Apple's Strongest Lock-In Is Free. It's the Color of a Text Bubble.

A teenager opens a group chat and sees one bubble in the wrong color. Not blue — green. The others know what it means before they read a word: someone has an Android phone. The reactions, the typing indicators, the high-res photos all degrade for the whole group because of that one outsider. Nobody at Apple has to do anything to make this happen. The social pressure does the work, the kid asks their parents for an iPhone, and a sale closes itself. That green bubble is the most efficient piece of strategy Apple ever shipped, and it never appears on an invoice.

The official story about Apple's lock-in is the App Store — the so-called 30% tax, the antitrust dragnet, the monopoly that regulators were finally going to break. That framing is mostly wrong, and the part that's right points somewhere else entirely. Apple's deepest moat isn't the fee it collects. It's the feeling it engineers — and the clearest proof is a decision its own executives made in writing, on the record, about a green bubble.

The text-bubble decision was made on purpose, in writing

In April 2013, Apple's own SVP of internet software, Eddy Cue, wanted to bring iMessage to Android as a full project. He pushed for it. He lost. The internal emails surfaced in the Epic v. Apple discovery, and they are remarkable not for what Apple decided but for how nakedly it reasoned. SVP Craig Federighi objected that an Android iMessage 'would simply serve to remove an obstacle to iPhone families giving their kids Android phones,' and SVP Phil Schiller agreed that 'moving iMessage to Android will hurt us more than help us.'⁴ Read that again. The argument against shipping a better product to more people was not that it would fail. It was that it would work — and in working, it would free families from the gravity that keeps the next phone an iPhone.

Note what this corrects in the popular telling. The story that 'Apple chose lock-in' is true, but it wasn't monolithic — a senior executive argued the other way and was overruled. That nuance matters, because it shows lock-in wasn't a passive byproduct of nice engineering. It was a contested strategic choice that the lock-in faction won, on explicitly stated grounds. The thesis of the whole walled garden lives in those emails: keep the family inside, and the family will keep buying.

The mechanism: a moat that costs nothing to defend

Most moats are expensive. You defend a brand with advertising, a patent with lawyers, a network with subsidies. iMessage lock-in is different: the enforcement is outsourced to your friends. Apple builds the messaging app, makes it default and exclusive to its hardware, and then the social cost of being the green bubble in the group chat is paid by users on users — at zero marginal cost to Apple. The geography of a teenager's group chat does the suppressing; Apple just declines to remove it. That is why the company would rather keep a superior product off billions of phones than collect a single dollar of goodwill from them. The lock-in is the product.

The App Store 'tax' is mostly a myth — and the courts said so

Here's where the popular outrage misfires. The headline number — 'Apple takes 30% of everything' — is not how the App Store actually works for most of the people in it. Under Apple's Small Business Program, developers earning under $1 million a year pay 15%, not 30%, and subscription commissions also fall to 15% after a subscriber's first year regardless of developer size.⁵ So the 30% rate is a real charge on large developers, not a universal toll. The majority of the catalog pays half that. The tax everyone is angry about is, for most of the App Store, a different and smaller number.

The legal story is even further from the folklore. In Epic Games v. Apple, the court found Apple did not violate federal antitrust law on nine of ten counts. The single violation was under California's Unfair Competition Law — a state consumer-protection statute — over Apple's anti-steering rules, producing an injunction that forced Apple to let developers link out to external payment.² That is not a monopolization ruling. It is a narrow finding about one contract clause. Anyone who tells you a court declared the App Store an illegal monopoly is repeating something that did not happen.

Then Apple proved the cynics half-right anyway

Having lost only the anti-steering point, Apple responded by complying in form while gutting it in substance: it let developers link out, then imposed a 27% commission on those off-app purchases — re-creating the toll through a different door. The Ninth Circuit, in December 2025, found Apple in civil contempt by clear and convincing evidence, ruling the 27% fee had a 'prohibitive effect' in violation of the injunction and that Apple had acted in bad faith.³ In May 2026 the Supreme Court declined to pause the contempt order, leaving the injunction in force.⁷ This is the tell. A company that fights this hard to preserve a fee it claims is incidental is telling you how much the fee is worth — even as the deeper moat sits untouched, because nobody can enjoin a green bubble.

And the money explains the ferocity. Services revenue — the App Store, iCloud, Apple Music — ran $96.2 billion in fiscal 2024 at a 73.9% gross margin, against 37.2% on Products.¹ The hardware is the toll booth; the services are the road. The walled garden's whole financial logic is that every locked-in user is a recurring, high-margin annuity, and the devices exist partly to keep that annuity from walking. Epic put the strategy in Apple's own alleged words: hook users on the integrated ecosystem so they wouldn't want to leave it.⁶

Isn't this just a product so good people don't leave?

The fair objection — and Apple's own — is that none of this is sinister. People stay because the products are genuinely excellent, interoperate beautifully, and respect privacy. High retention isn't a cage; it's a compliment. Industry surveys put iPhone retention near 92% and report that most iOS owners go on to buy more Apple devices.⁸ If your stuff is the best, of course people don't leave. That's a fair point, and it's partly true. But it doesn't survive the 2013 emails. A company confident that its product alone keeps customers does not block a better version of that product from reaching more of them. The Federighi argument only makes sense if you fear that interoperability would let people leave — which is an admission that something other than sheer quality is doing the holding.

The honest version is that both things are true at once: Apple makes superb products and deliberately withholds interoperability to convert that excellence into captivity. Even the famous loyalty numbers carry an asterisk — the much-cited '88% of U.S. teens own an iPhone' figure traces to a market-research survey, not a Census or an Apple disclosure, so treat it as a directional signal, not gospel.⁸ The point stands regardless of the decimal: the loyalty is real, and so is the engineering that makes leaving feel like a social demotion.

So aim the scrutiny correctly. The App Store fee is the part everyone argues about and the part Apple can lose — capped by regulators, litigated by Epic, narrowed by courts. The green bubble is the part nobody can touch, because it isn't a price or a contract; it's a feeling, planted in a billion group chats, enforced for free by the people you text. Apple didn't out-engineer its rivals into the walled garden. It out-located its product — choosing to be present in your relationships rather than merely on your phone — and then declined, in writing, to let anyone else in.