What is actually inside a smartwatch, and why the battery is the whole design

The theory of smartwatch design is well covered elsewhere. This is about the version you meet in practice.

The constraint is what people will wear

A watch has an acceptable thickness and diameter set by human wrists and fashion, not by engineering. That fixes the battery volume, and everything else is allocated from what remains. This is why smartwatch capability has advanced far more slowly than phone capability despite using related silicon.

Every added sensor and radio is charged against a budget that cannot grow.

The screen is the biggest consumer

Display power dominates whenever the screen is lit, which is why always-on modes have such a visible effect on runtime. OLED helps because black pixels are genuinely off, which is why watch faces are overwhelmingly dark by design rather than by fashion. Refresh rate is dropped aggressively in always-on states for the same reason.

Consumption also scales with brightness and with how much of the panel is lit, so a pale face with large filled areas viewed outdoors can cost several times what a sparse dark one costs indoors.

Optical heart rate has real limits

Photoplethysmography shines light into the skin and measures changes in reflected intensity as blood volume changes. It works well at rest and degrades with motion, poor contact, cold skin and darker skin tones under some sensor designs.

Chest straps measure electrical activity directly and remain more accurate during high-intensity intervals, which is why athletes still use them. Fit dominates the rest: a loose band lets ambient light reach the sensor and lets the sensor slide against the skin, which is why the standard instruction is a snug strap sitting above the wrist bone rather than on it.

Radios are expensive in power terms

Bluetooth Low Energy is cheap; Wi-Fi is expensive; cellular is very expensive, which is why cellular watches have noticeably shorter runtime. GPS is the other large consumer, and dual-frequency reception improves accuracy in cities at a meaningful power cost.

Most of the runtime differences between similar-looking watches come down to which radios are active and how often. Cost also rises with difficulty, because a radio at the edge of its range raises transmit power and resends lost packets, so the same connection is more expensive with the phone in a bag across the room than in a pocket.

Regulatory status matters more than marketing

Some features — electrocardiogram, irregular rhythm notification, blood oxygen — have regulatory clearance in some jurisdictions and not others. The same hardware can therefore expose different features depending on where the watch was sold.

The short version: a cleared feature is not a diagnosis, and every manufacturer says so in documentation that almost nobody reads. Clearance attaches to a specific claim on specific software, so an update can add or withdraw a feature on a watch already on your wrist without anything physical changing.

Implementations differ, and vendors are not obliged to document the differences.

Water resistance and lifespan are the same decision

Wearables are sealed with adhesive rather than screws, because a gasket and fasteners cost thickness that the battery has already claimed. Opening one to replace the cell destroys that seal, so a battery change is a service operation with new adhesive and a pressure re-test rather than something an owner does. The practical lifetime of a watch is therefore often the lifetime of its first cell, and cells lose capacity by calendar age as well as by charge cycles.

Mechanically, proprietary charging pucks are the second limit, because a device in perfect condition becomes unusable when the only charger that fits it is no longer made.

The takeaway

Everything a watch can do is negotiated against a cell you could hide under a thumbnail.

The constraint is almost always physical, and marketing rarely mentions which one.