“we couldn't just like start it from all the pay walls right because we have a lot of different paywall and then the same times that's very expensive to and it takes a lot of time and then effort to train the model to test the model so we start with something small so we test with couple features”

Roll out ML paywalls on a small feature subset first — full rollout is expensive and slow

Tinder scoped the initial ML paywall rollout to a handful of features rather than the full product catalog, reducing training cost and risk. The incremental approach let the team validate the model quality before expanding. The eventual multi-million-dollar lift came after proving the concept at small scale first.