Best smart mattresses
A mattress lasts eight to fifteen years. A connected device and the company running its servers frequently do not.
I earn a commission if you buy through some links on this page. It never changes what I recommend, and I publish negative verdicts on brands I earn from. Editorial independence.
Smart mattresses and connected sleep devices promise measurement: stages, heart rate, temperature, a score in the morning.
Two problems sit underneath that, and neither is about whether the technology works.
The lifespans do not match. A mattress is a ten to fifteen year purchase. A connected device depends on an app, a server and a company continuing to exist, and this category has a poor record: Eve went into administration in 2022, Brook + Wilde was liquidated in 2023, and Casper and Leesa have left the UK entirely.
Measurement does not equal improvement. Knowing you slept badly is not the same as sleeping better, and for some people it makes things worse.
The picks
The pick, if you want one answer
Origin LumbarCloud Mattress, 8.2 / 10. A mattress that does its job for fifteen years without depending on a server, which is what the money is better spent on.
| Product | Best for | From | Score |
|---|---|---|---|
| Origin LumbarCloud Mattress | Best mattress instead | £797 | 8.2 / 10 |
| Tweak | Best for the problem tracking usually reveals | Mid range | 7.8 / 10 |
| Silentnight Dual Climate Mattress | Best temperature answer without electronics | Mid range | 7.0 / 10 |
| Otty Original Hybrid Mattress | Best structural cooling | Around £500 | 8.1 / 10 |
Origin LumbarCloud Mattress: Best mattress instead
£797 · 8.2 / 10
Thirty-five centimetres and up to 8,000 springs. Smart mattresses sit in the premium bracket, and this is what that money buys in construction: more depth and more springs than mattresses at twice the price, with nothing to update or subscribe to.
Buy it if
You want the money to go into construction rather than electronics.
Do not buy it if
You specifically want measurement.
Read the full Origin LumbarCloud Mattress review
Tweak: Best for the problem tracking usually reveals
Mid range · 7.8 / 10
Scores 7.8. A different firmness on each side. The commonest finding from sleep tracking in a shared bed is that two people need different things, and dual firmness addresses that directly rather than reporting it back to you every morning.
Buy it if
You and your partner disagree and want that solved structurally.
Do not buy it if
You both want the same firmness.
Silentnight Dual Climate Mattress: Best temperature answer without electronics
Mid range · 7.0 / 10
Scores 7.0. Two seasonal sides, flipped twice a year. The flipping habit is the real benefit, and it makes the point that a mechanical answer to a temperature problem has no firmware and no subscription.
Buy it if
Temperature is the variable you wanted a smart mattress to manage.
Do not buy it if
You want active heating or cooling.
Read the full Silentnight Dual Climate Mattress review
Otty Original Hybrid Mattress: Best structural cooling
Around £500 · 8.1 / 10
Scores 8.1. A 16 centimetre spring core moves air continuously without power. Heat leaves a mattress when air moves through it, which is structural rather than electronic, and it keeps working in year twelve.
Buy it if
You wanted active cooling and would accept a passive answer.
Do not buy it if
You need a specific temperature held.
Read the full Otty Original Hybrid Mattress review
What consumer sleep tracking can and cannot tell you
Reasonably good at: total time in bed, rough sleep and wake patterns, movement, heart rate and its variability, and trends over weeks. Those are genuinely useful for spotting patterns.
Weaker at: sleep stages. Consumer devices infer light, deep and REM from movement and heart rate rather than measuring brain activity, and the accuracy against clinical polysomnography is limited. A nightly stage breakdown is an estimate presented with more confidence than it deserves.
Not able to: diagnose anything. If you have symptoms such as pauses in breathing during sleep, gasping awake, or heavy daytime sleepiness, that needs a GP rather than an app.
The problem with watching your own sleep
Worth knowing before you start
Clinicians have described a pattern sometimes called orthosomnia: anxiety about achieving a good sleep score that itself makes sleeping harder.
Sleep is one of the few things that gets worse when you try harder at it. If you notice yourself checking a score before deciding how you feel, that is the signal to stop tracking rather than to optimise further.
If sleep is a persistent problem, the next step is a GP rather than a sensor. The NHS says to see one when changing your sleeping habits has not helped, and that cognitive behavioural therapy is sometimes offered from there, in person or online. No mattress substitutes for that.
The service risk, spelled out
| Component | Expected life | Depends on |
|---|---|---|
| Spring unit | 8 to 15 years | Nothing |
| Foam comfort layer | 3 to 10 years, by density | Nothing |
| Pump, sensors, controller | Unknown | Parts availability |
| App and account | Unknown | The company continuing to exist |
| Cloud processing of your data | Unknown | Servers, and the terms you agreed to |
Before buying, ask three things: does the mattress still function as a mattress if the app stops working, is there a subscription, and what happens to your sleep data. Sleep data is health-adjacent information, and the privacy policy is worth reading rather than accepting.
Check the trading entity at Companies House too. It is free and takes two minutes, and it matters more here than on an ordinary mattress.
