The Rise of the App Store: Platform, Power, and the 30-Percent Tax

Zusammenfassung

On July 10, 2008, Apple opened the App Store — a single, curated marketplace through which all software for the iPhone could be purchased, downloaded, and installed. What followed was one of the most significant economic transformations in the history of software: a new distribution model that bypassed every existing retail channel, created a developer class of millions, generated hundreds of billions of dollars in revenue, and concentrated platform control in the hands of two companies — Apple and Google — to a degree that regulators would spend the next decade trying to understand and constrain. This article traces the App Store from its origins in Steve Jobs’s initial resistance to third-party apps, through the explosive growth of a new software ecosystem, to the antitrust battles and regulatory interventions that revealed the structural tensions in a model where the platform operator is simultaneously marketplace owner, rule-setter, payment processor, and competitor.

Steve Jobs Did Not Want an App Store

The history of the App Store begins with a refusal.

When Apple announced the iPhone in January 2007, Steve Jobs made no mention of third-party applications. The iPhone shipped in June 2007 with no mechanism for external developers to write software for it. Jobs’s public position was that developers could build web applications — sites optimized for the iPhone’s browser — and that this was sufficient. At the WWDC keynote in June 2007, he described web apps as “the sweet spot” for iPhone development.

The reasoning was not frivolous. The iPhone was a new platform, running a stripped-down version of Mac OS X on hardware with tight memory and battery constraints. Apple’s engineers had spent years wrestling with the stability and security implications of third-party code on Mac OS X. A web-based approach kept all execution sandboxed in the browser, preventing third-party code from crashing the phone, draining the battery, or compromising user data.

But developers and enterprise customers pushed back. Web applications could not access the device’s sensors, contacts, or camera. They could not run in the background or send notifications. They required network connectivity. The gap between what web apps could do and what native applications could do was enormous, and users who had grown accustomed to native software on Palm and BlackBerry devices noticed it immediately.

Within Apple, the debate continued through the fall of 2007. The most vocal advocate for native third-party apps was Scott Forstall, who ran the iPhone software team and had repeated heated arguments with Jobs about opening the platform. The chief opponent was Jobs himself, who worried that policing third-party developers would overwhelm his team and that poorly designed software would degrade the user experience. Board members and executives eventually pushed Jobs to relent; when he did, his conditions were strict: no applications would run on the iPhone without Apple’s explicit approval, and all applications would be distributed through a single channel that Apple controlled.

The iPhone Software Development Kit (SDK) was announced in March 2008. The App Store opened July 10, 2008, alongside the iPhone 3G. On its first day, it offered 500 applications. Within three days, it had recorded 10 million downloads.

The Economic Architecture: 70/30

The App Store’s financial model was simple and consequential: Apple would take 30 percent of every transaction, and developers would receive 70 percent. The model was borrowed from music licensing — the same split Apple had negotiated with record labels for the iTunes Music Store — and it applied to everything: paid app downloads, in-app purchases, subscription fees, and digital goods.

The 30 percent cut was not arbitrary. Apple’s argument was that it reflected the value the platform provided: the App Store’s distribution reach, the payment processing infrastructure, the fraud protection, the discovery mechanisms, and the security review process that protected users from malicious software. For independent developers without distribution relationships, retail shelf space, or payment infrastructure, the proposition was genuinely attractive. You built the software; Apple handled everything else.

The initial App Store was a gold rush. Applications that cost a dollar or two could reach millions of users within days. Games that would have required physical retail distribution could be downloaded globally overnight. The barriers to software distribution that had historically limited development to well-capitalized companies collapsed almost immediately.

Ethan Nicholas, a Sun Microsystems engineer, built iShoot — an artillery game — in his spare time and launched it on the App Store in October 2008. By January 2009, it was earning $37,000 per day. Steve Demeter, a developer in Cleveland, built Trism — a puzzle game — and earned $250,000 in two months. These numbers were exceptional, but they were not impossible, and their existence circulated widely, attracting developers to the platform.

Apple’s revenue from the App Store in its first year was modest — the company was taking 30 percent of a new market that was still small. But the trajectory was unmistakable. By 2010, App Store revenue was in the hundreds of millions. By 2020, it exceeded $60 billion. By 2022, Apple’s services segment — anchored by the App Store — was generating more than $78 billion annually, rivaling the revenue of entire industries.

