Not So Free Facebook

The IT industry has two types of products: those that save time (think of Google Search) and those that waste consume time (like Facebook). Though both are free, time spent on Facebook is sort of opportunity costs, typically equal to the user’s wage or whatever he does instead.

Even if the consumer formally pays nothing for either of the services, his behavior is not the same. That’s because of demand elasticities. One marginally relevant example:


These are demand curves. Percentage shows adoption rates. Nevermind the goods on the right. These are not IT and even not the developed world, but this is the most illustrative data of this kind around.

Most goods have elastic demand here. The blue curve also shows the striking difference in demand between zero and any positive price. This is very much like web products: the user base shrinks rapidly when the price becomes positive. For the freemium models, the premium user base is south of 5%. That’s why startups avoid pricing users at early stages.

Facebook also likes to pose itself as a free product. But it’s not really free. According to stats, an average user spends 40 minutes per day on Facebook. Though overstated, such usage is equivalent to $13 paid each day with the median US wage taken as opportunity cost.

Facebook, unlike Google, can set nominal access fees. Users already pay a lot for it, and equilibrium is around inelastic zone of the demand curve. Paywalled Facebook would make its shareholders happier because its current evaluation at $200 per user skyrockets with the enhanced cash flow. The current ad-based model is a dead end for Facebook because its ads target cold clients (compared to Google’s and Amazon’s visitors). While current earnings are very low for such a big company, Facebook’s P/E ratio of 75 is what investors are ready to pay knowing the forthcoming switch to a viable business model—and the paywall is one of them.

The logic of low elasticity under positive opportunity costs is relevant for other time-consuming services. Major newspapers had got paywalls long ago, but for other reasons: they have fewer users and high labor costs. Genuinely scalable web services are reluctant to experiment with payments and settle with nicely looking “premium” prices, like $5 or $10, which are loosely connected with costs and nearby offers, but never look like empirically grounded. Generally, these services prefer rules of thumb to experimentation. Maybe that’s a miss, since when the monthly fee is way below the hourly wage, demand is expected to be inelastic, so revenue opportunities must be around.

And yes, that’s possible because the IT industry is basically many monopolies complaining a lot about competition which isn’t there.