Health scores your users trust, without hiring a science team
Sleep, recovery and activity scores designed by our health scientists and computed inside your own Open Wearables deployment. A licensed module, validated by our science team before each release, with what every score measures written down so your team can explain any number to whoever asks.
Trusted by teams building on Open Wearables
What you get
Health scores built and validated by our science team
Build a morning check-in, a recovery view or a coaching plan on Open Wearables Health Scores. Each score is ready to ship and validated before release.
Designed by our health scientists and ML engineers, and validated by them before each release. You know what each score measures, so when someone asks why the number moved, you can answer.
Scores are calculated over the history already in your database, so the feature launches with weeks of trend for every user instead of an empty chart.
Webhooks push each new score as soon as it is ready. The API returns today's score, a given day or the full history, so there is no new pipeline to build.
Sleep, recovery and activity, each from 0 to 100. Your users get a reason to open the app every morning, and our science team keeps the scores validated as devices change.
The scores
Ready to ship, on one method
Sleep, recovery and activity each run from 0 to 100 on a shared method, so a user's numbers read the same way across devices and over time. Our science team adds new scores on that method when your product needs one.
Sleep score
How well a user slept, in one number, from what the night's sleep data says about its length, depth, regularity and disruption.
Recovery score
A morning read on whether the body is ready for load, for training and coaching products.
Activity score
One daily number for movement and its balance, rather than a step count on its own.
What you build with them
What your product does with a score
A score is the start of a feature, not the end of one. These are the moments teams build on top.
A number users check every morning
One score on the home screen, computed in your deployment overnight.
Coaching in your own voice
The same score reads differently for an athlete and for a patient. Your rules decide what each of them hears.
An answer to why it moved
Each score comes with what it measures, so your app can say what moved it last night.
Who it is for
Teams that put a score in front of their users
Digital health products
A daily score that gives users a reason to open the app, shown in your own interface and your own words.
Learn moreSports and nutrition brands
Recovery and activity behind training and fuelling advice, from the watch the athlete already owns.
Learn moreCorporate and employer wellness
One consistent score across a whole member population, whatever mix of devices they wear.
Learn moreHealth data platforms
Scores to offer your own customers, computed where their data already sits.
Learn more
How it compares
Where the score runs, and whether you can read it
Two questions decide most score choices: where the data is processed, and whether you can explain the number to the people who see it.
| A hosted API's score | Build it in-house | Open Wearables Health Scores | |
|---|---|---|---|
| [01] Where it is computed | In the vendor's cloud | In your cloud | Inside your own Open Wearables deployment |
| [02] How it is validated | By the vendor, rarely explained | By your own science team, once you have one | By our science team before each release, with what each score measures written down |
| [03] What it takes | An API key and a per-user contract | Health scientists, data and months of work | Open Wearables running, and the add-on switched on |
| [04] Choose this when | Speed matters more than control over the data | Scoring is your core product and you have the science team | You want scores in your product and need to explain where they come from |
Who builds the scores
Built by our health science team
The scores come from the scientists and engineers who maintain Open Wearables, the same people who answer questions about them on Discord.
Anna Zych
Neuroscience PhD
Leads the science behind the health scores: which inputs each one reads and how they are weighted. Writes Science Behind Wearables, a newsletter on what wearable data can and cannot tell you.
Filip Begiełło
ML Lead
Leads machine learning on Open Wearables. Builds the scoring layer: the code that computes each score inside your deployment and over the history it already holds.
Build scores on the devices your users already wear
Every provider lands in one standard schema, and the scores read that schema rather than any one device. Support another wearable later and the same score works on it, as long as its inputs are in the data.
Add health scores to your product
Health scores are part of Pro, computed inside your own deployment.
FAQ
What teams ask about the scores
How is a sleep score calculated?
Ours reads the length of the night, its sleep stages, how regular it is and how often it was interrupted, and turns them into one number from 0 to 100. Our science team designed it and validates it before each release.
What is a good sleep score?
Higher is better on a 0 to 100 scale, but the useful reading is the trend for one person over weeks rather than a single night against a fixed line. That is why scores are calculated over the history you already hold.
What is a recovery score?
A morning number from 0 to 100 on how ready the body is for load. Training and coaching products use it to decide what to suggest that day.
Is Open Wearables still free?
Yes. The core stays open source under an MIT license. Health Scores is a licensed module that runs on top of it.
Are the scores a medical device?
No. They are wellness scores, built to show trends and patterns over time, and should not be used as a clinical or diagnostic tool.
Which wearables do the scores work with?
They work across providers, because they read the standard Open Wearables schema. A score depends only on whether its inputs are present in the data, and that differs by device, so we check coverage with you.
How is this different from the Apple Watch or Fitbit sleep score?
Apple Watch, Fitbit, Garmin and Oura each calculate their own sleep score, with a method you cannot inspect. Ours is computed by Open Wearables from the raw sleep data the device syncs, so every user gets the same calculation whatever they wear. The device's own score is still stored, labelled with its source, so your app can show both side by side.
What happens to a user's score when they switch from a Fitbit to an Apple Watch?
Their score keeps the same method and scale, so the history before and after the switch stays comparable. The amount of detail can change: deep and REM sleep feed part of the score, and how much stage data a device reports differs between models.
Do we need Open Wearables running first?
Yes, the scores run inside it. If it is not running yet, deploy it yourself for free or have us deploy it.
Can you build a score we need?
Yes. Our science team builds new scores with you and validates them the same way. Tell us what you want to measure when you book a demo.
What does it cost?
Health scores are part of Pro, priced per deployment rather than per user. Pricing has the figures.