You wake up, check your wearable, and see a recovery score that looks worse than expected. The number is clear. The reason is not. Was it the late meal, a tense conversation, a warm bedroom, the beginning of an illness, or simply the normal error around a consumer estimate? The practical answer is to treat wearable data as a signal, then add a short record of the context around it. A device can estimate selected physiological or behavioral measures. It cannot reconstruct your lived situation or prove what caused the result. A voice note can preserve that missing layer while it is still easy to remember.
A score is a measurement layer, not an explanation
Wearables are good at making invisible changes visible. They can give you a consistent record of measures such as heart rate, step count, estimated sleep duration, or device-specific recovery scores. That continuity can be useful, especially when you review a trend instead of reacting to one morning.
But the apparent precision of a number can hide uncertainty. A 2024 umbrella review of wearable accuracy found that performance varied substantially by measure and device: heart-rate estimates were generally stronger than estimates of activity intensity or sleep. An earlier systematic review covering 158 publications similarly found that validity depended on the brand, metric, and setting, with energy expenditure performing particularly poorly. Both reviews combine studies of different devices, algorithms, and settings, so they cannot tell you exactly how the current firmware on your particular device will perform.
Sleep deserves extra caution. In a systematic review comparing Fitbit, Garmin, and WHOOP devices with polysomnography, agreement varied across devices and sleep measures. Separately, the American Academy of Sleep Medicine says that consumer sleep technology should not be used to diagnose or treat sleep disorders, although the data may help support a conversation with a clinician when interpreted in context (AASM position statement).
None of this makes the wearable useless. It changes the question.
Instead of asking, “Why did my body do this?”, ask, “What did the device observe or estimate, and what context should I preserve before I forget it?”
The context disappears faster than the number
Your dashboard will still show Tuesday’s score next month. It probably will not show that Tuesday included an early flight, a difficult client call, a restaurant meal, two glasses of wine, an unusually hard workout, or a child waking during the night.
That asymmetry matters. Numbers remain neat while memories become vague. When you review the week later, it becomes tempting to attach a simple story to the cleanest available variable:
- “My supplement stopped working.”
- “The workout ruined my sleep.”
- “My recovery is getting worse.”
Those interpretations may be possible, but the score alone cannot establish them. The U.S. Food and Drug Administration distinguishes certain low-risk general-wellness functions from software functions intended for diagnosis or treatment. In other words, a product's intended use and regulatory status matter; a health-related display should not automatically be read as a medical conclusion (FDA guidance).
Context does not eliminate measurement error or confounding. It gives you a better set of questions.
Use the Four-Layer Context Note
A useful wearable note does not need to become a diary entry. It needs four layers: signal, situation, subjective state, and plausible confounder.
1. Signal: record what the device actually showed
State the metric without interpreting it.
Sleep duration estimate: 6 h 12 min. Resting heart rate above my recent range. Recovery score lower than usual.
Avoid translating a device score into a medical conclusion. “Low recovery score” is a faithful observation. “My nervous system is damaged” is not.
If the app offers a chart, save a screenshot. It preserves the exact view without requiring you to retype every metric. When you later compare several entries, use the same measure and roughly the same time window.
2. Situation: capture what was different
Name the circumstances that a sensor could not know.
Late dinner, bedroom felt warm, and I worked until 10:30 p.m. after an argument.
The best situation note is specific enough to be useful but short enough to repeat. Calendar load, travel, caregiving, illness exposure, meal timing, training load, and unusual evening stimulation are all candidates. You do not need to log all of them every day.
3. Subjective state: describe how you actually felt
A wearable estimate and your experience can disagree. Keep both.
Woke up mentally clear but physically heavy. Energy 3/5. No headache. Motivation normal.
This is not “soft data” to be discarded. It is a different type of observation. Your goal is not to decide which source wins every morning. It is to notice when they move together, when they diverge, and whether either one predicts a decision you care about.
If you need a compact baseline, DailyLens’s guide to personal analytics for energy recommends starting with outcomes such as energy, focus, and stress before adding more inputs.
