Apple is stuck in its own revenue trap

The "Hey" e-mail app got a bit of attention this week. Made by BaseCamp, and published on iOS (and therefore subject to the terms and conditions of the iOS App Store), Apple rejected version 1.0.1, claiming that the app did not meet its guidelines. Two aspects made this rejection notable: version 1.0 was approved (and version 1.0.1 is minimally different), and Apple decided to "clarify" its terms and conditions after many people complained that the app was, in fact, in compliance with the published terms and conditions. (Apple's clarification was that certain rules apply to "business apps" and different rules apply to "consumer apps", and that the "Hey" e-mail app was out of compliance because it did not provide Apple with 30% of its revenue.)

Lost in the noise about the "Apple tax" and clarity of terms and conditions and consistency of rulings on said terms and conditions is an aspect of Apple that we may want to ponder.

Apple justifies its 30% cut of in-app revenue by offering the platform and services.

iOS is a capable platform. It does a lot. One can argue that the 30% rate is too high (or too low). One can argue that Apple holds a monopoly on apps for iOS.

I want to think about something else: Apple's model of computing, which allows it to justify the "tax".

Those services assume a specific model of computing. Not cloud computing, not web services, not distributed computing. A model of computing that was dominant in the 1970s (when Apple was founded) and the early 1980s. The model of local computing, of personal computing.

In this model, apps run on phones and applications run on laptops and desktops (and the Mac Pro tower). Apps and applications communicate with the user through the user interface. Everything happens on the local device. For Apple iPhones and MacBooks, computing occurs on those devices.

Compare that model to the model used by Google's Chromebook. In that model, the Chromebook is a simple device that sends requests to servers (the cloud) and simply presents the results. (IT professionals of a certain age will recognize this model as a variant of the 1960s timesharing, or IBM's terminals to mainframes. Both used simple terminals to invoke actions on the remote system.)

Back to Apple.

Apple must keep this model of local computing, to justify their take of revenue. They cannot move to a Chromebook model. If they did, they would lose their reason for the 30% tax. Developers are angry enough now at Apple, and while some decline to write for the iOS platform, many others "pay the tax" albeit grudgingly.

But what happens when computing moves to the cloud? A cloud-based app does little on the phone. The computing is on the servers. The app, at best, presents a UI and sends requests to the servers. Is the UI and an HTTP stack enough to justify the 30% "tax"? It's my opinion that such a simple system does not, and therefore Apple must keep apps in the older model of local computing, in which an app uses many services.

Apple has built a nice operating system and platform with its iOS, and it has built a trap with its App Store and 30% revenue cut. Apple is loath to give up that revenue. To keep that revenue, it needs to provide the services that it proudly hawks.

So, as I see it, Apple is stuck. Stuck with local computing, and stuck with a relatively complex platform. I expect Apple, for at least the short to middle term, to stay with this model. That means that apps on iPhone and iPad will stay in the local computing model, when means that they will be complex -- and difficult to support.

In the long run, I think Apple will move to a cloud-based model of computing, but only after everyone else, and only when Apple starts losing business. It will be a difficult transition, and one that may require new management of Apple. Look for a run of quarters with disappointing earnings, and a change in leadership, before Apple changes its App Store policies.