One of the things that impressed me most during my interview with Mint CEO Aaron Patzer was his focus on iterative development and rigorously testing new features before they are rolled out. As I've said, most of the Web 2.0 companies I see focus on building new features more than they do on analyzing what their users are actually doing with them. It's crazy. It's like they all work at Microsoft in 1996.
There's a conference coming up that focuses on measuring and improving the user experience: Startonomics, October 2 in San Francisco. The conference will be run by Dave McClure, an entrepreneur, investor, and familiar face to Web 2.0 conference goers.
McClure's main point is that when you're building a new service, you want to think about "conversion events," in other words, moving users from one state to the next (from browsing to exploring, exploring to buying, etc.). That has nothing to do with releasing features.
However, you can't overthink things or dawdle. Release early, watch the right metrics, and revise. McClure points to Slideshare and Teachstreet as companies that are working this way. (He's invested in both companies.) Who's doing it wrong? "Anyone who takes longer than a year to ship," he says. Examples of this include Chandler and Trillian's Astra. By being late, they are missing their market windows. The world's moved past them.
I've worked a little with Dave at various conferences, as well as with the Startonomics conference producer, Debbie Landa. (Her company runs the Under the Radar conferences where I often moderate start-up pitch presentations.) I like what these people are doing with this conference, and I think more entrepreneurs should pay attention to the message.
Here's McClure's five-minute start-up pirate talk. He says this is the pitch that has morphed into the day-long conference.