Apple's Moat Isn't Design. It's the Cost of Leaving — and It Compounds.

Try to leave. Not in theory — actually do it. You'd re-buy the apps, re-thread the group chats, untangle the photos, re-pair the watch that won't pair with anything else, find a new home for the saved passwords, and re-learn where everything lives. None of that is a contract. None of it is a locked chip. It's just the slow accumulation of a life arranged around one set of devices — and that accumulation is Apple's actual moat. Not the aluminium. The cost of walking away from it.

The official story is that Apple wins on design and innovation — that people pay the premium for the beautiful object and stay for the next beautiful object. That's the comfortable read, and it's wrong about the mechanism. Apple's own filings claim no structural moat at all; they describe its markets as 'highly competitive,' with 'aggressive price competition' and 'downward pressure on gross margins.'⁸ A company sitting on a design moat doesn't write that about itself. The moat is somewhere the 10-K won't name it.

Here is the thesis, plainly: Apple isn't a hardware company defended by design. It's an install base defended by switching costs that compound — each Apple device you own makes the next Apple device cheaper to choose and the whole stack more expensive to abandon.

Each device makes the next one easier and the exit harder

Most lock-in is binary: you're in a contract or you're not. Apple's is a gradient that steepens with every purchase. Buy an iPhone, and an Apple Watch is the obvious next watch because it's the only one that talks fully to the phone. Add AirPods, and they hand off between phone, watch, and Mac with no setup. Add a Mac, and your messages, copy-paste, and calls flow across all of it. Each addition does two things at once: it lowers the friction of the next Apple purchase (everything just works together), and it raises the friction of leaving (now you'd have to replace four things, not one). That's the engine. The install base isn't a count of customers — it's a count of these compounding entanglements, and it crossed 2.5 billion active devices in the quarter ended December 27, 2025.¹

You can see the engine's output in behaviour. iPhone customers held an 89% loyalty rate over the twelve months ending June 2025 — meaning roughly nine in ten upgraders bought another iPhone — against about 77% for Samsung.⁵ That twelve-point gap is the gradient made visible. It isn't that the iPhone is twelve points better. It's that the cost of leaving is twelve points heavier, and it gets heavier each year the entanglement grows.

Where the lock-in turns into cash

A moat that only protects market share is half a moat. Apple's converts directly into the highest-quality revenue it earns. The install base is the meter that the Services business reads. In fiscal 2025, Services revenue reached $109.16 billion — the first year it cleared $100 billion, up roughly 13.5% on the prior year.³ And this is the part that matters: the year before, Services ran at a 73.9% gross margin against just 37.2% for Products,² and Apple's FY2025 filing reports the Services margin rose further.³ So the locked-in customer doesn't just buy the next phone — they rent the App Store, the cloud, the music, the games, at roughly double the margin of the hardware that delivers them. The device sells once at 37%. The customer pays rent at 74% for as long as they stay. And the moat is what makes them stay.

Didn't regulators just confirm the lock-in?

Here the popular narrative trips over itself, and it's worth being precise because the truth is more interesting than the myth. The story you'll hear is that regulators 'proved' iMessage is a market-dominant lock-in tool. The record says the reverse: in February 2024 the European Commission concluded that iMessage does not meet the Digital Markets Act threshold to be designated a core platform service, so Apple faces no EU obligation to open it to rivals.⁶ The green-bubble-versus-blue-bubble pressure is real social glue — but legally it isn't the load-bearing wall everyone assumes.

What regulators did conclude is more telling for our thesis. They designated iPadOS a gatekeeper despite it not meeting the usual quantitative thresholds, and the stated reason was that 'end users are locked-in to iPadOS' and that 'Apple leverages its large ecosystem to disincentivise end users from switching.'⁷ That is the cumulative switching-cost moat described, in a regulator's own words, as a fact on the record. The lock-in they confirmed isn't a single feature like iMessage. It's the ecosystem itself — exactly the gradient, not the gadget.

The honest objection: isn't this just a great product people like?

The fair counter is that switching-cost framing can be a cynical way of describing a company that simply makes things people prefer — and Apple's own filings give that objection ammunition. If the market is genuinely 'highly competitive' with 'aggressive price competition' and 'downward pressure on gross margins,'⁸ then maybe the 89% loyalty⁵ is just satisfaction, not lock-in, and the moat is a flattering label for a hit product. That's a serious point, and it's partly right: a lock-in with a product people hated would erode, because the cheapest way to lower your switching cost is to be willing to suffer through the switch.

But notice what the satisfaction explanation can't account for: the compounding. A merely-good product wins each purchase on its own merits, fresh, against rivals. Apple wins the next purchase partly because of the last one — and the more you own, the wider the gap grows, independent of whether this year's model is the best on the market. The 37%-margin device increasingly exists to feed the 74%-margin rent.² You don't build a $109 billion recurring-revenue business³ on satisfaction alone; you build it on satisfaction that has been converted into entanglement. The product earns the entry. The switching cost keeps the rent flowing after the product is no longer the only reason to stay.

Apple's genius was never the chamfered edge or the launch-day demo, though those buy the entry ticket. It was recognising that a customer who owns four of your things is a fundamentally different asset than one who owns one — not four times as valuable, but exponentially harder to lose. The company sells aluminium at 37 points to acquire a relationship it rents back at 74. The moat is invisible because it isn't a feature you can point to; it's the quiet weight of everything you'd have to redo. Apple doesn't have to keep you from leaving. It just has to keep making leaving slightly more expensive than staying — and then let the years do the compounding.