4. Plausible confounder: name uncertainty explicitly
A confounder is another factor that could help explain the same observation. You do not need statistical training to use the idea well.
Possible confounders: room temperature, early illness, or device fit. One unusual night, so no conclusion yet.
That final sentence is valuable. It interrupts the urge to turn one association into a cause. It also creates a reminder to look for repetition before changing a routine.
Put together, the complete note can take less than a minute:
Signal: Lower recovery score; estimated sleep 6 h 12 min.
Situation: Late dinner, warm room, difficult conversation, worked late.
Subjective state: Mentally clear, physically heavy, energy 3/5.
Plausible confounder: Possible illness or loose device fit; no conclusion from one night.
Voice is especially useful here because context is often easiest to recall while you are still in the moment. A 30-second note can preserve tone and nuance that would feel tedious to type. If speaking is inconvenient, the same structure works in text. The method matters more than the input mode.
Review patterns without manufacturing causes
Daily data becomes more useful when you stop judging each score and review several days together. Once a week, ask three questions:
- What repeated? Look for a pattern across more than one isolated day.
- What else changed? Check workload, sleep opportunity, meals, alcohol, exercise, travel, illness, and device conditions.
- What decision would this change? If the pattern cannot inform an action, you may not need to track it.
Suppose lower sleep estimates appeared on three nights with late work, but two also involved a warm room and one involved alcohol. The useful conclusion is not “late work causes bad sleep.” A more honest hypothesis is: “Late work may be part of a cluster that reduces my sleep opportunity or changes my wind-down. I will keep the room cooler and stop work earlier for several comparable nights.”
Change one practical variable when possible. Keep the test reversible. Then review both the device signal and your subjective state. This is closer to disciplined self-observation than chasing the dashboard after every fluctuation.
If the issue is sleep, the broader basics still matter more than a single score. The DailyLens overview of sleep, energy, and focus can help you place the metric inside a larger routine instead of treating it as a verdict.
What wearable context cannot tell you
Context-rich journaling improves interpretation, but it has limits.
- It does not turn a consumer wearable into a diagnostic device.
- It does not prove that a meal, supplement, workout, or stressful event caused a physiological change.
- It cannot rule out illness, medication effects, sleep disorders, or measurement error.
- It should not be used to change prescribed treatment without professional guidance.
Persistent, new, or concerning symptoms deserve appropriate medical evaluation even if your dashboard appears normal. The reverse also applies: an unusual score without symptoms is a reason to check the device, the trend, and your context before assuming something is wrong.
How this fits into DailyLens
DailyLens is designed to keep the number and the lived day in the same record. You can attach a wearable screenshot to an entry, add a voice or text note, and track a small set of states such as energy, focus, stress, or sleep quality. The current workflow is manual context capture; it should not be confused with automatic syncing from Apple Health, Oura, WHOOP, Garmin, or Google Fit.
That deliberate boundary is useful. It encourages you to capture the part that no sensor can reconstruct: what was happening, how you felt, and what else might have mattered. If you are new to the habit, start with the simple principles in What Is Daily Journaling?, then add wearable context only when it helps answer a real question.
The goal is not a more impressive dashboard. It is a record that helps you make one calmer, better-informed decision.
If you want to pair wearable screenshots with voice-first context and lightweight state tracking, join the DailyLens early-access list.
Sources and further reading
- Fuller D, et al. Reliability and validity of commercially available wearable devices for measuring steps, energy expenditure, and heart rate: systematic review. JMIR mHealth and uHealth. 2020.
- Doherty C, et al. Keeping Pace with Wearables: A Living Umbrella Review of Systematic Reviews Evaluating the Accuracy of Consumer Wearable Technologies in Health Measurement. 2024.
- Schyvens AM, et al. Accuracy of Fitbit Charge 4, Garmin Vivosmart 4, and WHOOP Versus Polysomnography: Systematic Review. 2024.
- American Academy of Sleep Medicine. Consumer sleep technology: position statement.
- U.S. Food and Drug Administration. General wellness software functions and medical-device boundaries.
