Google’s DMA compliance plan calls Digital Markets Act’s usefulness into question, one pricing detail ridicules the EU

The European Commission’s competition chief, Executive Vice President Margrethe Vestager, stated in a recent Bloomberg TV interview that mobile app stores were the number one priority in connection with the implementation and, where necessary, enforcement of the DMA (March 6, 2024 games fray article). That makes a lot of sense.

Apple’s collision course is unprecedented, as is the EC’s rhetoric in response. Apple keeps telling users (consumers as well as corporate and government customers) that the DMA is a wholly bad idea. Apple imposes terms that make it extremely difficult for alternative app stores to be viable unless they carry apps that Apple doesn’t want in its own App Store (such as “adult content” and, because of a U.S. contract law issue, Fortnite) or for which Apple isn’t commercially flexible enough (non-fungible tokens).

By contrast, Google’s tone is more constructive, and the Android maker does not go as far as Apple in its attempts to thwart the DMA. But it goes more than far enough to achieve the same “the more things change, the more they stay the same” objective. And there is an insult, albeit a subtle one: Google’s DMA terms are worse than its non-DMA terms with respect to one pricing detail.

On balance, Google’s compliance plan does nothing to deliver a proof of concept for the DMA in its current form. If anything, the fact that a mobile ecosystem gatekeeper can render the DMA useless even by softer means than Apple’s underpins the case for legislative amendment. Not only has games fray been quick to identify the need for that (February 20, 2024 games fray article and March 3, 2024 follow-up), but a German state secretary (vice minister) and former Member of the European Parliament (MEP) has already gone on the record about that possibility (February 29, 2024 MLex report (mostly paywalled)).

Here are a few key observations on the Android app distribution part of Google’s DMA compliance plan, along with an item-by-item comparison with Apple’s approach.

Apparently no forced fragmentation (unlike Apple’s per-app-ID entitlements)

Some criticize Apple’s EU app rules as creating “friction” but that is an understatement. It’s a key objective for Apple to let app makers lose users if they avail themselves of certain liberties afforded to them under the DMA. Losing users is a fundamental problem, as user acquisition costs are substantial. It’s , not just a minor frictional loss or a slightly reduced conversion rate.

The way Apple goes about it is that for certain features, app makers must apply for an “entitlement” for an app. An “entitlement” is essentially a binary data point: it indicates whether an app may or may not do something, such as using other payment options. The problem is that such entitlements are defined at the app level, even if they are relevant only in one or more countries. So if Microsoft subsidiary Activision Blizzard King wanted to offer a version of Candy Crush that makes use of such an option, they’d have to either get all EU users of Candy Crush to download a new app (same name, but new app ID and then with the entitlement attached to it) or the rest of the world (unlikely anyone would do that, given that the EU accounts for only 7% of Apple App Store revenues).

Google’s compliance plan and published rules don’t indicate anything similar. Instead, Google just says that apps may offer certain options only to users within the European Economic Area (EEA), which is the EU’s 27 member states plus Norway, Iceland and Liechtenstein.

No fees on app installs via sideloading or alternative app stores

Unlike Apple, Google allowed alternative app stores even prior to the DMA, just that they never got much traction and the Google Play Store remains a monopoly by U.S. standards or dominant in its market as its called in the EU. For Apple, the fact that the DMA forced them to enable and allow alternative app stores for the first time anywhere in the world was an opportunity to then define new commercial terms. A seemingly low fee of 50 eurocents per install per year makes a disproportionate impact (January 28, 2024 games fray article).

Based on Google’s DMA compliance plan, there won’t be any such charge on Android. It would probably also raise issues with Google’s existing agreements with device makers if it suddenly imposed such a fee.

Same DMA-related terms for alternative in-app payment systems as Apple, but worse than Google’s own pre-DMA terms

Near the top of page 158 of its compliance report, Google explains that “[t]ransactions completed using the developer’s alternative billing system are subject to Google Play’s standard service fee, less an adjustment of 4 percentage points” under Google’s pre-existing User Choice Billing program, but under an EEA-specific program (i.e., Google’s DMA-related terms), there is only “an adjustment of 3 percentage points.”

So the DMA deal for developers is, with respect to the commission, worse than the pre-DMA deal (which they can still opt for) by one percentage point.

It’s a subtlety, but it must give policy makers pause. It shows there’s something wrong with the DMA in its current form that there is no hard and fast rule in the DMA, such as parity or Most Favored Nation clause, that could be enforced against Google.

