Math and the Mobile Weekend

A few years ago, I was sitting in a management review of a fairly new product, and we were dismayed to see a 7% fall in monthly revenue, when we had been expecting a small increase. It took an embarrassingly long time before it was pointed out that we were reviewing February, which was 10.7% shorter than January, so a 7% fall was closer to a 4% rise on a daily basis.

This has gotten even more interesting in mobile. It’s pretty clear that consumers engage with mobile devices more at the weekend (look at your Google Analytics logs). Is that important? This year the three months have been completely different. January and February both had 8 weekend days (26% and 29% of days), and March had 10 (32% of days!), April is back to 8 (the only other month this year with 10 weekend days is June). It’s also important to consider holidays, New Year’s Day, Presidents Day, MLK day etc can all impact this calculus.

Main point- in rapidly moving markets, make sure you are tracking (the right numbers across) relevant time frames. And expect June to be an unusually good mobile month this year [2013]…