I tend to think Princeton researchers -- and, more importantly, CNET readers -- are a pretty bright bunch. But if a new study from Carnegie Mellon University is to be believed, both groups are dead wrong, at least about the future of Facebook.
Late last month, a widely publicized Princeton study (PDF) using computer models derived from studying the spread of disease predicted that Facebook would lose 20 percent of its users between 2015 and 2017. Facebook immediately blasted the study, in a pretty humorous way, saying that if it used the same methodology as Princeton, the college itself would have no students by 2021.
During this kerfuffle, CNET asked readers if they thought Facebook would still be thriving in 2017. As you can see for yourself, 50 percent believed the social-media mammoth was going to go the way of the, well, mammoth.
Now, the latest study that's turned its attention to this oh-so-pressing problem, says the Big F, which just turned 10, will still be here for the foreseeable future. Bruno Ribeiro, a postdoctoral researcher in Carnegie Mellon's computer science department, came to this conclusion by applying a model he created using data from Web traffic data company Alexa to six years worth of user data from 22 membership-based sites like LinkedIn, Meetup.com, and -- yes -- Facebook.
Unlike the Princeton study, which looked primarily at Google searches for "Facebook," Ribeiro's model (PDF) looked at sites that have grown, as well as those that have suffered attrition, and analyzed various user trends like how likely active members were to become inactive, what role members played in getting others to join the site, and what role advertising played in increasing the site's traffic.
His conclusion? Facebook, along with LinkedIn, Huffington Post, and the Ashley Madison dating site -- which helps people find relationships, even if they're married -- will continue to thrive, while Flixter, OccupyWallStreet, and Tea Party Patriots won't. He does, however caution that, as with all predictions, unseen activities like the launch of a new social-media upstart that steals some of Facebook's thunder could affect his model.
One of the key components to a successful membership-based site, the study suggests, is users engaging in irritating behavior like twice-hourly updates on how fed up they are with the annual occurrence known as winter. "If this model is correct," Ribeiro said, "social network sites will try to make your friends' lives seem more interesting and your feedback on posts more urgent. From the model's perspective it is beneficial for companies to be encouraging this type of behavior."
Great. Just when we thought cute cat photos might be on the decline.