There is an implicit justification, though:

Under Google’s DMA-specific rules, developers can just offer an alternative payment system because they want users to use it. Under Google’s User Choice Billing program, which applies not only in the EEA but also in some other jurisdictions such as South Korea (where a change to the country’s Telecommunications Business Act mandated such an option), the user will be given the choice to use either an existing Google Play Billing account or an alternative option. So, arguably, if a developer convinces a user to use a different payment system despite Google Play Billing being offered as well, the developer is entitled to a better margin.

Also, some apps are not eligible for User Choice Billing while all apps are for the EEA program.

But the reason for the 4% margin under the User Choice Billing program is simply that payment processing is slightly more expensive in Asia than in Europe, and Google just wanted to see the same discount for all those markets.

In any event, the bottom line is that alternative payment systems are pointless because developers can’t bring down their costs (at least not to a non-negligible extent) and then split the savings with, or pass significant savings on to, users.

Linking out to external purchasing options

This is different from in-app payment alternatives because it’s not just about the payment: it’s about sending users to a website where they can make a purchase. Such websites could, for example, also offer items that are not available via Google Play.

In this context, Apple just makes sure (in the U.S., where Epic Games now has a permanent anti-anti-steering injunction in place, as well as in the EU) that no one saves money in the end. They just deduct the 3% that payment processing services are also going to charge (more or less).

Google’s related EU rules are slightly more elaborate and seemingly more generous, yet designed to strongly disincentivize that app developers opt for those terms:

• First, Google charges an “initial acquisition fee” of 5% for auto-renewing subscriptions and 10% for other in-app purchases (such as purchasing in-game currency or unlocking content) on all purchases made during the two years following the first such external transaction.
• Second, Google imposes an “ongoing services fee” of 7% for auto-renewing subscriptions and 17% for all other in-app purchases.

The difference between the two parts of the overall fee is that after the first two years, a developer could theoretically opt to stop using Google Play services (such as automatic updates), but only if the user consents.

Let’s focus on the standard rate because that’s where most of the money is made (and the logic is the same for those auto-renewing subscriptions, just that the numbers are different). So, after two years Google’s fees go down from a total of 10%+17%=27% to 17% if the developer keeps using Google Play services with a view to this particular app and a given user. And Google’s fees theoretically go away completely if the developer and the user agree to stop using Google Play services.

Mentally, we should always add the 3% that external payment processors will approximately charge. So for the first two years, there’s no saving relative to a Google Play in-app purchase. After two years, the saving will be a reduction from 27% to 17% or theoretically a total elimination of Google charges.

Mobile apps are a dynamic business, and that particularly applies to games, which are the largest segment of the in-app purchasing business. There are some games of impressive longevity that a significant number of users will still be playing not only after 2 years but even after 5 or more. But most of the business is really generated during the first year or two. During that period, an app maker will not save any money by offering an external purchasing option, linking out to a website. However, without having a clear business benefit (as it’s not very likely that a given user will keep making external purchases even more than two years down the road), the app maker is not in a position to offer the user any major incentive for clicking on the external link.

What makes it worse is that once the app maker sends users to external websites, the conversion rate (meaning the percentage of people who really do end up making the purchase) will go down significantly, if not dramatically, from just selling an item to an user within the app.

Therefore, it won’t make sense for app makers to send users to external websites for purchases. Apple’s rules achieve the same effect, just that in Apple’s case it’s easier to see and in Google’s case one has to add a couple of numbers and, more importantly, has to consider the impact on the conversion rate and the fact that most users don’t make purchases in an app over the course of many years.

The difference is just that Google can say: in the long run, app makers will benefit. Google just wants to get paid essentially the same commission (as for an in-app purchase) for the first two years. Then the commission will go down, potentially even to zero if Google no longer provides any services. So Google can argue that there is a theoretical opportunity, just that there’s no near-term return on investment and if one thinks it through, even the long-term deal is based because of the reduced conversion rate.

Is there an enforcement opportunity?

Compared to Apple’s disablement of web apps in the EU and temporary refusal to let Epic Games build an alternative iOS app store for the EU, Google isn’t being confrontational. However, if the measure of success for the DMA is that prices go down and rival app stores gain traction, the net effect is essentially the same. The DMA won’t change anything in the marketplace, even if the mobile gatekeepers make some minor concessions in the further process.

Enforcement of the existing rules against Google will be the same with respect to alternative in-app payment systems, but even harder than against Apple with respect to links to external payment options. And there is no reason to assume that the DMA will create a significant opportunity for alternative Android app stores in the EU. There is nothing with respect to alternative app stores that Google does specifically in the EU that hasn’t already been the case around the world for some time.

It would therefore make more sense to focus enforcement efforts on Apple to the extent that violations can be alleged without an exceedingly purposive interpretation of the DMA, and to look at the combination of Apple’s and Google’s EU app rules as evidence that the DMA should be modified sooner rather than later with respect to mobile app stores.