Why is my sleep score low? A low sleep score usually means one or two parts of your night fell short of your device's target. You slept less than usual, went to bed at an unusual time, woke up more, or spent less time in the sleep stages it tracks.
Key takeaways
A sleep score is a number from 0 to 100 that your wearable calculates each morning by comparing a few estimates about your night. Those estimates include how long you slept, when you went to bed, how often you woke up and, on most devices, how much time you spent in each sleep stage. Your device builds those estimates from your movement and heart rate, then rolls them into one number.
The recipe is different on every brand, so a sleep score isn't a universal measurement. A 74 on an Apple Watch and a 74 on an Oura Ring are adding up different things. The same night can land in different places on an Apple Watch, a Fitbit, a Garmin, an Oura Ring or a Samsung watch.
Apple doesn't count sleep stages at all. Oura only looks at the time you fell asleep. Garmin factors in your stress overnight. There is no standard or consensus among devices on what an actual good night's sleep looks like.
That's why the same night can look fine on your wrist and rough on your finger. Take eight full hours that started late and held less deep sleep than usual. Apple counts the hours and takes points off for the late bedtime. Oura, Garmin and Samsung also mark the night down for the missing deep sleep.
Duration carries the most weight on most devices. On Apple Watch, it's worth up to half the score. Adults need seven or more hours a night on a regular basis, according to a joint consensus statement from the American Academy of Sleep Medicine and the Sleep Research Society.
Apple gives 30 of its 100 points to bedtime consistency, and Oura scores your timing. Late nights and weekend sleep-ins cost points even when the total hours look fine. Beyond the score, a 2024 study of more than 60,000 UK adults found that regular sleep timing predicted long-term health outcomes more strongly than sleep length did.
In a study of more than 4,000 Finnish employees wearing heart rate monitors, evening drinking lowered measured recovery during the first three hours of sleep, and more drinks meant a bigger drop: roughly 9 percentage points after a small amount, 24 after a moderate amount and 39 after a large amount. Scores that factor in heart rate or HRV are built to pick this up.
In a small sleep-lab study, 400 mg of caffeine taken six hours before bed cut measured sleep by more than an hour.
Garmin puts your average overnight stress level straight into the score. Other devices can pick up a hard day through a higher heart rate while you sleep.
Wearables are good at telling sleep from wake and less accurate at telling sleep stages apart. If your score drops because of "low deep sleep," the estimate itself may be part of the story. More on this below.
Did you know? In that caffeine study, people lost more than an hour of sleep and didn't notice. Their sleep diaries missed what the lab measured. Sometimes your score catches something your memory of the night doesn't.
Drake C, et al. Caffeine effects on sleep taken 0, 3, or 6 hours before going to bed. J Clin Sleep Med. 2013;9(11):1195–1200. doi.org/10.5664/jcsm.3170
Apple's sleep score comes from three things: sleep duration (up to 50 points), bedtime consistency (30 points) and interruptions (20 points). Sleep stages don't count. If you slept a full night and still scored low, look first at bedtime consistency. A bedtime that shifted from your usual window is the most likely reason. Apple rates 81–95 as "High" and 61–80 as "OK."
Oura scores seven contributors: total sleep, efficiency, restfulness, REM sleep, deep sleep, latency (how long you took to fall asleep) and timing. Open the score and look for the contributor that dropped. Timing is the one people most often miss, because Oura scores when you slept against your natural rhythm. Oura calls 70–84 "good" and 85 or above "optimal."
Fitbit compares five measurements with targets tailored to your age and gender: duration, time to sound sleep, time in sound sleep, restlessness and full awakenings. Fitbit says most people average between 72 and 83. If your score dropped recently without any change in how you sleep, Fitbit updated its algorithm and notes that some people will see lower scores for the same sleep.
Garmin's score combines duration, time in each sleep stage, awake time and restlessness, and your average stress level overnight, all read from heart rate, HRV and movement. A hard day can lower it even after a long night. Garmin reported that its users averaged 71 in 2024, and in late 2022 only about 5% reached the 90–100 "excellent" range.
Samsung's score looks at total sleep time, sleep cycles, movement and awakenings, physical recovery (time in deep sleep) and mental recovery (time in REM). Because deep sleep and REM each carry weight, a night with enough hours but fewer estimated deep or REM minutes can score lower than you'd expect.
It depends on the device. On Fitbit and Garmin, 80–89 is "good" and 90 or above is "excellent." On Oura, 70–84 is "good" and 85 or above is "optimal." On Apple Watch, 81–95 is "High." Scores in the 90s are rare by design, and most people land in the 70s.
The line that matters more is your own. A 70 that's normal for you tells you less than a sudden drop from your usual 82. Read a week of scores before you read any single morning.
Sleep stages are the least reliable part. When researchers compared seven consumer sleep trackers with polysomnography (the lab standard for measuring sleep), most were good at telling sleep from wake. Stages were less accurate. All six devices that reported stages overestimated light sleep, and several misjudged deep sleep or REM (Chinoy et al., Sleep, 2021). A newer 2024 lab comparison of the Apple Watch Series 8, Fitbit Sense 2 and Oura Ring Gen3 found that all three were close to the lab on total sleep. The Apple Watch underestimated deep sleep by about 43 minutes, and the Fitbit by about 15. That study was funded by Oura, and Oura came out closest to the lab, so under the SoloMio Evidence Standard we're treating the ranking as provisional until independent researchers repeat it.
The formulas are private and they change. Brands publish the ingredients, not the full recipe, and updates can move your score without any change in how you slept.
Chasing the number can backfire. Sleep researchers have described people whose sleep got worse while they tried to perfect their tracker data. They named it orthosomnia (Baron et al., J Clin Sleep Med, 2017). The score is there to help you understand your night, and it works best when you hold it loosely.
Duration is only part of the score. Apple saves half its points for bedtime consistency and interruptions. Oura, Garmin, Fitbit and Samsung also score restlessness, how long you took to fall asleep, or how much deep and REM sleep they estimated. A long night that started late, or was broken up by wake-ups, can still score low.
A 70 is usually close to average. Garmin reported that its users averaged 71 in 2024, Fitbit says most people average between 72 and 83, and Oura counts 70–84 as "good." What matters more is how 70 compares with your own recent scores. A steady 70 is less telling than a sudden drop from 85.
They score different things. Apple's score comes from duration, bedtime consistency and interruptions, and leaves sleep stages out. Oura adds deep sleep, REM, efficiency, time to fall asleep and timing. A night with plenty of hours but little estimated deep sleep can look fine on one and middling on the other.
Consumer trackers are generally accurate at telling whether you're asleep or awake and at estimating total sleep time. They're less accurate at identifying sleep stages, often overestimating light sleep and misjudging deep sleep. Treat the stage-based parts of your score as estimates, and trust trends over several nights more than any single reading.
One low score is common and usually reflects an ordinary change, like a late night or an evening drink. If you regularly wake up unrefreshed, snore loudly or feel very sleepy during the day, whatever your score says, talk to a healthcare professional.
SoloMio offers wellness guidance, education and lifestyle support. It doesn't diagnose, treat or replace a doctor. If you have a health condition, are pregnant or take medication, talk to a professional before changing your routine.
We're opening SoloMio to a small first group, who'll help shape what it becomes.
SoloMio turns the signals your body already sends into a language you can understand. Here's who we built it for and the six checks every recommendation has to pass.
.png)