Explaining the disproportionate impact of Apple’s 50-cent fee in response to the EU Digital Markets Act

The first thing to clarify is that the question is not whether developers should be able to free-ride on Apple’s intellectual property rights. However, Apple itself had a certain business model for the Mac where developers would pay for development tools (a market in which there should also be a possibility for third parties to compete, though the operating system maker will be in the pole position) and can then make their apps available directly to users. Developers of iOS have to pay Apple not only an annual fee (as opposed to Google’s much lower one-time fee) for membership in its developer program but also on the distribution side. And if Apple had to actually enforce its patents and copyrights, it would face significant challenges (such as interoperability privileges under statutory law or case law).

This article is not going to discuss what amount would be fair and reasonable, but to explain why Apple’s new EU terms thwart the DMA by preventing market entry, and why a reasonably strong case can be made that they are discriminatory. Under the DMA, Apple must provide access on FRAND terms, and inother contexts (such as standard-essential patents) the ND part of FRAND is much easier for the courts of law to judge than the FR part.

Apple’s focus on 99% of developers instead of 99% of what matters

When Apple announced the new terms, it argued that 99% of developers would not lose or even save money if they opt into the new terms in the EU, and for less than 1% of developers it would be better to keep the old terms (which they can if they want, though they will then not be able to distribute through alternative app stores).

This is an old pattern: Apple’s “small business” program and a “small developer” class-action settlement were also designed to provide potential benefits to a large number of developers that collectively account for a small part of app revenues. But the latter is what actually matters to end users.

That focus on a percentage of developers doesn’t make sense when a company normally argues it puts users first (as did Tim Cook in his Epic Games v. Apple testimony) and doesn’t even care about developer satisfaction (as was revealed by the same testimony).

Apple won’t charge the 50-cent fee if developers opt for the new terms with respect to apps that are installed by less than 1 million users during a given year in the EU. The vast majority of developers never reach that number. And even if one exceeds it marginally, it means the fee remains negligible. But in that highly concentrated business, it’s only a small number of developers that really count.

It’s not just that the 99% of smallest developers don’t have a problem because they don’t reach, or don’t significantly exceed, the 1-million threshold. There’s also a large number of apps that don’t offer in-app purchasing of the kind that Apple taxes: Amazon’s online retail business is an example.

Even for the little guys, the per-install per-year fee is risky

Let’s assume a developer historically reached, say, 10,000 or 100,000 users per year in the EU. And let’s also assume that this app does generate revenue from in-app purchasing, which Apple would tax.

It would seem a no-brainer that this developer would just opt into the new terms for the EU: the percentage, depending on annual revenue, goes down from 30% to 20%, or from 15% to 10%. So, why not? There still is a problem and that’s the unpredictability of what will happen. In the event of a sudden surge, which could be triggered by a YouTuber talking about the app, the download figures could get out of control. And then, unless the app has a very strong business model, the little guy might be ruined.

That is a risk that the little guys don’t have under Apple’s old terms or elsewhere (Android, PC, Mac, video game consoles, whatever).

Comparing 50 cents per install per year to realistic revenue expectations of freemium apps

At first sight, an amount of 50 eurocents (US$0.54) per install per year may seem low. Some might think that this is what Apple should just be entitled to because of its intellectual property rights involved.

It’s just that freemium apps typically get only a limited part of their user base to pay. For example, 2%. In that example, Apple’s fee would be paid for 50 times as many users as those who will actually make a payment. 50 cents than becomes 25 euros per paying customer.

App makers don’t publish their lifetime value (LTV) figures, but an article by online marketing firm Udonis states some “benchmarks”:

• hyper-casual games: $0.20 to $0.40
• casual games: $1 to $3
• mid-core games: $2 to $5

Note that LTV is not annual revenue but lifetime. And even if a user may not play a hyper-casual game anymore, it will get updated unless they deleted it, and Apple will charge the 50 cent the next year. Developers could prevent that only be not updating their app at all, but that would then affect all users and not just the ones who aren’t really using it anymore.

Alternative app stores are hit really hard

More than anything, the 50-cent per-install per-year fee is designed to prevent the emergence of alternative app stores apart from “adult content” and maybe some niche markets. And alternative app stores will be the only way to get apps to iPhone users in the EU other than Apple’s App Store, as Apple doesn’t allow the direct installation of apps from the web, but merely calls installation of apps through alternative app stores sideloading.

