Consumer wearables love a sleep score. Eighty-two last night, seventy-four the night before, a little ring that turns green. Product teams copy the pattern because it photographs well and because every vendor demo leads with it. Then a physician advisor asks what 82 means for a person with sleep apnoea, shift work, or a new SSRI, and the room discovers that the score is a proprietary mash of duration, disturbance, and whatever the firmware felt like weighting this quarter.
Health App Analytics does not forbid composites. It forbids composites without an appendix. If you cannot show time in bed, time asleep, number of long wakes, and whether the watch was actually worn, you do not have a clinical artefact. You have a game mechanic. Pulse Map’s dashboard-critique week fails boards that lead with a score and hide the traces behind a tap that nobody in the review ever makes.
Coverage, again, is the unfashionable hero. Nights with a usable trace. Nights the device sat on the nightstand. Nights the user travelled across time zones and the stage labels became fiction. An employer-wellness desk we sat treated “average sleep score” as engagement; pairing had quietly died. Once they led with usable nights, the score became optional colour, which is all it deserved.
Design implication: put the trace first for anyone who will make a care decision. Keep the score for the person using the app at the sink, if your product truly needs a glance. Do not feed the glance into the executive board as if it were an outcome. Sensor Storytelling briefings spend a surprising amount of time on this one swap. It feels small. It changes who argues with you.
If you are mid-redesign, bring three nights of raw stages — gaps included — to the desk, or sit the wearable ingest block inside Pulse Map. We would rather look at an ugly night than a polished 82.