Bump , the app that lets users "bump" smartphones together to share contacts and photos, is launching another service to let users pay the same way. It's called Bump Pay.
It's a simple solution, powered by PayPal, and designed to solve the headache of splitting drink or dinner tabs. Similar solutions exist, including one from Venmo  and even PayPal's own Bump-enabled app  which came out in 2010. (ING Direct's tech  is also made possible by Bump.) But now Bump, an app that's been downloaded by more than 80 million users, has decided to introduce a new app, rather than update its original.
"We certainly could have put it into the Bump app, and in the future, that's not totally out of the question," cofounder and CEO David Lieb tells Fast Company. "But we wanted to test this idea out as a separate app to see if Bump Pay is interesting and useful enough to people, rather than conflating it with Bump right now."
It's a problem many developers--at companies large and small--face when introducing new features. When Bump added calendar and music sharing to its app for version 2.0, the features saw little traction: Just tens of thousands of songs were shared daily, while calendar sharing represented less than 1% of usage. So Lieb spent a lot of time cutting feature after feature  (including the ability to share music, calendars, and apps) to whittle his service down to a core set of simple tools--a decision that caused a backlash when he released the pared-down Bump 3.0. (Google, Facebook, and Twitter, too, have endured criticism for similar actions.)
"It was really painful for us internally because we did have small numbers of people who really loved those features," he says. "People start to complain about it, and tell you how much they hate you because you took out these features."
To avoid the issue of radically changing the core Bump app again, Lieb's team decided to launch Bump Pay as a standalone app, under the umbrella of what the startup is calling Bump Labs.
"We decided that when we test these news ideas, we'll test them externally, so we'll never have a case where we gave it to people, some people didn't really like it while a few people loved it, and then those few people get upset when we decide to not keep supporting it," Lieb says. "Can we test out these things in a really simple way that doesn't largely affect the Bump app?"