Two Gardens: Apple’s Curation vs. Google’s Openness

The fundamental strategic choice that distinguished Apple’s App Store from Google’s Android Market (later Google Play) was the trade-off between curation and openness. Apple maintained a mandatory review process for every application: no app could be distributed through the App Store without Apple’s approval, and Apple reserved the right to reject or remove applications for reasons that were sometimes arbitrary, inconsistently applied, and disclosed only after rejection. Google required applications to meet published policies but reviewed apps primarily through automated tools rather than manual inspection, and its review process was faster and more permissive. The result was two distinct experiences: Apple’s iOS offered a more consistently curated catalog with fewer malicious or low-quality applications, at the cost of slower developer iteration and more platform control. Android offered faster deployment and more application diversity, at the cost of more malware, more inconsistency, and a fragmented device ecosystem. Neither approach was strictly superior — each reflected a coherent set of priorities. Apple’s model traded developer freedom for user trust; Google’s traded platform control for developer velocity.

Android Market and the Second Ecosystem

Google’s response to the App Store arrived three months later, in October 2008. Android Market launched alongside the first Android device, the HTC Dream (T-Mobile G1). Where the App Store required Apple’s review and approval, Android Market was initially more permissive: developers could publish applications with minimal review.

The Android ecosystem faced structural challenges that Apple’s did not. Android ran on hardware manufactured by dozens of different companies — HTC, Samsung, Motorola, LG, Sony — each with different screen sizes, processors, memory configurations, and software modifications. This fragmentation meant that a developer writing for Android had to test against dozens of device configurations rather than a small, controlled set of Apple hardware. Applications that worked perfectly on a Samsung Galaxy behaved unexpectedly on a Motorola Droid.

Google addressed fragmentation imperfectly and gradually over the following decade: introducing compatibility guidelines, requiring certification for devices that used the Google Play Store (Android Market was renamed Google Play in 2012), and working with hardware partners to standardize minimum specifications. But fragmentation never disappeared entirely, and it remained a persistent cost for Android developers compared to their iOS counterparts.

The Android market also developed different economics. Android’s user base was ultimately larger than iOS’s in raw numbers — Android phones were available at every price point globally, while the iPhone remained premium-priced — but iOS users consistently spent more per capita on applications and in-app purchases. Developers targeting revenue rather than reach prioritized iOS first for years. Only in the mid-2010s did the revenue gap narrow enough that simultaneous iOS and Android launches became the norm rather than the exception.

In-App Purchases and the Free-to-Play Transformation

The first App Store applications were paid downloads. You paid $0.99 or $2.99 or $4.99, and you received the full application. This model, familiar from software retail, was challenged almost immediately by a different approach: the free-to-play model, enabled by in-app purchases (IAP).

Apple introduced in-app purchases in October 2009. The mechanism was straightforward: an application could be distributed for free, and could then offer the user the option to purchase additional content, features, or virtual goods within the application. From Apple’s perspective, in-app purchases were subject to the same 30 percent commission as paid downloads. From the developer’s perspective, the lower barrier to download increased install rates dramatically — and the ability to monetize within the application opened revenue models that were not possible in a paid-up-front world.

The consequences were far-reaching and not entirely anticipated.

Zynga had pioneered in-app purchase mechanics in Facebook games — FarmVille, launched in 2009, generated revenue by selling virtual goods (faster crop growth, decorative items) to players who had started with a free game. The same mechanics translated to mobile. Games like Clash of Clans (Supercell, 2012) and Candy Crush Saga (King, 2012) were downloaded hundreds of millions of times for free and generated hundreds of millions of dollars through in-app purchases, primarily from a small percentage of players — “whales” in industry parlance — who spent heavily.

Loot boxes — virtual containers that could be purchased with real money and contained randomly determined rewards — emerged from the same mechanics. Players paid for a chance at valuable virtual items, with the odds of receiving any specific item typically obscure or deliberately difficult to calculate. The similarity to gambling was not lost on regulators. Belgium classified loot boxes as gambling in 2018 and banned them for children. The Netherlands, UK, and various other jurisdictions investigated or restricted them. Apple and Google responded with varying degrees of disclosure requirements — apps offering loot boxes were required in some markets to disclose the odds of each outcome — but the fundamental mechanics remained widespread.

The transformation in revenue models reshaped the gaming industry fundamentally. Premium games — games that required payment before play — became a smaller and smaller fraction of App Store revenue. The economics of mobile gaming increasingly favored games designed to encourage ongoing spending over games designed to provide a complete experience for a one-time purchase. The implications for game design were significant: games optimized for in-app purchase revenue often differed structurally from games optimized for player enjoyment.

The 30 Percent Under Pressure

For a decade, the 30 percent commission was a cost that developers paid without public complaint. The complaints existed — they circulated in developer forums and conference hallways — but no developer wanted to alienate a platform that controlled access to hundreds of millions of customers.

That changed in August 2020, when Epic Games deliberately triggered a confrontation.

