Platform Failure Is a Strategic-Identity Crisis, Not a Feature Gap

Based on the research ofConstantinides, Cennamo, and Aaltonen, "The Evolution of Digital Platform Ecosystems: An Identity Domain Perspective," Information Systems Research, 2026

That is the uncomfortable implication of new research from Panos Constantinides, Carmelo Cennamo, and Aleksi Aaltonen, published in *Information Systems Research*. Their study of Apple's iOS and Google's Android from 2007 to 2024 introduces the idea of an "identity domain", the organizationally held sense of what a platform sponsor's ecosystem is *for*, which in turn legitimizes everything from architecture choices to governance rules. Apple's identity domain, they argue, centers on seamless, premium user experience, which justifies vertical integration and a tightly controlled walled garden. Google's identity domain centers on data ubiquity, which justifies horizontal licensing, open scaling, and proprietary control points buried deeper in the stack. Neither company is more "open" or more "closed" in the abstract; each is internally consistent with a different theory of what it is building. The strategic failures of BlackBerry and Windows Phone, the authors argue, are not cautionary tales about hardware or app counts. They are cautionary tales about strategic dissonance, platforms whose moves stopped cohering with any legible identity at all.

Why Apple and Android Never Converge

Apple and Android have spent nearly two decades competing for the same device category while refusing to imitate each other's structure, and the identity-domain framework explains why that refusal is rational rather than stubborn. Apple's premium-UX identity makes vertical integration, owning the chip, the OS, the App Store, and increasingly the services layer, the only strategy that reinforces the brand promise of a seamless, controlled experience. That coherence pays: in its fiscal 2025 fourth quarter, Apple posted record quarterly revenue of $102.5 billion and an all-time-high Services revenue of $28.75 billion, up roughly 15 percent year over year, pushing full-year Services revenue past $109 billion. Services is the monetized residue of identity-driven lock-in; it only works because the hardware and software layers beneath it never stop reinforcing the same premium promise.

Google's data-ubiquity identity points the opposite direction. Android's value to its sponsor was never meant to come primarily from the OS license; it comes from putting Google's services and data pipes in front of as many end users as possible, which requires giving manufacturers, carriers, and regional OEMs enough latitude to customize hardware and go after price points Apple will never touch. The European Commission's own gatekeeper filings under the Digital Markets Act capture the structural difference without meaning to: Apple was designated for App Store, Safari, and iOS, three tightly coupled, Apple-operated surfaces, while Alphabet was designated across Google Search, Android, Chrome, Google Play, and YouTube, a looser federation of properties that each independently extend data reach. Regulators treat both as "gatekeepers." The identity-domain lens says they are gatekeepers of fundamentally different kinds of value, and that difference is not a technicality, it is the entire strategic logic each company is built to protect.

BlackBerry's Keyboard Was Never the Problem

The standard BlackBerry postmortem is about hardware nostalgia: the company clung to its physical QWERTY keyboard while the market moved to touchscreens, and it paid for the attachment. That story is true but incomplete. According to Harvard Business School's Digital Initiative, BlackBerry, founded in 1984, built its dominance on a specific identity: secure, efficient communication for professional and government users, the ecosystem that made "CrackBerry" a term for executives who couldn't put the thing down. By 2008, that identity had carried BlackBerry to a peak of 19.5 percent of the global smartphone market, the same year its then co-CEO dismissed the touchscreen trend outright: "the most exciting mobile trend is full QWERTY keyboards. I'm sorry, it really is. I'm not making this up."

The iPhone and Android overtook BlackBerry in global smartphone share by 2010, and BlackBerry's response was neither to double down on its security identity nor to credibly adopt Apple's design-led one, it tried to do a diluted version of both. The BlackBerry 10, launched in January 2013 after years of delay, drew praise from some critics but sold poorly; by then, five years of consumer loyalty had calcified around Apple and Android, and BlackBerry's late pivot toward consumer touchscreens and app-store parity didn't look like a company executing a new identity. It looked like a company that had lost its old one without finding a new one to replace it. BlackBerry's enterprise-security niche still exists today; it is simply no longer attached to a mass-market consumer ecosystem, which is exactly what identity-domain theory predicts happens when a sponsor tries to chase a rival's value logic instead of deepening its own.

Windows Phone Collected Everyone's Compromises

If BlackBerry's failure is a story about abandoning a clear identity, Windows Phone's failure is a story about never establishing one. Microsoft's mobile OS borrowed design language, the tile-based "Metro" interface was genuinely distinctive, but it offered neither Apple's end-to-end coherence nor Android's open-licensing scale and developer gravity. The result was a platform that never cleared a demand threshold strong enough to hold developers. At its regional best, in the five largest European markets, Windows Phone reached an 8.2 percent share in July 2013, per Kantar Worldpanel figures; globally, Strategy Analytics put its best quarterly share at only 3.9 percent that same year. Microsoft's effort to force the issue by acquiring Nokia's devices business in 2013 only concentrated the exposure rather than fixing the underlying problem. The last meaningful Windows 10 Mobile feature update shipped in October 2017, and Microsoft formally ended support for Windows 10 Mobile on December 10, 2019, a quiet death for a platform that, strategically, had never fully come alive. A platform without an identity domain has no principle for deciding which features earn a place and which don't; it ends up accumulating whichever compromises keep it in the conversation, satisfying no one's core need especially well.

A platform without an identity domain has no principle for deciding what to build next, only a list of other people's features to imitate.

Two Tests for Two Kinds of Gatekeeper

The identity-domain framework is not just a retrospective device for explaining dead platforms; it is a live design question for every platform sponsor deciding how far to stretch, and for every regulator deciding how to draw a line around market power. Constantinides, Cennamo, and Aaltonen explicitly extend their argument to the EU's Digital Markets Act, which designates "gatekeepers" largely through quantitative thresholds: an EU turnover of at least €7.5 billion in each of the last three years, or an average market capitalization of at least €75 billion, combined with at least 45 million monthly active EU end users and 10,000 yearly active EU business users, sustained for three consecutive years. The European Commission applied exactly this test in September 2023 when it designated six gatekeepers, Alphabet, Amazon, Apple, ByteDance, Meta, and Microsoft, across 22 core platform services. The thresholds are clean, auditable, and hard to lobby around at the margin. They are also indifferent to the fact that an Apple-shaped gatekeeper and a Google-shaped gatekeeper extract and defend value through almost opposite mechanisms.

None of these choices is free, which is itself the point the paper is making about regulation: a single enforcement template applied to Apple's walled garden and Google's open federation will always be either too blunt for one or too loose for the other, because the thresholds measure scale, not the strategic logic that scale is in service of. The same blindness is what let BlackBerry and Windows Phone drift for years before the market's verdict became undeniable, nothing in their boardrooms was tracking identity coherence as a leading indicator, only market share and feature checklists as lagging ones. Strategy that looks coherent on a slide can still be quietly dissonant with what an ecosystem's partners and users believe the platform is for, and by the time the dissonance shows up in the numbers, as it did for BlackBerry in 2010 and for Windows Phone through the mid-2010s, the identity that would have made recovery possible is usually already gone.