Fake Steve Jobs unlovingly refers to those who base their business models on giving stuff away for free as ‘freetards’. I just watched this video of Chris Anderson, editor in chief of Wired and author of blog and book The Long Tail, talk about ‘free’ as a business model. Chris is writing a book on this topic that will come out early next year. Now, call me a picky economist, but there’s a gigantic difference between free and cheap. The things that Chris talks about becoming ‘effectively free’, such as data storage, bandwidth, and processing power, are never going to be completely free. They are relatively cheap and getting cheaper, but will always have a non-zero price. This obviously makes quite a significant difference from an economic point of view. I think Chris understands this, but I guess a book titled Cheap is not going to sell as many copies as one titled Free.

Anyway, there are a few cases where giving something away for free can make sense from a business point of view, and these cases may become stronger when the cost of underlying inputs gets cheaper. One example is where you sell a system of complementary things to the same consumers (razors and blades, mp3 players and music, etc). It may make sense to give away one of the complementary components if you can charge for the other(s) and sell more of the system as a whole. The other is where you have a two-sided platform that generates value by bringing two groups of consumers together and for some reason you find it optimal to give away one side for free while charging the other group.

The latter strategy is what a lot of the ‘free’ business models that Chris talks about are based on, like Google search or Gmail. Google doesn’t charge you for performing searches or using Gmail because it makes money by selling advertising to you. Although Chris doesn’t call it as such in his talk, it’s a two-sided platform with web users on one side and advertisers on the other. Google has decided that it can make the most money by giving search and email services away for free and earning its revenues from advertising and so far it seems to be working out pretty well for them.

One idea of Chris’s that I like is what he calls “every abundance creates a new scarcity” (although ‘every’ is a pretty strong claim to make). For example, he says that bandwidth and storage and so on are becoming cheap, but attention is becoming increasingly scarce. Call me picky again, but this isn’t quite right. The available amount of attention isn’t becoming more scarce, but the reduced costs of distributing content to people have made competition for attention more intense. From an individual content creator’s point of view, attention is becoming scarce because she faces more competition from other content creators for the same limited amount of attention as before, rather than attention itself becoming more scarce. Anyway, it’s an interesting insight that as the cost of certain inputs go down in the content business, competition for attention goes up. From a business point of view, this could be bad. Reduced costs are good, but competitive entry is not good for profits.

An idea of Chris’s that I didn’t like is “if the marginal cost of something is close to zero, why not treat it as zero and give it away?”. That’s a guaranteed way to make a loss, unless you have complementary things that you can sell to raise revenue to cover those “close to zero” costs. And I also think he’s underestimating the real costs of providing services like Gmail. Storage may be cheap (US 40 cents per gigabyte as he says), but providing a couple of gigabytes of email storage plus an email service to a lot of people is still going to be a costly business overall. For example, suppose Gmail has 10 million users, each of which are allotted 2 gigabytes, but each user on average only uses 10% of their storage space. Then you need about 200 megabytes of storage per user, but you’ll need to double that because you need to keep a backup of every bit of data. So you need 4 million gigabytes of storage, which is going to cost US$1.6 million. That’s peanuts to Google, but still significantly more than zero, and that’s just the cost for the hard drives. Then there’s the servers to run the service, the bandwidth, a place to put the servers, electricity to run them, technicians to operate them and develop the site, and so on. And this is all only on the loss-making side of the business. On the other side (selling advertising), you need a mechanism for selling ads, billing, collecting debts, and so on. All this sounds far from free to me, and I’m just wildly guessing, but providing email service to 10 million customers could well cost $10 million per year, which is a decent amount of advertising to have to sell. Cheaper storage and bandwidth have improved the economics of the online email business, but it doesn’t guarantee that providing online email for free is going to be money-making, nor does it mean that an ad-supported ‘free’ model is the only way to go.