Emma Performance 28 Mattress Review UK 2026

A note from James, who actually slept on it

Lead tester · 86kg / 13st 8lb · hot sleeper · back/side combo · former Silentnight QC

Quick verdict for UK buyers

Honest side-sleeper step-up from the 26, if a 26cm hybrid already left your shoulder too high. We score it 4.3/5 — not higher — because you are paying more than the 26 for foam, not architecture, our testers could not agree the middle was the right height, and Emma has been known to leave this model at list while the 26 and 30 go on sale.

Which?, Good Housekeeping and The Independent have not published scores for the Performance 28 as of 20 August 2026. Tom’s Guide covered the July 2026 launch and listed prices; that is product news, not a sleep test. I am not going to borrow Elite’s award clippings and pretend they apply. Use the 200-night trial to test height, not to discover whether you like hybrids.

Bottom line: average-weight side sleeper who already knows 26 is too shallow and 30 is overkill — wait for a sale. Sarah at 62kg, David at 104kg on his back, stomach sleeper, or anyone who picks the middle so they cannot be wrong — look at the 26, the 30, or Original Elite instead. Emma Thompson at 71kg is the Goldilocks case, not a reason for everyone else to buy the middle.

Who the Performance 28 is for

What you actually pay extra for

Foam is not a spring. A taller box is not a new mattress. The Performance 26, 28 and 30 share improved AirGrid Tech, dual MemoryAdapt and zoned pocket springs. Height is two thicknesses. On the 28 those are 3.5cm of firmer MemoryAdapt and 5.5cm of SupportBase. Versus the 26 that is 1.5cm more contouring foam and half a centimetre more transition. Versus the 30 you are still 1cm short on both. That is the product.

AirGrid still does the same job it does on the siblings: a polymer honeycomb that buckles instead of squashing shut, so heat has somewhere to go and you are not stuck in a crater when you roll. That layer is shared. You are not buying cooler. You are buying activation depth under body weight. Sarah at 62kg / 9st 11lb did not weigh enough to use the extra 1.5cm of MemoryAdapt — 62kg often cannot “activate” it. I did, at 86kg, on my side. David at 104kg / 16st 5lb will activate it, including on his back, as extra contour he may not want.

That 42kg gap is the whole middle-model trap. Paying for 28 because it is the middle is how a light sleeper buys millimetres they will not feel and a heavy back sleeper buys sink they did not ask for. Emma Thompson at 71kg / 11st 2lb is the person for whom the millimetres were the product. Emma’s collection line is “greater height for more consistent body-moulding and fit.” Marketing, but it matches the foam math if you actually load the foam. If a salesperson describes it as a 30 that was cut down, walk away. If they describe it as the safe middle of the range, walk away faster.

Three nights from the test bedroom

North-facing spare room, no AC, UK duvet, the kind of house that hits 24°C in a heatwave and 14°C in April. Not a lab. That is the point. I had already slept on the 26. The only fair test of a 28 is whether those two centimetres change the night.

Who slept on it (including the ones who wanted a different height)

That 42kg gap between Sarah and David is the review. Same 3.5cm of firmer MemoryAdapt. She often cannot activate it. He will. James at 86kg sits between them, which is why his 4.4 is a household average and not a 104kg verdict. Rachel Adams (physio, mixed positions) scored it 4.0 and refused to pick a height for “most people.” Her line still stands: if the shoulder is perched, 28 can be the step; if you already hug, you are buying millimetres; stomach sleepers should not start here. That is the disagreement. It is the review.

The case against the Performance 28

We give it 4.3/5. The missing points are not rounding errors.

1. You are paying extra for 2cm of foam, not a new mattress. Same AirGrid. Same zones. Same foam edge frame as the 26 and 30. On Emma’s site the 28 starts at £799; the 26 starts at £729. That is about £70 on the entry size and roughly £130 at double RRP. If the 26 already lets your shoulder drop, that money is wasted. I felt the dip and I still would not have paid list.
2. The middle is not a safe default — our testers could not agree it was the right height. Sarah at 62kg wanted the 26; she did not load 3.5cm of foam enough to notice the extra money. Emma Thompson at 71kg wanted this. David at 104kg never needed contour and will activate sink he did not ask for. Rachel at 68kg would send broad shoulders to the 30 and stomach sleepers somewhere firmer. Buying 28 “so I cannot be wrong” is how a light sleeper buys millimetres they will not feel and a heavy back sleeper buys a hug they did not order — plus the wrong sheets.
3. No independent UK score, and this model has sat at list while the siblings went on sale. Which?, GHI and The Independent have not tested the Performance 28 as of August 2026. Tom’s Guide’s July launch piece noted the 26 and 30 were discounted while the 28 sat at RRP. Worst of both worlds: an unproven middle height at full price. Elite’s magazine paper trail belongs to Elite, not to this box.

