Hey Apple! Apple’s App Store Controversies
This week, a new email service, Hey.com, had their app halted (and maybe potentially removed as it had been previously listed by what Apple has called ‘an error’) because they do not offer the ability to purchase the service through the app. Without paying for the app, the Hey service is useless. Thus, Hey offers a subscription fee of $99 per year on their own site. Hey don’t want to use Apple’s payment service as Apple takes a 30 percent cut of the subscription in the first year and 15 percent thereafter.
There are several issues going on here. The first has been a persistent claim that Apple’s processes for app approvals have been confusing. This goes beyond payment issues but also includes content as well. The second is that Apple invests in the app store and also in software so that apps can be written. Any app that is free or is paid for off-site, does not contribute to those costs directly. Finally, Apple’s policies are not unusual. For instance, Google has the same policies through its Google Play store through which most Android users install apps.
The Hey issue seems to have blown up in discussions of developer discontent. This has happened at the same week that the European Commission has launched an investigation into Apple for antitrust violations with respect to App Store policies. These are mostly in the back of complaints by Spotify. (I wrote about those previously here). It seemed timely, therefore, to take a quick look at some of the basic economics of app store pricing. (Also something that I did research on some time ago).
What does Apple’s in-app payment option do? From a user perspective, it reduces transaction costs. Assume that these costs are t (> 0) if you buy a subscription or extra app functionality elsewhere and are 0 on the App Store. These costs are not simply convenience but lots of other things that Apple (and Google for that matter) require that, from an economist’s perspective, reduce transaction costs. Things like reminders when subscriptions are going to renew or when free trials run out that are the very sort of things that might otherwise cause consumers to pay for apps they don’t want. Of course, developers may want them to pay for those which is another type of conflict. Either way, they are a cost to consumers so t captures that.
Apple charges developers 30 percent of revenue (mostly but with subscriptions, it is a little different) if these purchases are in-app and nothing otherwise. Let this share be a variable, a. So if p is what consumers pay through the App Store and d is what they pay from developers directly, the developer gets (1-a)p if the consumers buy in-app and d otherwise. Note that there is no requirement (anymore) that p = d and that developers charge the same price via both routes. (An example of this is LinkedIn).
For the most part, developers use the Apple in-app system and set p = d which is the equivalent of them going ‘all in’ on Apple. But there are exceptions. Spotify and Netflix, for instance, opt not to have the in-app route at all. This is what Hey wanted to do.
Why couldn’t they do this when Spotify/Netflix could? The answer that Apple (and Google) give is that these services are ‘readers’ of content that can be consumed elsewhere. By contrast, Hey is not really that and is a service that both reads and writes and other stuff. There is an economic difference. For Hey, the cost of use is approximately zero for each user. For Spotify/Netflix, the cost of use is positive as they have to pay content providers. Thus, to require Spotify/Netflix to pay a 30 percent share of revenue when they have significant marginal costs is going to potentially make it difficult for them to cover those costs.
Just to put a finer point on that, note that for (1-a)p > c requires that a < (p-c)/p. While if c were close to zero, this would hold for a = 0.3, there is a good chance it may not hold when c is significant. As “reader” apps generally have a higher c, carving out an exception for them makes sense.
This, of course, doesn’t tell us whether Apple banning apps like Hey that want the Spotify/Netflix option is a good or bad idea. Apple’s claim is that they don’t like it when apps have a bad experience for users one of which could be downloading an app and not being able to use it. Of course, if Apple allowed links directly to the developer’s site for payment options, that experience wouldn’t be as bad. But broadly speaking the transaction costs would still be there. There is at least a theory that Apple believes that having low transaction costs as much as possible is good for its ecosystem and developers who use its in-app purchases. However, I do wonder if there is room for more options.
To see what the constraint is, note that, when payments occur outside of Apple, Apple does not get a 30 percent revenue cut. What it does get is the price consumers pay for its hardware. If Apple knows that it is going or receive app store revenue from each user, it can reduce the price of the hardware accordingly. In other words, if Apple push developers to pay it a cut of revenue, Apple can reduce iPhone prices. If they do not, then iPhone prices will be higher. If all consumers were the same, there would really be no difference. It would just change the flow of payments. But if consumers are different, then we need to think about where deadweight losses might emerge. In that situation, one can imagine an argument whereby people only pay for apps who are otherwise more wealthy than the average and thus, taking a revenue cut and lowering iPhone prices leads to a lower deadweight loss than the other way around. Put simply, seen in this way, the 30 percent cut is a form of price discrimination. It is, possibly, an iPhone discount for customers who do not pay for apps.
Which brings us back to Hey. It is a premium email service. If Apple do not require a payment from apps like it, this is wiping $30 a year from the life-time value of customers who buy an iPhone and use Hey. If half of all iPhone users would buy an app like that, not requiring a payment means that, on average, Apple does not receive $15 a year from the life-time value of consumers and, using a 2-year upgrade cycle, that is $30 on average from each phone. So it is not about Hey itself but about the ecosystem. And that App Store cut adds up.
This also explains why Google has not decided to opt for different policies. It could do so but that would mean that it would reduce earnings from Android licensing which, in many respects, is more important to it than to Apple because it doesn’t make its own phones. If they dropped the 30 percent fee, they would have to increase Andriod licensing fees which would increase prices of phones for everyone.
That said, there is an aspect of Apple’s behaviour that is troubling. There are reports that developers are afraid of publicly complaining about Apple’s rules for fear of retribution. That should disturb us and it should disturb Apple. Apple should want everyone, customers, developers etc, to be satisfied in its eco-system. If that is not the case, it indicates a significant problem. Voice is the mechanism by which large companies can be held to account on product quality. If it is being stifled, that belies our assumption a company is committed to quality.
