I wonder how much of their time the second time around was saved because they had already built so much back-end? I know the gist of the article is to concentrate on ideas and user experience instead of features, but having all that work in place had to be a big boost for them.

I think some of the best programs I've seen were built once, thrown away, then built again. Once you start coding, it is easy to put the blinders on and think only in terms of features and not users. Likewise, it's easy to think of a million user experiences you want to provide without having any clue how the pieces are coming together. In a way, they had two startups, one for each problem. Seems like it worked out well for them.

We were able to carry over some of the back-end work from the first version, but not the majority. Much of the old code was fairly convoluted, and we wanted the new codebase to be easier to maintain and build on. The good news is that we're very happy with how the new code is working for us, and it should serve us well for a long time.

You're absolutely right about how easy it is to get so wrapped up in the features that you completely forget how anyone will actually use them. From the moment we started developing the new ImThere, we really shifted the way we worked to avoid falling back into that trap. We're always looking for ways to get even better at thinking about new functionality from a user's perspective. We test every new line of code that goes out the door extensively not just for bugs, but also for usability and how it affects the flow of the site. It takes more time to do, but we've found it completely worthwhile.

Thanks for the great comment!

- Gorman