Our team has also created a model to compare the value Play delivers against the revenue share paid by a developer on their consumer spend. Two key points that we believe may underlie Tinder's decision-making:

• Despite generating the highest apps revenue from Play, Tinder severely under indexes on the % of installs coming from Play driven sources vs. dating peers and games. In 2018, only 8% of Tinder's installs came from Play driven sources vs. 15% for OKC and 27% for Meetic. This bolsters their assertion that 'people mostly come to Play looking for Tinder'. • The relative discovery value they have been deriving from Play has decreased since 2017, due to an increasing share of acquisitions coming from Search and Search driven re-installs. In parallel, their revenue has been increasing y/y. Thus the team estimates that if you compare the value of non-Search driven discovery vs. revenue share paid, Tinder is now deriving only 10% of the revenue share value vs. the 30% share they pay.

Paul Feng added: 'The analysis does a nice job quantifying Tinder's "people mostly come to Play looking for Tinder" argument. As part of the Magical Bridge AND Payment Policy efforts, we're looking at ways to tier pricing — we can potentially use this. Consider a pricing tier based for apps / brands that bring an inordinate number of users / FOPs / NPUs to the store.'