Users can rarely judge fit themselves; the system needs to show signal quality

What it is

Optical paths under skin and electrode impedance are invisible. “Feels snug” can still leak light; “feels loose” can sit in the optical sweet spot. Fit cannot be accepted by proprioception. The system has to translate signal quality into an actionable prompt: reseat, tighten one hole, pause measurement.

Why it happens

People feel pressure and slip; they do not feel whether AC perfusion is enough or whether common-mode rejection has collapsed. Quality metrics have to come from front-end physics: PPG AC/DC, agreement with a physiological heart-rate range, electrode impedance, mains energy fraction, coherence between acceleration and the physiological spectrum. The metric must also separate “not worn” from “worn but poor,” or users retry while already seated. Timing decides usefulness: a one-shot flash on a settings page while a pretty heart rate still shows after daily slip keeps invisibility until after the decision.

Studying it

Ask users to adjust to “feels right” by feel alone while logging objective SNR, then give a quality prompt and let them adjust again; compare error. Factors: prompt form (color dot, text, whether it names an action), whether the prompt stays up. Outcomes: how many reach usable SNR, overtightening rate, ignore rate. Contrast with “no prompt but the UI still shows a number” to see how many treat bad data as true. Obvious laboratory lift-off is easy; the field is slight light leak, and the prompt must be more sensitive.

Where it stops holding

A quality warning that blinks continuously becomes a new false alarm, and users disable the prompt rather than reseat. Clinically trained users may read waveforms; consumer users cannot be assumed to share that skill. The indicator itself discloses “the body is being measured” and may need a mute in social settings. If the metric is calibrated on a laboratory population, dark skin or hair can be systematically labeled “poor quality,” and the prompt becomes exclusion.

Applying it

• Show quality for the whole period a measurement may drive UI decisions, not only at pairing.
• Put an action in the prompt: “tighten one hole,” “move up off the wrist bone,” not only a red mark.
• When quality is below threshold, withdraw the number or mark it unusable so a bad heart rate cannot keep driving adaptation.
• Verify by creating a slight light leak or high impedance without a visible hardware gap, and counting how many people correct from the prompt alone; if they cannot, or no prompt appears, the indicator failed.

Related

• Same group: C9.09.1 How well an electrode or sensor seats on skin directly sets signal-to-noise ratio · C9.09.2 Sweat, hair, and changing skin alter contact impedance during wear · C9.09.3 Loose or misplaced wear systematically shifts the measurement baseline
• Adjacent: C9.10 Individual Calibration and Baseline Drift · C4.29 Visibility of Capture State
• Search: signal quality indicator · wear detection · PPG SQI