Facebook's mission is to connect the world and provide tools that help you do so.
Facebook dating is a recently launched feature that allows you find people who are matches for you basis a personalised quiz that you take.
Users get matches and then they can start conversation with matches.
Goal here is to drive engagement
Assuming the goal is here for product metrics in BAU state and not launch driven metric
User journey here
user signs up > user takes the quiz > user shows preference for people he/she likes > match > engagement > user may churn off
I feel that sign/user takes the quick = handleded under growth metric
primary metric: i want to focus on people showing preference for people as it's a leading metric and also is a great input metric plus indicates engagement well
defensive check metric: Users may want to take converstion off facebook which runs the risk of churn however user may come back after a while to try dating again
monthly churn - people who become inactive suddenly ( since they took conversation offline, this should be easy to detect )
re-activation time- after how long do people come back to FB dating
other health metrics
Matches created daily
% of matches leading to conversations
New conversation created everyday