Lenny Rachitsky: evaluation

Transcript context

…That's right. Right. I'll give you a very tactical example is when I joined Instagram in January 2016, believe it or not, the onboarding, the sign-up flow at Instagram had literally no logging to it. It had logging of how many people started and how many people ended. Right. And I joined in January and it was like we had to write a roadmap. And so the roadmap looked like this. Okay, we know this amount at the top of the funnel and this amount at the bottom end of the funnel. And there was eight steps in the funnel and we don't know what is going on. Right. So the first bit of understand work was we needed to do the instrumentation of that funnel as fast as possible to get the data to figure out where the drops were happening and what to fix. But because we knew of what was happening at the top of the funnel and the bottom of the funnel, we can go and play around with the experience and see what was broken. So we ran a bunch of tests of stuff that was obviously broken to see what would improve. Right. And so it was like a mix. And so what we did was we set up time where we did, the beginning of like the first couple of weeks, add the logging, ship it to code, re-evaluate in the middle of the quarter, look at the full funnel and then add more things that we can do later on once we got that. But in order to get that done, that involved understand work with growth marketing to figure out what was the schemas for the instrumentation, right, the engineering to actually do that logging. Right. Data science to like pull together funnels and dashboard it, all of that had to come together all at one time. I really like just this concept of calling it understand work. I think that alone is a powerful tool because we're going to spend more time on understand work. It feels like it gives meaning to stuff that otherwise people would brush aside.…

Stored transcript either side of the excerpt. The highlighted words are the published quote; the surrounding text is unedited source, never generated.