The new Backpack went live yesterday, and today 37Signals took the unorthodox (for them) step of releasing first day sales figures on their blog. "It was the best day in Backpack’s history," writes founder Jason Fried. So far, so good. While we don't know how that stacks up to the average Backpack sales day, nor how sales of Backpack will effect future sales of Basecamp, it doesn't really matter -- it is still far too early to make any judgments. But that's not what this post is about.

The day after we wrote about Backpack we wrote a post using Twitter as an example talking about how it is smart to let your customers help define your app. In it, we even linked to a chapter from 37Signals' book Getting Real about finding the right customers. While the main thrust of the article was about providing APIs and open access to products (something that 37Signals already does for many of their apps, including Backpack), the idea in general can be applied to the Backpack update, at least in theory.

Looking over the Backpack sales numbers this morning, I started thinking about my Backpack post last week, and I began to notice a contradiction between it and the post about letting customers define your product. In one post we wondered if perhaps 37Signals was "not eating their own dogfood, so to speak." While the other could be used to point out that if the changes to Backpack were a result of dogfood consumption, then it is likely that what they were doing is letting users define their application.

While I'm not privy to the motivations of the company, it seems likely that multiuser support, and the other features added to Backpack yesterday were things that current users of the app were asking for. As they wrote in their book, if a feature is worth adding, your customers will keep bugging you about it. "If it's really worth remembering, they'll remind you until you can't forget."

Changes as big as the ones 37Signals launched this week are not done on a whim. My guess is that these are mostly things customers had asked for -- and not just once, but a lot, over a long period of time.

So maybe, rather than losing focus and falling prey to feature creep, 37Signals was doing what we'd advised just a day earlier -- letting their customers define their app. Backpack started out as a simple organizational tool aimed at a single user. Perhaps customers helped define new use cases that prompted 37Signals to adapt the application to better suit the way in which people were using it.

I still worry that the changes have set it on a crash course with Basecamp, and I really hope that when the two apps collide the result is the a la carte service I envisioned last week. For now, though, maybe the changes to Backpack are just what the customer ordered.