Roughly one-third of the population uses a health or fitness wearable. Recent surveys show that over 30% of Gen Z respondents use wearables to track their daily health. We live in a culture where health is judged by “sleep scores” and “stress scores.”
These metrics offer precious little insight beyond confirming a problem exists. Often, the data is intuitive: If you slept four hours after a long flight, you don’t need a watch to tell you your sleep was poor. If you’re working long hours at the office and foregoing exercise, it’s little surprise to see a high stress score.
While numerical judgment has been accepted as a hallmark of the quantified-self movement, we need better insights. Wearable use and reliance by consumers and physicians will increase, given the FDA’s new guidance documents that “will promote more innovation with AI and medical devices.” These guidelines aim to ease the regulation of wearables and increase the integration of wearable data into clinical workflows.
Yet as wearables become more sophisticated and AI integrates into medical devices, we’re at a crossroads. We can either double down on the scoring model—adding more complex algorithms to generate ever more “precise” numbers—or embrace what AI actually makes possible.
Contextual coaching understands individuals’ unique patterns, circumstances, and needs. The scoring paradigm hinders the delivery of personalized health guidance. Perfecting abstract scores won’t build systems to help people understand and improve their health in meaningful ways.
How we got here
Wearables have a long history, from the first glasses to early Polar heart rate monitors used by athletes. The health wearable market rose further with the iPhone and health apps. The Apple Watch made smartwatches mainstream, then introduced sophisticated sensors for heart and blood oxygen monitoring.
Even now, the wearable market is poised for expansion, in part due to greater awareness about heart disease, diabetes, and other chronic conditions. The global wearable technology market was worth $103.1 billion in 2025 and is projected to more than double to $229.97 billion by 2033, according to Grand View Research.
Early generation wearables excelled at collecting data, including sleep stages, heart rate variability, glucose trends, and nutrition. However, they failed to offer insight. For example, a standard interface couldn’t interpret the multicausal relationships among low heart rate variability, a late meal, and a high-stress workday for millions of users.
Industry leaders distilled massive datasets into a single, digestible number, a classic software solution for complexity. Scores, however, felt scientific and objective. Product teams designed challenges and leaderboards, creating a “minimum viable product” for data communication, even though they miscalculated data more than 20% of the time.
Why scores don’t work
Health scoring prioritizes measurement over meaning. A score tells you what, but rarely why. When a user sees a “bad” number, they are left with unanswerable questions: Is the score low because of stress, illness, or hormonal fluctuations? Was this score in my control (too many cocktails) or out of my control (the flu or a cold)?
Without context, the number becomes a source of anxiety. Users can feel judged by a “failed” score, which increases cortisol levels and further degrades recovery. The tool intended to motivate positive behavior ultimately undermines it through a reductive, grade-based model. Some reporting suggests that wearables have even led to mental health problems.
The inability to scale interpretation is a problem of the past. Rather than a vague, anxiety-provoking score, wearables are evolving to provide actionable, context-driven insights for care and lifestyle management. AI can now synthesize high-dimensional data, cross-reference it with contextual information (like travel or mood), and respond in natural language. Guidance is no longer delivered as a percentile, but as a conversation.
In the past, if a user woke up to a sleep score of 63, they might have felt tired and as if they had failed. Better data and technology allow a wearable to tell users: “Sleep was disrupted last night, likely because your bedtime shifted later than usual. Your recovery markers suggest you are catching up on a sleep deficit. Tonight, try an earlier wind-down ritual and avoid intense late-evening exercise.”
Or the wearable will acknowledge that, as a new parent, your sleep was likely interrupted by your baby, resulting in a temporary but expected shift in sleep patterns. This output results in better guidance based on context and understanding.
The shift from scores to adaptive systems
Health platforms can move beyond the scoring metaphor and focus on contextual coaching. This means providing daily motivators that adapt to the realities of a user’s life—a stressful workweek or a travel day—rather than assuming every day must be a perfect optimization exercise.
This shift is already happening. Products like Incora Health’s smart earrings and Nowatch’s screenless device represent a new generation of wearables that capture physiological signals without requiring users to constantly check or interpret a numerical score. These devices don’t display sleep scores or readiness ratings.
Instead, they collect data passively and surface guidance only when contextually relevant. By minimizing visual feedback and emphasizing subtle, contextual cues, they reduce the cognitive and emotional burden of being graded throughout the day. The hardware itself reflects the philosophical shift: As AI-powered interpretation improves, the interface can recede, allowing guidance to emerge naturally rather than demanding constant attention.
Metrics still matter. But the next decade’s most effective health experiences will prioritize both action (providing a single, realistic next step) and humanity (reducing the pressure of self-judgment). Wearables and health platforms will lean into each person’s health journey rather than assigning a score or number to everyone.
Theresa Neil is the founder of Femovate.