What actually improves sleep, in order of effect
A dark room. Blackout blinds or lining do more than any device. Light is what shifts your body clock earliest.
The right duvet tog. Seven to 10.5 suits most modern UK bedrooms, and 13.5 is a convention from draughtier housing. Waking hot is more often this than the mattress.
The right pillow height. Free to check: lie in your sleeping position for ten minutes and have someone look from behind.
A consistent wake time. More effective than a consistent bedtime, and it costs nothing.
Then the mattress. And if you share a bed, two single duvets solve more shared-bed complaints than any technology.
What not to buy
A connected mattress that stops being a mattress if the app dies. Ask explicitly.
Anything with a subscription for core functions.
Sleep stage data treated as fact. Consumer devices infer stages rather than measuring brain activity.
A tracker, if you are anxious about sleep. Measurement can make that worse.
Technology instead of a dark room and the right duvet. Both do more.
Also considered, and why they are not here
A roundup is only as useful as what it leaves out, so here is what was assessed and rejected for this list.
| Product | Why it is not on the list |
|---|---|
| Zeeq Pillow | Not scored. A connected pillow whose useful functions depend on an app and a service continuing to exist, which is the whole argument of this page in a single product. |
| Oura Ring | Scores 7.0. A tracker rather than a mattress, and the better shape for this category: if the service ends you have lost a ring, not a bed. |
| Simba Hybrid Pro | Good construction at £1,049 with no electronics. A refurbished Hybrid Ultra has listed at around £699 against £1,749, which is the best value in that range. |
| Origin Hybrid Pro | The highest scoring mattress here at 8.4 for £374, which is the comparison any premium smart mattress has to beat on construction. |
How this list is made, and who made it
Independence
No brand pays to appear on this site and no placement is sold. Some links earn a commission if you buy, which never changes a score, an order or a verdict. Several of the lowest scores here are on products that carry commission, and several recommendations carry none.
Who writes this. Kyle Klinkradt has worked in the mattress and bedding industry for over ten years across retail, distribution and brand partnerships, and has handled, sold or assessed products from every major UK bed-in-a-box brand. Every review on this site is written by one person, so the scores mean the same thing from page to page.
What the assessment is. Construction specifications against what they physically do: spring height and where in the mattress it sits, spring count for the size, gauge, whether the base is sprung or foam, foam density where it is published, comfort layer depth and material. Then the commercial terms that decide what you are actually buying: trial against returns policy, guarantee length against the company's UK trading history at Companies House, and the regulatory record.
What it is not. This is not a laboratory. I do not publish pressure-mapping data or claim to have slept eight years on every mattress here, because nobody has and pages that imply otherwise are not telling you the truth. Where a judgement rests on construction rather than long-term use, the review says so.
What gets a product removed. Leaving the UK market, a specification change that makes the review inaccurate, or a corrected fact. Brands that have withdrawn are kept on the site with the score removed rather than quietly deleted, because people still search for them.
See editorial independence, corrections and the author page.
How these were chosen
Every product here has its own review on this site with a score out of 10, and the score is a judgement rather than an average. Nothing appears in a list because it pays more.
A score is against the market, not against its own range. The cheapest mattress in a lineup is usually priced to make the model above it look reasonable, and the flagship of a budget range has risen into a bracket where better brands compete. Both are judged against what else you could buy for the money.
Everything listed can actually be bought in the UK. That is not a given: Casper withdrew in July 2020 and Leesa no longer sells here, and both still appear in current British roundups with affiliate links attached.
I do not pad the list. Where only three or four products genuinely qualify, there are three or four. A list of seven when the category contains four good ones is a list with three bad recommendations in it.
Prices move constantly. Judge the cash price on the day rather than any percentage beside it. See every review on this site or where the real savings are.
Related guides
Other pages on this site covering the same ground.
- Best zip and link beds and mattresses
- Best wool mattresses
- Best waterproof mattresses
- Best travel cot mattresses
Questions people ask
Are smart mattresses worth it?
The lifespans do not match: a mattress is a ten to fifteen year purchase and a connected device depends on an app, a server and a company continuing to exist. Ask whether it still works as a mattress if the app stops, and whether there is a subscription.
How accurate is consumer sleep tracking?
Reasonable for time in bed, movement, heart rate and trends over weeks. Weaker on sleep stages, which consumer devices infer from movement and heart rate rather than measuring brain activity, so a nightly stage breakdown is an estimate.
Can a sleep tracker make sleep worse?
It can. Clinicians have described a pattern sometimes called orthosomnia, where anxiety about achieving a good score makes sleeping harder. If you check a score before deciding how you feel, that is a signal to stop tracking.
What improves sleep more than technology?
A dark room, the right duvet tog at 7 to 10.5 for most modern UK bedrooms, a pillow at the right height, and a consistent wake time. If sleep is a persistent problem, a GP is the next step rather than a device, and the NHS treats cognitive behavioural therapy as something offered sometimes rather than as the standard answer.