Epic had released Fortnite on iOS in 2018 and had been paying Apple’s 30 percent commission on all in-app purchases — Fortnite’s V-Bucks currency being the primary revenue mechanism. On August 13, 2020, Epic updated the Fortnite app to offer a direct payment option that bypassed Apple’s payment system, charging players 20 percent less than the App Store price — and not paying Apple’s commission. This violated Apple’s developer agreement explicitly and obviously.

Apple responded within hours, removing Fortnite from the App Store and terminating Epic’s developer account. Epic filed a lawsuit against Apple the same day — the suit had clearly been prepared in advance. Epic simultaneously filed suit against Google, which had removed Fortnite from Google Play for the same reason.

The Epic v. Apple case went to trial in May 2021. Epic’s legal theory was that Apple had an illegal monopoly over iOS app distribution — that by requiring all iOS apps to be distributed exclusively through the App Store and all in-app purchases to use Apple’s payment system, Apple had foreclosed competition in the iOS app market. Apple’s defense was that the relevant market was not “iOS app distribution” but “gaming” or “smartphones” broadly, and that within that market Apple faced vigorous competition from Android.

The judgment, issued by Judge Yvonne Gonzalez Rogers in September 2021, was a partial victory for both sides. The court found that Apple did not meet the legal standard for a monopoly under federal antitrust law — the relevant market definition was contested, and Epic failed to establish Apple’s dominance in a sufficiently broad market. Apple prevailed on nine of ten claims.

But on the tenth claim, Apple lost. The court issued a permanent injunction requiring Apple to allow developers to include links in their apps to external payment systems — the “anti-steering” provision that had prohibited developers from even telling users that cheaper purchasing options existed outside the App Store. This provision, the court found, was an unfair business practice under California law.

Both sides appealed. The injunction was stayed pending appeal. The legal battle extended for years, but the case had done something more important than its immediate legal outcome: it had made the App Store’s economics a public and political issue.

Regulatory Intervention: The EU Acts

While the American legal challenge proceeded slowly through the court system, European regulators moved faster and more comprehensively.

The EU Digital Markets Act (DMA), which entered into force in November 2022 and became applicable to designated “gatekeepers” in March 2024, was written with Apple’s App Store — among other platforms — clearly in view. The DMA designated Apple as a gatekeeper for iOS and the App Store, and imposed obligations that the Epic litigation had failed to achieve through American courts.

The DMA required Apple to allow iOS users to install applications from sources other than the App Store — sideloading — and to allow third-party app stores to operate on iOS devices. It required Apple to allow developers to use third-party payment systems without Apple collecting its commission. It prohibited Apple from requiring apps to use Apple’s in-app purchase mechanism.

Apple’s response was to comply technically while minimizing the practical impact of compliance. The company introduced a Core Technology Fee — a charge of €0.50 per install per year for apps that exceeded one million installs — which applied to apps distributed outside the App Store. Critics argued this fee made alternative distribution economically unattractive for popular apps. Regulators began investigating whether Apple’s implementation of DMA requirements was compliant with the regulation’s intent.

The European regulatory intervention was the first time that legislation had directly mandated changes to the App Store’s fundamental business model. Its global significance was potentially larger: if Apple created a more open iOS environment for European users, the pressure to offer the same globally would intensify.

Dead End: Nokia Ovi Store and BlackBerry App World

The App Store did not emerge from a vacuum. By 2008, every major mobile platform operator had understood that software distribution was a critical piece of the mobile ecosystem. What distinguished the App Store from its predecessors and contemporaries was not the idea but the execution — and the failure of competing platforms illustrates why execution mattered more than arrival.

Nokia’s Ovi Store, launched in May 2009, was the largest platform for mobile software distribution before the App Store by raw numbers: Nokia sold more phones globally than any other manufacturer, and Symbian — Nokia’s operating system — ran on hundreds of millions of devices worldwide. Nokia had the distribution, the brand recognition, and the installed base. What it lacked was developer adoption and a coherent user experience.

Symbian development required using Carbide.c++, a development environment that developers consistently described as painful to use. The application signing process required purchasing certificates from a certificate authority, submitting applications for testing, and waiting weeks for results. Applications crashed unpredictably across Symbian versions — S60 3rd Edition and 5th Edition had significant incompatibilities. The Ovi Store’s download experience was slow, the catalog was difficult to navigate, and the payment infrastructure was inconsistently available across markets.

Nokia recognized these problems. The company launched a series of initiatives to improve the developer experience — reducing signing requirements, improving development tools, providing more developer support. But by 2010, the window was closing. iOS and Android were attracting the most ambitious mobile developers, and the application quality gap between Ovi and the App Store was widening rather than narrowing. A developer choosing where to spend six months building an application did not choose Symbian.

