I'd be more interested in a comparison of engagement (checkins per account per month) over time. The first metric can hide declining engagement if there are sufficiently many latecomers who haven't burnt out yet.