Field Note · Independent analysisProduct teardown
Apple Ecosystem Fragmentation
The company most associated with coherence, examined at its seams.
Apple sells one idea above all others: it just works, and it works the same everywhere. That promise is why people buy the second and third device. It is also, in practice, where the ecosystem is weakest. The same task (sharing a file, managing a notification, finding a setting, moving a piece of work from one screen to another) feels different on iPhone, iPad and Mac, and the difference is not always because the device demands it.
Where the seams are
Inconsistent mental models. A gesture, a menu location or a name that differs across platforms forces the user to relearn something they already know. Context-switching cost. Moving mid-task from Mac to iPhone means reconstructing where you were, not only reopening the app. Feature parity gaps. Capabilities that exist on one device and not another break the expectation the ecosystem has spent years building. Design language drift. Some divergence is right, since a pointer is not a finger, but a good deal of it is historical accident, the residue of separate teams shipping on separate calendars.
The people who notice most are power users, who develop private workarounds, and new users, who expected the promise and find the seams first. Both groups quietly discount the value of owning the whole set.
What I would build
A unified interaction language: one set of gestures, animations and feedback patterns, defined once and adapted per input method, so that learned behaviour transfers. A context continuity layer that carries not just the open document but the working state (position, selection, intent) across the hand-off, so that the second device picks up rather than restarts. Adaptive parity: functional equivalence as the rule, with device-specific exceptions made deliberately and explained, not discovered. And progressive disclosure that is consistent across devices, so complexity is revealed on the same schedule whether the screen is six inches or twenty-seven.
Design principles
Familiarity over novelty: reuse what the user already learned on their other device. Preserve context and intent, not just files. Adapt to the device's strengths without abandoning the shared pattern. And make behaviour predictable: a user who has used a feature on one device should be able to guess how it works on the next.
What I would measure
Task completion when a task starts on one device and finishes on another, against the same task on a single device. Time to first productive action after a hand-off. Support and search queries of the form 'where is X on iPad': the ecosystem's own confusion, written down by its users. And multi-device engagement over time, because that is the number the promise is supposed to move.
The learning
Coherence is a product, not a by-product. It has to be owned by someone whose remit crosses the device teams, or it erodes one reasonable local decision at a time. The lesson generalises to any company with more than one surface.