Valuating social networks – it’s not that easy (TechCrunch remix)

In a recent post, TechCrunch has been trying to model the “real market value of social networks“. While the effort is interesting and gives some ballpark figures for Facebook/MySpace-like SNS, there are a number of issues with the model – in addition to those highlighted in the comments – that prevent it to be applied in a general way.

1. Not all SNS are created equal

Let’s take an easy example: a SNS that requires no invitation, and a simple email with no check to register. In this case, we can suppose a large number of registered users, most of them dropping dead after registering as there are not enough links between them. Trust is low and targeting for ads almost impossible as there is no user profiling aside from the content users post.

Another example: a SNS that is invitation-only, requires real name, age, gender, and allows you to connect only with friends you know already and accept the connection. Quite restrictive. Assuming it works, the result is a high-trust network with close links.

To give names on those two extremes, the first one would be Mop.com in China, the second Cyworld in Korea. Others are likely to gravitate somewhere in between. Now what if Mop and Cyworld had the same number of users? What if Mop tried to monetize with ads while Cyworld does it with digital goods? Which has more revenues?

So back to TC’s article, how much is a SNS worth? “It depends on how you value a user”. Without even looking at the valuation of a user based on the ad market potential, users are already not equal as their relationships due to the DNA of the SNS differs.

The issue of demographics has also been pointed by Dave McClure and others in the comments. There is a professional SNS in Japan, M3, which has 150,000 users, they recorded 100 million USD in revenues last year. That’s over 650 USD per user. How? Well, those 150,000 users are doctors and guess who advertises? Aha.

2. It assumes SNS can only make money from ads

It comes largely from a US perspective and is understandable: if you operate in a rich market for online ads, why look further? If you don’t have the luxury of a rich online ad market, you have no other choice but to find alternative business models. The personalization (skins, fonts, background music, avatars, etc.), casual games, alerts and mobile business models have all be proven in several places (not only in Asia).

What would be QQ’s valuation in China considering the online ad market is 1 billion for 220 million internet users (<5 USD per user? That ain't too much compared to 132 USD in the US). And how could a model be global if it leaves aside the world's largest internet market in terms of users?

Again, "It depends on how you value a user", and users should then be valued beyond their ad sales potential (else you could buy QQ for a dime vs. its current valuation over 10 billion USD on the HKSE).

What to remember

Eventually, the model seems fairly alright to compare SNS with a style similar to… MySpace or Facebook, relying mostly (entirely?) on ads. Anything that derives from this will go off the chart.

Our point here is to show that “SNS” as a term has been stretching too much and it would be time for more granularity to describe the great variety of DNAs, business models, demographics, etc.

We introduced last year the “ARFU” (Average Revenue From Users) as a new metric to distinguish pure ad-based SNS from those with mixed or non-ad models. Maybe other new terms would be welcome to avoid treating every SNS as a MySpace or Facebook lookalike.