I thought it might be fun to get people's views on analytics they've used which have challenged cherished or long-held beliefs.
One from us is the time online postgraduate students, predominantly mature-aged workers, log in. It was long held they logged in after work and on weekends, and particularly late at night when the kids go to bed.
Using Pyramid we produced heatmaps for our various cohorts of session log-in times and days -- and what do you know? Online students do log in evenings, but they also peak at 10am Monday, with Monday 9am-1pm a heavy time for new sessions too. Guessing they turn up to work and start working out what they need to do, catch up on discussion forums, print materials, ... The higher number of sessions Sunday from noon to 10pm, much higher than the equivalent time on Saturday, was also interesting.
Any analytics that have caused you to re-think your views?