Show one ad only after a lesson ends, and test it against retention

Cem KansuAnd if you're just watching metrics, I'm sure that might look fine, but when you test it, you're like, "What the hell is this?" Then you should know what you've introduced and obviously avoid the things that you wouldn't want to see as a user.

Cem KansuHence, our first test, and honestly even today, that's our only ad placement, is after you've completed a lesson, we show you an ad. Not during, not before, but always after, because we believe that doesn't interrupt your learning.

Cem KansuSo one is the qualitative. It should be a good user experience, and you should know that by testing the app yourself every day. The second one is quantitative. Measure everything you can, like is retention fine, is app closes fine? What's happening to your reviews?

Cem KansuYou might have to make some trade-offs. Ideally you wouldn't, but realistically, if you go from zero ads to some amount of ads, there is some risk that some curves might come down, but I think you should know what that cost is.

Cem KansuAnyway, that's through AB testing, that's how we measured it to be... Honestly, our retention was maybe a slight hit, but mostly neutral for most of our ad tests over time. That's one.

Cem KansuThe third one is there is, of course, you go from a free product to selling something, users won't always say thank you for doing that. Imagine if Uber started completely free and one day they said, "Okay, now we're going to charge for our Ubers." Users would hate that too.