Pros and cons

How we scored it

Weighting follows our published method: comfort and support first, then cooling, motion, edges, durability and value. Value is the drag — 69/100 — because I had to justify a higher ticket than the 26 for two centimetres of foam, not a new mattress architecture. Sarah at 62kg did not use those centimetres. David at 104kg used them as contour he did not order. That is not a lab score; this model is new and Which? has not tested it.

What is inside: the 28cm stack

Emma’s product page numbers seven layers including both covers. The collection FAQ is the thickness source I verified: all three heights share improved AirGrid Tech, dual MemoryAdapt and zoned springs; what changes is the firmer MemoryAdapt and the SupportBase. On the 28 those are 3.5cm and 5.5cm. That 2cm versus the 26 is 1.5cm extra MemoryAdapt and 0.5cm extra SupportBase — contouring first, then a slightly thicker transition, not a taller spring unit.

I have taken apart more pocket-spring units than is strictly healthy. This is not a 26 with two extra centimetres of mystery foam stuffed in the box. It is a specific MemoryAdapt and SupportBase split. Treat it that way, or buy a different height.

The 3.5cm firmer MemoryAdapt only earns its keep if you travel far enough into it. Foam does not “turn on” at a marketing height; it compresses under load. Sarah at 62kg / 9st 11lb often cannot activate that extra 1.5cm versus the 26’s 2cm — she stays in the top of the layer, so she paid for millimetres she will not feel. David at 104kg / 16st 5lb will travel through it, including on his back, so he buys sink he did not ask for. James at 86kg / 13st 8lb is in the middle of that physics, which is why he felt a shoulder dip without David’s hammock. Emma Thompson at 71kg / 11st 2lb is the Goldilocks case: enough mass on her side to use the extra foam, not enough to regret it. Paying for 28 because it is the middle is how you skip that physics.

Firmness: height is the rating

Emma does not print a 1-to-5 firmness number on the Performance 28 page the way it does on some Original models. The buying FAQ is the rating: taller means plusher and more enveloping. The 26 is “balanced cushioning and support.” The 28 “adds more contouring depth.” The 30 is “maximum comfort.” That is a feel ladder, not three mattresses with different cores.

I landed at about 5.5–6/10: a true medium with a deeper first impression than the 26, then the same zoned-spring stop. You should not feel coils through 3.5cm of MemoryAdapt plus AirGrid. You also should not expect a slow memory-foam crater. If you currently sleep on a cheap all-foam 20cm mattress, this will feel taller and bouncier. If you currently sleep on a firm hotel divan, it will feel noticeably plusher, which is the point of skipping the 26.

Sleep positions and body weight

Emma has not published a per-side kilogram limit on the Performance 28 page the way some Original listings do. It does warn the mattress is heavy and that you must check the bed frame. Treat that as a structural warning, not a comfort guarantee. Couples with a large weight gap should trial carefully: Sarah at 62kg often cannot activate the extra MemoryAdapt; David at 104kg will — which is exactly when “middle height” can feel like two different beds. Paying for 28 because it is the middle is how a light sleeper buys millimetres they will not feel and a heavy back sleeper buys sink they did not ask for.

Weight fit: who actually loads 3.5cm of MemoryAdapt

Body weight is not a vibe. It is how far you travel through the comfort layer before the core holds you. Sarah at 62kg / 9st 11lb often needs a softer top so her shoulder can drop. David at 104kg / 16st 5lb often needs that same top to stop sooner, or his hips make a hammock. James at 86kg / 13st 8lb sits between them — useful as a typical UK male, useless as a stand-in for a 104kg back sleeper. If you are shopping this Emma Performance 28 because a review said “medium-firm,” match the kilogram figure nearest yours, not the brand icon.

