Tinder Reports Rise in Sustained Conversations After Algorithm Change

Tinder says the number of sustained conversations on its dating app increased 2% year over year in August, following a change to its recommendation algorithm introduced in July. CTO Vinay Kuruvila explaind the specifics in Tinder’s September 3rd CEO Connection series livestream, titled “Inside Tinder Product Velocity“.

The company defines a sustained conversation, or “Spark,” as an exchange that goes back and forth at least six times. Tinder’s updated recommendation system is designed to prioritize potential matches based on the likelihood that they will reach this level of interaction.

The change, referred to internally as Queue Unification v2, combined Tinder’s previously separate recommendation systems into a single system. Tinder Chief Technology Officer Vinay Kuruvila said the unified system is focused on optimizing for “sparks,” rather than balancing multiple objectives such as likes, retention and revenue.

Match Group CEO Spencer Rascoff said the system now reorders recommendations with the goal of increasing the likelihood that users will not only match but continue talking.

The August results represent an improvement from earlier in the year. Sustained conversations were down 4% year over year in the second quarter and fell 1% in July before returning to growth in August. Another measure, Spark coverage, also increased. The percentage of users reaching a Spark was up 2% year over year in the second quarter, followed by a 5% increase in July and a 6% increase in August.

The recommendation overhaul has not yet been completed across Tinder’s entire user base. Kuruvila said in a September 3 presentation that the unified system was still being rolled out across different user segments.