Fascinating if true. But I have a doubt.
Recently there’s been a splash about a paper that says Facebook will lose 80% of its users before 2017. The authors, John Canarella and Joshua Spechler (aerospace engineers from Princeton) have modelled MySpace’s rise and fall with epidemiological techniques and compared it to data about Facebook. Their calculations show Facebook’s user figures are about to take a nosedive.
Granted, Facebook’s user figures have started levelling off recently. But the paper doesn’t take into account that MySpace failed partially because of Facebook’s rise.
I agree with the authors in that epidemiological modelling shares some similarities with social network spread. Ideas do, after all, spread through contact between ‘infected’ and ‘susceptible’ people. Which social network to join is just one of these ideas. However, the irSIR model used in this paper doesn’t account for diseases competing for hosts, and a successful infection barring new diseases from infecting a previously infected host.
This is exactly the case with the evolution of social networks we’ve seen in recent years.
In general, people tend to use the social network that their friends are on. If a new network offers better functionality, (cf. Facebook’s clean functionality vs. Myspace’s cluttered nightmare) people will abandon an old network for a new one. But network effects are important. A new network can become the default because everyone else is on it. This locks users into the incumbent network.
Before Facebook’s rise, there were dozens of social networks. MySpace, BeBo, Friends Reunited, Orkut . . . they came and went. I argue that, through network effects, Facebook has obtained a dominant position that will not fall for the same reason the other networks did.
Just look at the comparative search data:
Even at its peak in 2007, MySpace had less than 20% of the interest Facebook has today. That’s a whole different situation.
My point is this: Cannarella and Spechler say they’ve discovered a pattern in which social networks rise and fall. But does their model take into account the fact that most of the previous social networks fell because another one sprung up? And does the vast size and scale of this new network mean it operates by different rules? I think it might.
The one thing that makes me wonder is that downeard trend in search volume you can see in the figure above.
What do you think? Will Facebook fizzle and die? Will it be replaced by something else? Or will Facebook become the next email – a technology staying with us for decades even though it is outdated and inefficient?