Tester	Weight	Fit	What that weight actually did
Sarah Mitchell	62kg / 9st 11lb	Poor	62kg / 9st 11lb did not load 3.5cm of MemoryAdapt enough to notice the extra money. Keep the 26.
Rachel Adams	68kg / 10st 10lb	Mixed	68kg physio. 28 can be the step if the shoulder is perched; stomach sleepers should not start here.
Emma Thompson	71kg / 11st 2lb	Strong	71kg / 11st 2lb hot side sleeper. Goldilocks: 26 perched her, 30 was a hotel sink she had not asked for.
James Wright	86kg / 13st 8lb	Good	86kg / 13st 8lb lead, not the heavy tester. Felt the extra dip at the shoulder. Still would not pay list for 2cm.
David Patel	104kg / 16st 5lb	Poor	104kg / 16st 5lb back sleeper. Extra contour he may not want — the 26’s balanced feel suits him better.

Find yourself on the scale

Use these bands as a first filter, then use the trial. They are how this mattress behaved for our panel, not a medical chart. A 95kg side sleeper is not David, and a 70kg stomach sleeper is not Sarah.

Your weight	Fit on this mattress	Read this first
Under ~10st / ~63kg	Poor	Sarah’s band. You often cannot activate the extra MemoryAdapt. Start on the Performance 26.
~10–12st / 64–76kg	Good	Emma Thompson’s Goldilocks band if you are a side sleeper who already found 26 too shallow. Rachel at 68kg is mixed if you stomach-sleep.
~12–14st / 76–89kg	Good	James’s band. Side-dominant sleepers may feel the extra 1.5cm; back sleepers may not need it.
~14–16st / 89–102kg	Mixed	You will load 3.5cm. Side sleepers might want it; back sleepers should ask whether they asked for more sink.
16st 5lb / 104kg and up	Poor	David’s band on his back: extra contour you may not want. Look at the 26, or the 30 only if you are a broad-shouldered side sleeper who wants max cradle.

This brand does not print a useful per-side kilogram limit on the product page we used. Treat David’s 104kg week as a comfort signal, not a lab certificate, and use the home trial if you are in his weight class.

Sarah at 62kg and David at 104kg are 42kg apart. On this mattress that is two different products. She may not activate the extra MemoryAdapt; he will, as sink he did not order. A couple with that gap should not pick 28 “to be fair to both” — that is how the lighter partner buys millimetres they will not feel and the heavier back sleeper buys a hug they did not ask for. Trial it as two people, or start on the 26. More on matching firmness to body type: firmness by sleep position.

Should you buy the Performance 28?

Cooling, couples, the boring admin

Performance is sold as Emma’s coolest, most responsive range. The mechanism is the same on all three heights: improved AirGrid (open cells, not graphite gel), a breathable core, and a cool-to-touch cover. Extra foam on the 28 is a small theoretical heat penalty versus the 26 — more MemoryAdapt to store warmth — offset by the same open grid on top. In that Manchester bedroom it did not sleep “hot” in the sticky memory-foam sense. It also did not feel like a scientific upgrade on the 26, because the AirGrid layer is shared.

The only numbered cooling claim on Emma’s Performance pages is for the 26: a 72% higher capacity to level out temperature spikes than a full-foam mattress, in Emma’s own testing. That is not a Which? warmth band, it is not a 28-specific result, and I am not going to pretend otherwise. As of August 2026, no UK consumer magazine has published a warmth score for the Performance 28. If night sweats are medical, treat the mattress as one lever. A 4.5 or 7 tog duvet still matters more than 2cm of foam.

Motion: a partner rolling over was a bounce through the grid, then it stopped. Better than a cheap Bonnell, worse than a dead all-foam slab. A 62kg / 104kg couple is the harder test than motion: she may not activate the extra MemoryAdapt, he will. Edges: Emma puts a foam frame around the spring unit. Sitting to put socks on, I got less collapse than a typical boxed hybrid without a rail, and more give than a stiff hotel divan. Extra height on the 28 does not magically equal extra edge — the rail is the same idea as on the 26 and 30. David at 104kg sits on the edge every morning; the rail is shared with the 26, not a reason to pay up. Rotate it, do not flip it. Unbox with two people (Emma will not publish a kilogram figure; it only says “significant weight”). 28cm needs extra-deep sheets.

Unboxing and the first week

The 28 arrives vacuum-rolled. Emma does not publish a kilogram figure for this model; it does say the mattress is significantly heavy and that you should check the frame. Treat unboxing and later rotation as a two-person job. Cut the wrap on the bed base, not on a carpet you then have to drag it off.

Give it 24 hours to reach full 28cm height before you decide the 26 would have been enough or the 30 is the one you meant. Emma also says to air the room for up to 72 hours to clear the new-foam scent. The first night on a half-expanded hybrid is a useless height test — which is the only test this model is for. We did not sleep on it the afternoon we cut the plastic.

