The Real Cost of App-Store Dependence
The subject of this final synthesis is Apple’s App Store as a gatekeeping system for digital goods and services, and the question is what developers and end-users lose when one company controls the route from app access to payment and support. The answer is larger than the commission. This series supports a more precise judgment: steering can reduce dependence only when it opens the whole customer journey, leaves developers with a workable economic choice, protects users without turning safety into a blanket deterrent, and produces evidence that the relationship has actually changed.
The sources support a strong claim about the shape of the problem, but not an automatic verdict about its outcome. The CMA’s proposed UK Steering conduct requirement is provisional. The European Commission’s report describes enforcement and compliance work that was still continuing. The evidence includes stakeholder submissions, redacted figures, provisional assessments, and observations from different jurisdictions. A responsible conclusion must therefore separate what Apple controls, what a remedy could change, and what would still need to be measured.
Dependence is a bundle of control
The visible price is Apple’s commission. The deeper cost is the loss of options around that price. For developers of digital goods and services, the CMA says Apple generally restricts both the use of other payment systems inside the app and the ability to tell users about offers available elsewhere. Those restrictions operate at different points, but they reinforce one another.
The App Store is the route to a large and distinct group of users. That reach makes the store more than a convenient sales channel. A developer may have a website and still depend on Apple because abandoning the App Store would mean losing access to users who have chosen Apple devices and expect native apps there. The CMA reports large UK ranges for apps, developers, and monthly users, as well as a redacted £0–5 billion range for annual digital-goods spending, although several figures are redacted or presented only as ranges. The exact count is less important than the bargaining condition: a channel that is difficult to replace can impose terms with less competitive pressure.
The first cost of that condition is reduced distributional choice. A developer may be able to imagine another route but not use it at a comparable scale. The App Store’s reach also creates a two-sided dependence. Developers make the platform more valuable by supplying apps and services; users make it more valuable by bringing demand. Apple’s control over the gateway lets it mediate that exchange and set conditions for participation.
The second cost is reduced communication. A developer that cannot tell a user about an outside offer cannot make that offer compete at the moment of choice. The customer may never know that another price, subscription structure, or payment method exists. The restriction therefore affects the market before money changes hands. It prevents comparison from entering the relationship.
The third cost is transaction and payment dependence. Apple’s In-App Purchase system is not simply one processor among many when alternative systems cannot be used for the relevant digital transaction. The CMA explains that Apple acts as the merchant of record for IAP, taking on financial and legal responsibility for the sale. That can be valuable. It can simplify payment and give users a familiar process. But it also means that the developer cannot freely decide which provider should process the transaction or which commercial arrangement best fits its service.
The fourth cost is relational. The business that designs a digital service may not control the billing, refund, or support interaction that follows a purchase. The CMA records developer evidence that restrictions on steering can prevent direct relationships with customers, limit access to customer information, and leave developers with less control over refunds. A developer may remain responsible in the customer’s eyes for a failed payment or cancellation while lacking the information and authority to resolve it directly.
This is also an information cost. A developer that receives less information about who is buying, how offers perform, or what problems customers face may have a weaker feedback loop. It may be less able to tailor a service, coordinate support across platforms, or test a new pricing model. Direct customer contact does not guarantee good decisions or justify unlimited data collection. It does give the developer more control over the feedback needed to improve its own product.
The fifth cost is a constraint on product design and competition. The CMA says developers reported limits on pricing structures, distribution choices, digital and physical bundles, cross-platform support, and personalization. It also records concerns that developers competing with Apple’s first-party apps can be weakened when they must pay a commission and cannot direct users to a lower-cost route. A rival may have less money to invest, less ability to differentiate an offer, and less control over the customer relationship than the platform’s own products.
The sixth cost is the exclusion of potential service providers. Alternative payment processors and merchant-of-record services cannot compete for a role that Apple reserves to its own system. This affects more than the developer’s invoice. A provider with a different fraud system, account model, support service, or pricing structure cannot test whether it would create value for developers and users. The absence of that competitive test can keep the platform’s terms higher or less responsive than they would be in a contestable market.
These costs are not all measured with equal strength. The restrictions themselves and the CMA’s proposed concerns are directly described in the consultation.