8 Reasons To Ship a Feature
As product managers, shipping features is kind of our thing, right? A product manager is usually the only person who is engaged in every step that needs to take place for an idea to become a feature. Listing all of those steps could be the subject of another blog post (or several). This blog post is about the last step: analyzing the feature itself, after building it, and deciding whether it’s successful enough to become an official part of the product moving forward.
Why not just ship every feature we build? I mean, we already put the engineering time into building it. It’s essentially free to keep now. I understand this thought, but features that are not successful should either be killed or iterated on/tried again. They didn’t solve the problem they were supposed to, so at best we’d be adding complexity to the product for no reason and at worst, we’ve introduced a change that makes our product worse for users.
Here are some reasons we commonly use at Yammer to justify shipping (i.e. keeping) a feature we have built.
- The feature worked: The feature increases product complexity, but the A/B test results show increases to core metrics. (this is the best one)
- Reduced complexity: The test results show flat metrics, but the feature decreases product complexity.
- Strategic value: The test results are flat or bad, but we need to ship it to ship a future project that we expect to be positive.
- Helps some users we care about: The test results are flat or bad, but the feature improves Yammer for a segment of users that we’ve prioritized for some reason.
- Better design: The feature is not very complex, the test is flat, and we feel that it moves our design/UI forward.
- Judgement call: The feature is impossible to test for some reason but we like it.
- We committed: The feature satisfies a customer or admin request that we agreed to.
- Checks a box: The test results are flat or bad, but the feature makes us more secure, compliant, localized, etc. in a way aligned with our strategy in that respective area.