Minutes of use is important because it starts to show how engaged the public is with a given service, particularly if you divide the number by unique visitors over the same time frame, getting average minutes per user. The more minutes, for example, the more time you have to deliver ads, rather than assume that each visitor stays long enough to see any appreciable number. But the volatility of these numbers is astonishing. Over a one year period, Twitter is up by a factor of 37. Facebook gets seven times the traffic. MySpace? Down by nearly a third. And Tagged.com? Up almost ten times. (I guess all those spam emails does something after all.)
How does a company plan for those types of rapid changes? Not just suddenly having to scale up to handle the traffic, but scaling down as the in crowd moves like a mass of lemmings to the next online happening? It's tough and might explain some of who so many social media companies have a hard time with monetization. Abrupt changes are not something that advertisers want, and basing a business model in part on your popularity gets tricky when you don't know by how much it might change tomorrow.
Illustration via stock.xchng user svilen001, site standard license.