Some Economics of iPhone Pricing

With the introduction of the new iPhone, Apple has changed its revenue model quite significantly. With the first iPhone, the shared revenue model for Apple/ATT was a relatively high price for the phone and “reasonable” monthly rates for the cell service itself. Apple sold the phones themselves or to ATT, who likely marked them up quite significantly to make a margin on their sale, and Apple also got a cut of the revenues from the cell phone plans. That was an interesting revenue model, and a departure from the industry standard, which is characterized by the cell phone carriers subsidizing the purchase of handsets by consumers and no revenue sharing with the handset manufacturers. The new iPhone is being advertised as "twice as fast, and half the price," and the overall Apple/ATT pricing model is more like the industry standard, with ATT taking a loss on the sale of the phone itself, while no longer paying Apple any revenue from the monthly charges.


Why did they switch to the industry standard revenue model, with ATT now subsidizing the phones and Apple not getting any of the revenue stream from the monthly charges? To answer this, I think the best thing to do is to consider why the industry standard model makes sense for anyone. The answer to this was really given in an earlier post of mine on the economics of iTunes/iPod pricing: see here. Interestingly, the essential pricing features of the iPhone/cell service bundle are not different from that of the iPod/iTunes. Of course, one big difference is that with the iPod/iTunes bundle, Apple owns both parts, whereas with the iPhone/cell service we have Apple and ATT each controlling one part of the bundle. More on that issue at the end of this post. For now, let's think how we would price the iPhone/cell service bundle if we controlled both parts.

With the iPod example, I argued that the optimal pricing scheme would be a relatively low price on the iPod and a markup over marginal cost on the songs. The rationale is based on price discrimination across different classes of consumers. In a perfect world, you would want to set the price of songs at marginal cost to get all consumers to buy the "efficient" number of songs, generating a lot of what we call consumer surplus. Then you would make each consumer pay for the hardware, the iPod, exactly what that stream of consumer surplus is worth. This scheme would maximize profits from the bundle. Of course, we don't live in a perfect world, most notably, Apple cannot figure out what each consumer values the bundle at and make them pay exactly that. If we take the simple case where Apple can charge only one price for the hardware/iPod, then they will want to price the iPod "low" and price the songs at more than marginal cost. Such a pricing scheme manages to get even low-value consumers to buy the iPod, while getting a lot of revenue from those consumers who will purchase a lot songs.

These ideas extend naturally into the cell phone market. An integrated handset/cell service provider would want to price the handsets "low" to get the low-value, price sensitive consumer into the market (there are a lot of such folks) while extracting significant revenue through the monthly charges (which are clearly priced way above marginal cost).

There is a further reason for this "subsidized" handset cost model in the cell phone market that is not present with iTunes and songs. In the cell market, you can offer consumers a lot of different monthly plans, thereby further segmenting and discriminating across the consumer segments. Again, think that your basic objective is to get all types of consumers into the market, charging each type as close as possible what they are willing to pay for the bundle of the handset and the service. One of the key constraints you face in doing such pricing is cannibalization -- if you offer one type of consumer a low overall bundle price, your high end consumers might find that kind of scheme attractive as well, and you will lose the ability to extract maximum revenue from those high end consumers.

If you charge a high handset price, this cannibalization problem will be real tough. With a high handset price, the only way to get the price sensitive consumers into the market is to offer some monthly plan at a very low price. But such a plan might also be attractive to your high end consumers. With a subsidized handset price, you can charge relatively higher monthly prices and still get the price sensitive consumer into the market, and you can charge even higher monthly fees to the folks who are going to talk a lot and put a real high value on that service.

For other analogies, think of IBM in the old days pricing its computers, or copy machine pricing. IBM used to lease its mainframe computers at rockbottom prices, extracting revenue from consumers' use and purchase of the punch cards (Ha! Anyone out there remember those?) Copy machine pricing is similar. A lot of restaurant franchise revenue models also use this. Individual restaurant franchise owners must pay a lump sum for the franchise, but then they also pay a markup on their purchase of ingredients. You get better overall revenue extraction by keeping the franchise fee "low" and using the purchase of ingredients to essentially extract differential value based on usage. Cell phone pricing is very similar.

So the industry standard model makes sense, and Apple/ATT have moved to it. Why did they not take it on in the first place? Mistake? Maybe. Or maybe it was related to a dispute between the two companies. There is tremendous opportunity for disagreements between Apple and ATT in their joint venture. Some disagreements will just be over different beliefs about market size and growth. When two parties to a transaction differ in their beliefs about value, you will often see contingent pricing contracts: we disagree over the value, so instead of one or both of us having to take a risk, let's make the price contingent on actual value to be determined later. That is essentially what the revenue sharing did, giving Apple more revenue the more customers that would sign up. Perhaps Apple was putting its money where its mouth was, showing ATT that it was willing to take the risk of not signing up nearly so many customers as Apple believed they would.

These ideas on possible disagreements between Apple and ATT raise an interesting question. One of the beauties of iPods/iTunes is that Apple owns both the hardware part of the bundle and the "consumables" part of the bundle. That minimizes disagreements and distortions in the pricing and delivery of the overall bundle. (The integrated ownership by Apple of the hardware and operating system in their core computer market is another example of how well integration can work.)

So would the joint value of the iPhone and cell service be greater if they were both owned by the same company? Should Apple and ATT get together much more formally? Interestingly, the market capitalization of Apple is about $150 billion while the market cap of ATT is only $1.2 billion. And Apple has cash on hand of around $9 billion (end of March 2008).