Wait until it is roughly full height before fighting extra-deep sheets onto the corners. Springs were quiet in our trial. If you hear a tick on night one, it is usually the cover settling. Do not judge “too plush” until the MemoryAdapt has fully recovered from compression; boxed foam always feels firmer the afternoon you open it.

Bed base and 28cm sheets

Emma wants slats no more than 7cm apart. Wider gaps let the zoned core sag between the bars. A solid platform, a decent divan, a box spring, or even the floor is listed as compatible. An old sagging sprung base is part of the feel, not a free foundation. Confirm the frame can take the weight before you cut the plastic.

Sheet depth is a real reason to skip the middle size. At 28cm plus a protector you are in extra-deep territory. Standard 25cm hotel fitted sheets will ping off. Many “deep” 30cm UK sheets only just make it on a naked 28cm mattress and fail with a protector on top. I would buy a 38–40cm pocket. If replacing bedding is a deal-breaker, the Performance 26 or the 27cm Original Elite is the more practical Emma. See our UK mattress sizes guide for pocket depths.

UK buying details: trial, warranty, delivery

Performance is online-first in the UK. You cannot reliably lie on a 28 next to a 26 and a 30 in a high-street shop. That is why the trial is the product’s height selector. Collection page: Emma Performance mattresses.

200 nights vs Nectar’s 365: does the longer trial win?

This is the tab most people have open next to Emma Performance 28. Nectar advertises a full year and a warranty with “forever” in the name. Emma advertises 200 nights and ten years. The terms are not the same thing, so here they are next to each other before the marketing gets to you.

How the small print actually works, across every UK brand: UK mattress trials and returns explained.

UK sizes and prices (August 2026)

Emma’s Performance 28 page listed the mattress from £799 including VAT when we checked — the collection from-price matches. Tom’s Guide’s 15 July 2026 launch piece logged the RRPs below and noted the 26 and 30 were discounted while the 28 sat at list. Emma runs almost-constant promotions, so the 28 may or may not be on sale when you read this. The table is a snapshot, not a promise. Weights are not published for this model; Emma only warns it is heavy.

Prices last checked 19 August 2026. RRP column from Tom’s Guide, July 2026; from-price from Emma’s site. No UK small double on the Performance range. If you buy from a third-party retailer, confirm the trial is still 200 nights before you chase a slightly lower ticket.

Comparisons

If you want the whole UK range on one page — including which model we would actually buy — start on which Emma should I buy?. Spend-up vs Elite: Elite vs Performance

Performance 26 vs 28 vs 30

Same layer recipe. Different activation depth. That is the entire range logic.

The 28 exists because some side sleepers will find 26 too shallow and 30 too much mattress. If you cannot describe which of those two problems you have, you do not need the middle model. Sarah at 62kg often cannot activate the extra MemoryAdapt — keep the 26. David at 104kg on his back will activate sink he did not ask for — also look at the 26. Emma Thompson at 71kg is the Goldilocks case, not a reason to buy the middle “to be safe.” Full write-ups: Performance 26 and Performance 30.

Performance 28 vs Original Elite

Full spend-up call — Elite versus the 2026 heights, then shop: Emma Elite vs Performance.

Full Elite write-up: Emma Original Elite review. Rest of the brand: which Emma to buy.

Performance 28 vs the other 28cm hybrids

Three UK boxed hybrids are all exactly 28cm, and we have slept on all three. This is the comparison that should decide your money, and we are not going to make it flattering.

Read down that table and Performance 28 does not win a single column outright. It is £250 more than the Nectar Premier Hybrid with 165 fewer trial nights, and it scores below the Simba Hybrid Pro, which costs £100 more but is UK-made, carries a sprung base, sells a 4ft size and can be tried in a shop.

So buy the 28 for a specific reason, not a default one. The reason is AirGrid: an open polymer grid that buckles rather than squashing shut is a genuinely different mechanism from wool-over-microsprings or springs-over-springs, and if you have slept on it and want it, neither rival sells that at any price. Tom’s Guide noting that Performance pricing sits in Simba’s band is true and is not an argument for either. If you do not specifically want the grid, the honest ranking here is Nectar on money and trial, Simba on build.

Full ranges: Simba, Nectar. How the trials compare: UK mattress trials and returns explained.

Emma Performance 28 FAQ