Alternative app stores need a storefront app that will also have to pay those fees. But the bigger problem is that the overall set of terms makes the following things practically impossible for alternative app stores:

• to attract free apps that don’t do in-app purchasing (for example, storefront apps of online retailers, as they can keep Apple’s old terms and pay nothing);
• to attract freemium apps that don’t have a very high LTV; and
• to create any win-win between app store and developers, or win-win-win between app store, developers and consumers, as there won’t be enough of a margin.

The first two bullet points are key because even if an alternative app store may not make any money by distributing free apps, and very little by distributing freemium apps that don’t have a high LTV, those apps would nevertheless enrich the app catalog.

The last bullet point requires a bit of an explanation. If an app developer contemplates distributing through an alternative app store, the 50-cent fee is a given. But if the same developer accepts the 50-cent fee and goes through Apple’s App Store on the basis of the optional new terms, Apple will charge a 20% commission on top of the install fee. Assuming that an alternative app store needs a commission in the 10% range to be profitable, and unlike Apple’s own store it also has the cost of paying for the installations of its storefront app, then there just isn’t a margin that enables a bottom-line benefit. If there was a significant margin, the developer could

• make more money per user (in which case the developer will benefit only if the alternative store reaches most users; otherwise the reduction in volume will more than eat up the higher margin);
• pass the savings on to users as lower prices mean higher volume; or
• a mix of both (split the gains of trade).

Here, there will be no significant benefit, and that’s why developers will usually decline to distribute through a new app store. What will furthermore dissuade them from doing so is that if the experiment failed and they’d have to return to Apple’s App Store, their users would have to uninstall the app they downloaded from an alternative app store and then find the same app on Apple’s App Store and install it from there. Some users wouldn’t do that, or they’d try but get lost somewhere along the way. That’s just one example of non-financial terms here that will furthermore up the ante for alternative app stores.

So who’s going to benefit from alternative app stores? Hardly anyone other than the “adult content” industry

Alternative app stores will effectively become an interesting option only if an app, or a category of apps, would otherwise not be available on the iPhone at all because of Apple’s App Store rules.

There are apps that Apple doesn’t allow. Epic Games announced that they would bring Fortnite’s iOS version back in the EU through an app store of their own, but that won’t be possible if Apple takes the position it terminated (lawfully in the opinion of the U.S. courts) Epic’s developer agreement. Apple should not get away with that in the EU, but some public or private enforcement may be necessary and would take time.

While Apple distributes apps and books promoting bogus medicine (homeopathy), they take issue with some other health-related stuff, such as vaping. They don’t allow adult content (to put it that way) as John Gruber mentioned on his Daring Fireball blog. That new opportunity for native iOS adult content apps may in the end be the only economically meaningful result of Apple’s new rules (explanation on X (formerly known as Twitter)). The economics in terms of per-install revenue would enable such apps, unless the makers of such products felt that web apps (which are so far the only way they can serve iOS users) are good enough for their purposes.

EU policy makers obviously had something in mind that would open up the market across all app categories. And that’s simply not going to be the case for years to come. Apple is prepared to duke it out in court with the European Commission. In fact, it’s already doing so over the gatekeeper designations, and for now Apple accepts only the designation of the iOS App Store but disputes that the macOS, watchOS and iPad app stores are included. With respect to iPad, that position is unreasonable; it does appear very reasonable to treat macOS differently; and watchOS is in a gray area where reasonable people could come on either side of the argument.

Discrimination theories

Alternative app stores won’t be able to offer app makers a set of terms that comes without the 50-cent per-install per-year fee. That is a form of discrimination or self-preferencing.

One could furthermore argue that Apple’s new terms (which are the combination of the new terms with the option to keep the old ones) are inherently discriminatory because app makers are taxed only if they want to distribute paid apps or generate in-app purchasing, but online retailers, ride-sharing services and others make a lot of money on iOS without paying Apple anything.

Then there is also a potential argument that involves discrimination and market concentration: the per-install per-year fee favors those who incorporate a wide range of features into a single app. That is an issue in the Brazilian antitrust investigation of Mercado Libre’s complaint: Apple doesn’t let them offer Disney content to those who subscribe to Mercado Libre’s equivalent of Amazon Prime. Amazon is a way bigger company and can give people everything in one app; Mercado Libre needs to partner with Disney, which works for them on Android, but Apple’s rules don’t allow it. Here there’s now a similar situation that favors conglomerates like Amazon over everyone else.

There will be complaints, and the European Commission is far from unlikely to take action. But it will take time, and in a “you win some, you lose some” scenario, Apple would likely defend its rules to an extent that there still wouldn’t be competition in iOS app distribution. Apart from adult content apps…