How we test, and why nothing is scored yet
One scorecard, published before any scores. Every review states exactly how it was tested, and the testing basis is never dressed up.
Testing-basis labels
Every review carries one of three labels. We never imply hands-on testing when it is the third.
| Label | Means | Scored? |
|---|---|---|
| Hands-on (N nights) | We used the product and report the number of nights. | Yes |
| Loaner unit (disclosed) | The manufacturer provided it, and the page says so. | Yes |
| Published sources and community feedback | A summary of what other reviewers, makers and users report. Every current review is in this category. | No |
The 100-point scorecard
| Criterion | Points | What we measure |
|---|---|---|
| Sleep-data accuracy and trend reliability | 25 | Total sleep time and wake detection compared with a sleep diary and a second device, and consistency over many nights. Sleep-stage accuracy is weighted lightly because stage minutes vary so much between devices. |
| Comfort and wearability at night | 20 | Fit, skin irritation, side sleeping, noise and heat, and whether the battery lasts the whole night or week. |
| Actionability of insights | 15 | Does the app tell you something you can change? Coaching quality, alarms and integrations. |
| True cost of ownership | 15 | Hardware plus 1-year and 3-year subscription cost, and which features sit behind a paywall. |
| Privacy and data control | 10 | Data export, deletion, third-party sharing policy and cloud dependence. |
| Reliability and support | 10 | App stability, connectivity, warranty and return policy. |
| Innovation and unique value | 5 | Anything competitors cannot do. |
| Total | 100 |
Sub-criteria are adjusted per category; a bed cover has no ring sizing to judge, for example.
How we will test
- Cross-device validation. Wear two or more devices on the same nights and log total sleep time, wake time and resting heart rate or HRV against a manual sleep diary. Raw nightly numbers will be published.
- Trust the trend, not the stage minutes. A Terra Research comparison of 2,144 overlapping nights found total sleep time agreed across devices within about 20 minutes (median), while sleep-stage minutes disagreed by more than 250 minutes on average (source). That is why stage accuracy carries limited weight. More in our accuracy guide.
- True cost. Device price plus subscription over one and three years, shown on every review today.
- Re-checks. Reviews are re-checked every 90 days or after a major price, firmware or model change, with visible dates.
How we handle sources
- Every factual claim links to where it came from. Company claims are labelled as company claims.
- Ratings from other publications are quoted with attribution and never presented as ours, and never placed in our structured data.
- Anything we could not confirm is listed under "What we could not verify" rather than left out.
- Prices carry the date they were checked.
Who does the work
SleepLab Guide is owned by Ben Shamon, who will receive, wear and test the products. Research, source-gathering and drafting are assisted by an AI agent, which does not claim any hands-on experience. See about and editorial policy.
Questions
Why does nothing have a score yet?
Because we have not tested anything yet. We only score products we have used ourselves for a stated number of nights.
Do brands pay for reviews or placement?
No. We do not accept payment for reviews, scores or placement. Free or loaned units are disclosed on the review.
Do affiliate commissions affect scores?
No. Commission rates never influence a score, a ranking or whether we cover a product. We currently have no affiliate links.
Can a brand approve a review before it runs?
No. Brands can correct factual errors, but they do not see or approve review text before publication.