass as a company in a specific geography -- i.e. Seattle.
The first mistake: we weren't aggressive enough in customer acquisition. We had the idea of using the address book as a hook into customer acquisition but we never spent the energy and focus to really maximize the use of the address book in growing the social network. We thought that using the web address books was too "spammy". In retrospect, this perspective on customer acquisition was a little too ivory tower and not enough boot strappy. In fact, there's a broad spectrum of what constitutes spam in the eyes of a consumer. We/I didn't understand the subtleties of consumer perceptions sufficiently to build the app that grew our social network AND that consumers wouldn't perceive as "spammy". Linked In and Facebook today have finessed and pushed these consumer perception boundaries -- but back in 2004 and 2005 it was difficult to see just how willing of consumers were to send out invites from their address books.
The second mistake: we expanded out of Seattle in August 2005 and went national. We did this because Insider Pages expanded nationally and we were feeling the competitive pressures to keep up. The fact is that neither company had figured out what it took to create a community directory that achieved critical mass. I remember when we were raising money for the series B round and Bill Gurley of Benchmark challenged me on the decision to nationally. I defended the decision at the time. Looking back, I think the decision to go national in Aug 2005 was an error. We ended up spending the better part of 6 to 9 months building a local search web site that could be accessed nationally -- going national was more work than we expected. Ultimately, this decision prevented us from focusing on the customer acquisition problem I mentioned above as well as other improvements that would have made our product more sticky and compelling.
More mistake and lessons learned blogging to come....