The European Commission is not happy with Apple’s efforts to comply with the Digital Markets Act (DMA). They have been leaking as much for a while now, mainly to the Financial Times¹, but now they have come out and said so explicitly in a press release (as opposed to a press leak). Here is the core of the EC’s position:

Under the DMA, developers distributing their apps via Apple’s App Store should be able, free of charge, to inform their customers of alternative cheaper purchasing possibilities, steer them to those offers and allow them to make purchases.

Describing the situation this way underlines the fundamental misunderstanding of what drove the success of the App Store that lies at the heart of the DMA. The EC appears to thinks that success is solely due to coercion of developers by Apple, but in actual fact that success (benefiting both Apple and developers) is driven primarily by user experience — which, yes, results in no small part from the constraints which Apple places on developers.

Too much of the commentary frames this situation as a tug-of-war between Apple on one side and developers on the other — but it is actually a three-way-conversation, with users in the mix as well. Sometimes users’ needs will align with Apple’s priorities, and sometimes with developers’, and sometimes they will have their own. But I hardly ever see the users’ perspective discussed, so that is what I want to do here.

A badly scratched iPhone on the ground

Squeezing users

The way the EC has framed the situation is that evil Apple is forcing developers to overcharge users in order to extract maximum profit from that exchange. Now Apple is hardly innocent here; some of the anti-steering rules do get ridiculous. Customers would, all things being equal, prefer a cheaper product. But what if the product they are getting for a lower price is not the same?

The benefit of a single payment system is precisely that it is unique and singular. As an Apple user, I have one place to look for all of my subscriptions. Even better, if I want to cancel one, Apple lets me do that with a couple of taps, with no hard sell or obstructively tedious process. I can also see the same thing for my kids, and share subscriptions (where allowed) with my family.

If there were multiple “purchasing possibilities”, all of that would instantly becomes far more complex and fraught. For a start, I would have to enter my payment details in each and every different payment system, meaning that I would have to evaluate for each one whether I felt it was trustworthy. I would also have to sign up to each individual system’s rules for cancellation and refunds, instead of benefiting from a single clear overall set of rules.

The system as it is has generated the explosion of apps that we see in the current App Store, precisely because it removes complexity and fear from the decision whether to try out a new app or subscription. The counter-example would be Blackberry, which did allow for J2ME apps to be installed from almost any source — and hardly anybody did, because the process was both tricky and scary.

If the EC forces Apple to relax its rules, one result will be that people will be much more reluctant to download a new app or pay for one through an unfamiliar payment portal. This reluctance will probably not be terminal, because the App Store now exists and apps have a certain momentum behind them, but those factors did not spring up fully-formed without any input. They are the result of concerted effort and investment by Apple over a period of many years, and the company is entitled to the fruits of its labours.

Could Apple design their systems differently? Maybe, and they could certainly be a little more circumspect in how they enforce their rules. Most egregiously, during the Covid lockdowns of 2020, many in-person experiences (exempt from paying Apple’s 30% cut) were forced to pivot to digital formats — and Apple demanded a 30% cut from struggling gym trainers in the middle of a pandemic, refusing to back down until Facebook got involved. In fact, I will stipulate that this entire situation with the European Commission could have been avoided if Apple had been just a little bit more flexible. But that’s not who they are, so here we are.

Who pays for apps?

These are the specific areas that the EC propose to investigate:

Apple’s Core Technology Fee, under which developers of third-party app stores and third-party apps must pay a €0.50 fee per installed app. The Commission will investigate whether Apple has demonstrated that the fee structure that it has imposed, as part of the new business terms, and in particular the Core Technology Fee, effectively complies with the DMA.

Apple’s multi-step user journey to download and install alternative app stores or apps on iPhones. The Commission will investigate whether the steps that a user has to undertake to successfully complete the download and installation of alternative app stores or apps, as well as the various information screens displayed by Apple to the user, comply with the DMA.

The eligibility requirements for developers related to the ability to offer alternative app stores or directly distribute apps from the web on iPhones. The Commission will investigate whether these requirements, such as the ‘membership of good standing’ in the Apple Developer Program, that app developers have to meet in order to be able to benefit from alternative distribution provided for in the DMA comply with the DMA.

The almighty Euro

Let’s start by talking about money. To be clear, Apple’s Core Technology Fee applies to apps that have more than one million downloads — and Apple added additional criteria to the CTF to cover hobbyist apps with no revenue, and to avoid bankrupting apps that suddenly cross that threshold with a three-year “on-ramp” of payments.

The relevant background here is that the App Store market is distorted by free apps that monetise without payment from users. These apps are massively profitable, but because their business model does not involve anyone paying for the app itself, they do not deliver any direct revenue to Apple. Is anyone arguing that an Instagram user is not worth quite a lot more than half a Euro?² Netflix used to allow people to purchase subscriptions through their app, but removed that option to avoid paying Apple’s cut — so again, zero revenue to Apple for a pretty major app.

The usual objection is that these “free” apps also have value to Apple — and to be sure, an iPhone without Instagram or Netflix would be less valuable and less successful. However, that situation is not a fact of nature that just arose out of nowhere. Instagram in particular simply could not exist without smartphones, so that between Meta and Apple is a fully symbiotic relationship.

Apple has accommodated these “free” apps on the App Store without extracting revenue directly, but has demanded that they at least abide by the strictures of the platform, despite recurring attempts to evade those controls and abuse users by the likes of Meta and Epic. The CTF is Apple’s way of making sure that these abusive vendors cannot simply leave the App Store, further weakening the controls Apple puts in place to protect its users, without any sort of consequences.

