The Quest for Perfect Price Discrimination

A man reads the online version of Britain’s Times newspaper in London in 2010. The Times became the first British newspaper to fully charge readers to access its stories online.

Photo by BEN STANSALL/AFP/Getty Images

Felix Salmon had a long post Wednesday on state-of-the-art thinking in the field of media paywalls, and it seems to basically come down to price discrimination:

Talking to Mather and MediaPass, it’s clear that their idea of “best practice” doesn’t rely much on meters at all. They have the numbers, remember: they know what kind of walls are best at maximizing revenues, and what kind of walls just end up turning readers away. And crucially, one of the biggest lessons they’ve learned is that it’s a mistake — at least from a purely financial perspective — to treat all readers equally. Some readers have a much greater propensity to pay than others; ideally, you want to extract a lot of money from those readers, while also allowing the vast majority of your visitors — the ones who will never pay you anything — to still consume your content and view the associated ads.


And although readers hate the kind of extreme opacity practiced by the FT, where there’s basically no rack rate and nobody knows what anybody else is paying, from a revenue prospective it makes a lot of sense. The FT knows quite a lot about its registered readers, so it can be quite effective at charging the highest prices to people with the greatest willingness to pay, while charging much lower rates to readers in, say, India.

Take Slate, which, as I’m sure you’ve noticed, doesn’t have a paywall. The reason, basically, is that this isn’t like selling cars, where it costs us a lot of money to obtain each vehicle and thus we need to be compensated for every unit we ship. Each extra click on the site has at least a little monetary value to us as an advertising platform, and the cost of serving an extra is basically zero. So for any ad-selling website on the Internet it’s a minor tragedy any time a customer walks away because he didn’t want to pay.

In principle, the way any site would want to hire this would be by embedding a little gremlin in the brain of each reader and charging everyone their exact willingness to pay. People who’d be happy to read Slate but only for free would get to read it for free. Prosperous people who love Slate would get charged quite a lot. People who love it every bit as much but have less objective ability to pay would be charged less. But some fabulously wealthy people would get to read for cheap simply because the gremlin would credibly inform us that rich as those people are, they’ll gladly pay $11 a year for Slate but not a penny more. It’s either a beautiful dream or a terrifying one, depending on how you think about things. But while it’s never going to happen, it does represent something of an idealized version of how a paywall would work. And it’s why you should never expect to see simple pricing schemes in this field. A successful paywall is going to be based on some complicated considerations. Indeed, complexity itself can be a virtue—a way of delivering a discount to those who are sufficiently thrifty to care about figuring it out.

People are often upset about the weirdness and lack of transparency in the pricing of health care services, but other industries—airlines and hotels most notably—are also based largely on sophisticated price discrimination and at least one theory about the future of media is that this industry will be too.