Nokia’s situation was worsened by internal organizational dynamics. The company’s engineering culture was hardware-focused; software had long been treated as a necessary complement to handset design rather than a primary product. Reorienting toward software-first thinking required cultural change that moved more slowly than the market. When Microsoft acquired Nokia’s device division in 2014, the Ovi Store was already an afterthought.

BlackBerry App World, launched in April 2009, faced a different but equally fatal version of the same problem. BlackBerry’s installed base was substantial and its users were loyal — the device was dominant in enterprise and among users who valued physical keyboards and email security. The BlackBerry platform had genuine strengths. But BlackBerry’s development environment, like Nokia’s, was significantly less accessible than Apple’s iOS SDK, and the company’s culture was focused on enterprise customers whose application needs were specialized rather than on consumer applications that could drive broad developer interest.

BlackBerry’s larger problem was that its core use case — secure mobile email — was being absorbed into iOS and Android. As those platforms improved their enterprise security capabilities and received IT department approval, the unique value proposition of BlackBerry narrowed. By the time BlackBerry attempted a platform relaunch with BlackBerry 10 in 2013, iOS and Android had established ecosystems so much larger that the comparison was demoralizing. BlackBerry App World had fewer than 100,000 applications when Apple’s App Store and Google Play each had more than a million.

The lesson of Nokia Ovi and BlackBerry App World was not that the App Store model was uniquely clever. Multiple companies understood the model. The lesson was that a software distribution platform’s value was inseparable from the developer experience, the payment infrastructure, the discovery mechanisms, and the hardware quality that surrounded it. Arriving first was not enough. Nokia and BlackBerry had distribution relationships, brand recognition, and installed bases that dwarfed Apple’s in 2008. What they could not provide was the combination of accessible tools, reliable infrastructure, and user experience that made developers choose to build for a platform and users choose to stay on it.

The Developer Class

The App Store created a new category of economic actor: the independent mobile developer.

Before the App Store, software distribution required either a publisher (for retail games and applications), an enterprise sales relationship (for business software), or a sophisticated website with payment processing and download infrastructure. The App Store eliminated all of these requirements. A developer with an Apple Developer Program membership ($99 per year), a Mac with Xcode, and an application could reach every iOS user globally within days of submission.

The implications were initially democratic. Developers who had previously worked within large software companies could strike out independently. Students with strong programming skills could compete for attention against established publishers. International developers in markets without strong domestic software publishing industries gained access to a global customer base.

By 2020, the App Store had generated more than $200 billion in cumulative payouts to developers. The installed developer base numbered in the millions across iOS and Android. The app economy had become a significant sector of the software industry.

But the democratic framing obscured structural inequalities that became increasingly visible over time. The App Store’s discovery mechanisms — search, editorial curation, category rankings — heavily favored applications with large marketing budgets and established user bases. A new application competing for the same keywords as an established application with millions of existing users was at an enormous disadvantage regardless of quality. The App Store’s economics increasingly favored either large companies with the resources to acquire users at scale or exceptionally viral products that spread organically.

The App Store also created a form of platform dependency with few precedents in software history. A developer whose application was rejected, removed, or had its category eliminated by an App Store policy change had limited recourse. The platform’s terms of service were non-negotiable and could change without warning. Apple’s App Store guidelines ran to tens of thousands of words, were subject to interpretation by reviewers who applied them inconsistently, and could reject an application for reasons that did not apply to a competing application already in the store. Developers who built substantial businesses on App Store distribution were dependent on a single chokepoint they did not control.

The Lasting Transformation

The App Store’s most significant contribution was not a specific business model or a particular technology. It was the demonstration that software distribution was a platform business, and that controlling the platform was more valuable than controlling any individual application.

Before the App Store, the dominant mental model of software was: developers build applications, users buy applications, publishers distribute applications. The App Store introduced a different mental model: platforms control access to users, developers pay for access, and platform operators set the rules that govern the entire ecosystem.

This model has spread far beyond mobile software. The economics and control dynamics of the App Store — a single curated marketplace with mandatory commission, content review, and discovery mechanisms controlled by the platform operator — became the template for the gaming platforms (Steam, PlayStation Store, Xbox Games Store, Nintendo eShop), smart TV operating systems, in-car software platforms, and emerging embedded systems platforms that followed.

The 30 percent commission, which felt generous to developers in 2008 when it replaced zero revenue or complex retail arrangements, became a contested norm. As the App Store grew from a developer convenience into an indispensable distribution channel, the commission’s character changed: it was no longer primarily a payment for services rendered but a tax on access to a captive market. The regulatory and antitrust response to that transformation — still unresolved as of 2024 — will determine what the next generation of software distribution looks like.

For the broader transformation of software into services, see The Rise of SaaS. For the infrastructure on which app-based services run, see The Cloud Computing Era.