Installing apps outside the App Store

The EC also complains that Apple has made it too complicated to install an alternative (lower-case) app store, or to deploy an app directly from a web site. The question is, just how easy should it be to do those things?

In my misguided youth I ran Windows for a few years, and it was just generally accepted best practice to blow away your entire OS installation at least every twelve months to get rid of all the accumulated barnacles and cruft that came from installing apps. Less technical users would simply refuse to install any new apps or even to run updates because they knew it was dangerous. The iPhone made it safe to try out a new app, because it can do very limited damage in its sandbox, and it’s easy to remove it entirely. Some of those benefits are due to technical constraints in iOS itself, but some come from the App Store.

Given that situation, and the weaker controls that Apple can enforce outside of the App Store, it seems reasonable to warn users that they are about to do something that is technically more dangerous than what they are used to in the App Store. In fact my preferred option for this sort of side-loading would have been to enable it only for developer accounts, or accounts that have specifically accepted some additional limitation of liability. By the time the DMA came around, that would certainly not have been acceptable, but I do wonder what would have happened if Apple had enabled such an option years ago to take the pressure of that particular request from technical power users.

Who’s in charge around here?

The final point is about which developers are eligible for these mechanisms. Developers say, not unfairly, that to require them to have been in the App Store for two years to take advantage of the new distribution mechanisms becomes a chicken-and-egg problem. If the app you want to launch is not allowed in the official App Store, but you need two years in the App Store to be able to launch outside it, you are effectively locked out of ever launching your app.

Apple has been revising those criteria, and will no doubt do so again, but the obvious background is that if Apple were to throw the whole system open, there can be absolutely no doubt that hordes of burner developer accounts would invent whole new categories of abuse, and Apple would be stuck playing whack-a-mole. If instead the vendors offering apps directly from their websites are known to Apple and have a track record, it is easier to ensure that they are not up to more than the usual amount of mischief.

I am not saying that Apple’s proposal is perfect, but it is not unreasonable, and once again, it has benefits for users in terms of how trustworthy they can assume app vendors are.

Integration is the key

All of these EC proposals are, in the words of Ben Thompson on the Dithering podcast, “targeted at the heart of Apple, which is integration”. Apple’s ecosystem is valuable in large part because it is so tightly integrated — but it is integrated precisely because Apple controls all the components. If Apple had to open every one of its features up to third parties, that seamless integration — which users love! — would inevitably be weakened, as the simple calculation of “Watch works with iPhone” would turn into some sort of horrible compatibility matrix, with different vendors, platforms, and even versions of individual products all offering differing levels of support.

In fact, the recently-announced delay to some headline iOS 18 features, which has been interpreted by some as spiteful, can just as easily be parsed as Apple’s fear of interoperability questions. Will the EC require that Apple support SharePlay from iPhone to Windows that is as seamless as it is to a Mac? What about Android to Windows? SharePlay is a competitive benefit that is only available on Apple’s products, but that is because it arguably can only exist in that integrated ecosystem.

Why not both?

Where are we going

Of course Apple is still not immune to shooting itself in the foot, because in the middle of this fracas, the App Store moderation team decided now was the perfect time to ban an emulator app. Apple has a long-standing policy against anything that can run downloaded code, for fairly obvious security reasons; anything that is going to run on an iPhone needs to be vetted first. The developers of the UTM emulator removed just-in-time compilation (JIT) from the build they submitted to Apple, but it was still rejected, on the grounds that a PC is not a games console. Apparently the only emulators that are allowed are the ones that emulate games consoles, like Riley Testut’s Delta, the game emulator that fired the starting gone in the Alt Store race — not general-purpose computers. The fact that Apple is using its control to enforce this rule even outside its own App Store — and doing so at this time — is what is liable to come back to haunt them in the future.

In all fairness, Apple is making moves in the direction of better interoperability, most recently with the launch of AccessorySetupKit to enable what is described as an “AirPods-like setup experience”. The key point — and the reason why I doubt this sort of move will be seen as a remedy by the EC — is that this increased ease-of-use requires the makers of those accessories to sign up to using an Apple-provided framework.

In the other direction, public standards also embrace ideas from Apple’s proprietary frameworks. USB-C has modes that deliver many of the advantages of Thunderbolt, and Qi charging is evolving to be much closer to MagSafe than it started out. I would love for the EC to push these efforts — on both sides — to go further and faster. I just do not think that mandating specific remedies, and then quibbling with how those are presented to users, is going to be effective in terms of providing a better experience for end users.

If the EC gets its way with the DMA, some developers might be able to release apps that would not have been approved by Apple, or using different payment processors. Most of those developers will be massive US companies like Meta or Epic, not struggling indie developers crushed under Apple’s boot-heel. Situations like Delta and UTM are the exception: those apps have tiny numbers of users in proportion. For users, this outcome would mean subscriptions spread across a dozen different systems, each with is own arcane cancellation and refund rules, and no unified view of what they (or their kids) are paying for.

I struggle to see how that is a good outcome for European consumers.

🖼️ Photo of broken phone by Lucas van Oort on Unsplash

I mention this because the FT’s journalist on the “Big Tech vs the EU” beat, Javier Espinoza, seems to have an almost willful misunderstanding of both sides of his topic. He stands out because I generally find the FT’s reporting to be pretty good, although of course this could always be Gell-Mann amnesia. ↩
I can’t find an authoritative public source, but estimates seem to range around an Instagram user being worth around $50/year in revenue to Meta. ↩