Mattress buying is unusual among large purchases because the thing you are assessing cannot be assessed quickly. You spend a third of your life on it, the effects take weeks to appear, and the test available to you in a shop — lying down fully clothed, in a bright room, for ninety seconds, while someone watches — bears no relationship to the actual use.
This is why the trial period matters more than any specification. It is the only mechanism that lets you evaluate the product the way you will use it. Everything else in this guide is about narrowing the field before you get to that point, so the mattress you sleep on for a month has a decent chance of being right.
Sleeping position narrows the field first
How you sleep does more to determine what suits you than any other factor, and it is the question a shop assistant is least likely to ask.
Side sleepers need the mattress to yield at the shoulder and hip so the spine stays straight. Too firm and those points are pushed up, bending the spine sideways all night. Back sleepers need support under the lumbar curve with less give at the shoulders. Front sleepers need firmness above all, because a soft mattress lets the hips sink and arches the lower back into a position it will complain about by morning.
If you share a bed with someone whose position differs from yours, that conflict is the central problem to solve, and it is usually solved with a medium mattress plus a topper on one side rather than with a compromise that suits neither.
Firmness ratings are not a standard
There is no industry definition of medium-firm. One manufacturer’s medium is another’s firm, and the same mattress feels different to two people of different weights, because firmness is a relationship between the surface and the load on it, not a property of the surface alone.
A lighter person does not compress the comfort layers as far, so a mattress described as medium can feel firm to them. A heavier person compresses through the comfort layer to the support layer beneath, so the same mattress feels softer and then abruptly harder. This is why weight belongs in the conversation and rarely appears in it.
Treat firmness labels as a rough sort rather than a measurement, and rely on the trial to settle it.
The same logic explains a complaint that comes up constantly: a mattress that felt perfect in the shop and wrong at home. In a showroom you lie on it for two minutes, usually on your back, usually clothed, on a floor-level base, in a warm room. At home you are on a bed frame at a different height, under a duvet, moving between positions for eight hours. Almost none of the variables match, so it would be surprising if the impressions did.
Heat is the variable most people fail to anticipate. A mattress that cradles you closely has more of its surface in contact with your body, and contact is how heat transfers. Anyone who sleeps hot should treat a deep, enveloping comfort layer as a warning rather than a selling point, regardless of what the cooling technology on the label claims to do about it.
Construction, and what each type gets right
| Type | Strength | Weakness | Suits |
|---|---|---|---|
| Pocket sprung | Support with good airflow; movement is isolated | Comfort layer quality varies enormously | Most people, and anyone who sleeps hot |
| Memory foam | Pressure relief at shoulder and hip | Retains heat; can feel slow to move on | Side sleepers, light sleepers with a restless partner |
| Latex | Responsive, durable, cooler than foam | Heavy and expensive | People who want foam comfort without the sink |
| Hybrid | Springs for support, foam for pressure relief | The compromise can be mediocre at both | Couples with different positions |
Note that no row in that table is best. Each is a set of trade-offs, and the right one depends on the position and heat questions above. Anyone telling you a single construction is superior is describing a preference.
The trial period is the specification
Since you cannot judge a mattress quickly, the returns policy is doing the work a test would otherwise do. Read it properly, and check four things.
- How long is the trial, and does it start from delivery or from order?
- Is there a minimum break-in period before you are allowed to return it? Thirty days is common and reasonable.
- Who collects it, and is that free? Collection charges can be substantial for something this size.
- Is a refund or an exchange offered? An exchange-only policy is much weaker than it sounds.
A generous trial from a seller with straightforward collection is worth more than a modest improvement in specification. It is the difference between a decision you can undo and one you cannot.
It is worth being sceptical about very long trials advertised as a headline. A hundred nights sounds generous, but if collection is chargeable, or the refund is issued as credit, the practical protection is much weaker than the number suggests. Read the mechanism, not the duration.
The things people forget
Your base matters. A mattress designed for a slatted frame behaves differently on a solid platform, and slats spaced too widely will let a foam mattress sag between them. Check the maximum slat gap the manufacturer specifies, which is usually narrower than the gaps in an old frame.
Pillows matter almost as much as the mattress and cost a fraction as much. If your neck hurts, the pillow is the more likely culprit and the cheaper thing to change first. It is worth exhausting that possibility before concluding the mattress is wrong.
And timing matters. Mattresses follow a predictable discount cycle, and paying full price is largely avoidable if you can wait — the pattern is mapped out in the best times of year to buy each category.
What to know before you buy
- Identify your sleeping position honestly, including the one you wake up in rather than the one you fall asleep in.
- Read the trial terms before the specifications. The trial is what protects you.
- Check the slat gap your frame allows against the manufacturer’s requirement.
- Change your pillow first if the complaint is neck rather than back.
- Ignore firmness numbers across brands. They are not comparable.
- Give it a fortnight before judging. Almost every mattress feels wrong for the first few nights.
How long does a mattress last?
Long enough that this is a purchase to take seriously, and long enough that the difference between a good one and a poor one compounds nightly for years. The usual signs it is finished are a visible dip where you sleep and waking up stiffer than you went to bed. If either appears, replacement is overdue rather than optional.
Do I need to rotate it?
Yes, unless the manufacturer says otherwise, and it is the single cheapest thing you can do to extend its life. Rotating head to foot every few months evens out the compression. Flipping only applies to double-sided mattresses, which are now uncommon.
Is an expensive mattress better?
Up to the middle of the market, generally yes — better materials genuinely last longer and support better. Above that, price increasingly buys marketing, showroom presence and finish. The framework in five questions before any large purchase applies here as directly as anywhere on this site.
The practical approach is to narrow by position and heat, ignore firmness labels, then choose between the shortlist on the strength of the trial period rather than the specification sheet. Order the one you can most easily send back, and let a month in your own bedroom make the decision that no shop floor can.
