The Thing about A/B tests — To Learn or To Ship: That is the question.
The Product Podcast
00:00
How to Build a Heuristic for a Two Ship Test
When you're working with a two ship test, basically you answer all your product trade off questions by saying, will not solving this make us feel bad about this being in the product? You can get by not solving a lot of edge cases because it's not going to show up in the data. If you're committed to going back and iterating, this will save you a lot of time. On average, we want to have four or five of our tests fail for every four or five tests that succeed.
Play episode from 08:09
Transcript
