How we test

Imagine two air fryers side by side on a worktop. Same size, same price, both rated 4.7 stars. One will crisp chicken thighs for a family of four in 22 minutes. The other will steam them for half an hour and leave you scrubbing the drawer. From the outside there's no way to tell them apart. That's why this page exists.

Below is the method every one of our testers follows, in the same order, every single time. Nothing about it is secret or clever. It simply takes a lot of weeks.

1. Everything starts with the primary goal

Before we order anything, we write one sentence answering a plain question: what does "best" really mean for this product? We call it the primary goal, and it's the first line on every card on our tests in progress page.

Some real examples from the current bench:

• Cordless vacuum cleaners: which one cleans a whole three-bedroom home, carpets and pet hair included, on a single charge.
• Noise-cancelling earbuds: which one blocks the most engine noise on a long-haul flight and stays comfortable for six hours.
• Pellet and gas barbecues: which one holds a steady temperature through a 10-hour low-and-slow cook in cold weather.
• Automatic pet feeders: which one never skips a meal or jams across 30 days of twice-daily feeding.

The primary goal picks the winner. Everything else is a tiebreaker. It keeps us honest, too: it's written before a single parcel turns up, so we can't move the goalposts after falling for a clever feature.

2. The shortlist

We start broad. Best sellers at Argos, Currys, John Lewis and Amazon.co.uk, brands that are growing quickly, models that keep cropping up on forums and Reddit threads, and every product you ask us about by email. That normally gives us 12 to 20 candidates.

Then we whittle it down. Anything with a pattern of safety complaints, a product recall, dodgy-looking reviews or no reachable customer service is out before we spend a penny. We also drop models about to be discontinued, because a glowing review of something you can't buy helps nobody. What's left is usually 4 to 7 models, the ones a normal person is genuinely choosing between.

3. We pay the shop price, like you

Every test unit is bought with our own money from ordinary retailers, at normal prices, delivered like any other order. We don't request review units and we don't test free samples sent in return for coverage.

Buying this way also shows us what you'll experience: how the box arrives, how long delivery takes, whether the manual makes sense and whether the "quick setup" really takes nine minutes or forty. Those details go into the review.

All our testers live in the UK, so we buy the model sold here, from the same shops you use, from Currys and Argos to Boots and Lakeland. That means the UK plug, the 230-volt motor and the model number you'll find on the shelf. Bedding comes in UK sizes and every price is given in pounds.

4. Weeks of everyday life

Our testers use the products at home, on their own timetables. The hoover does the Sunday tidy-up after a birthday party in Clive's 1930s semi in Sheffield, springer spaniel included. The air fryer makes tea on a Tuesday in Tracey's kitchen in Leeds with two hungry boys round her ankles. The cabin bag goes through gate sizers at Manchester Airport with Karen, who took 71 flights in the past twelve months.

Most tests last between 21 and 60 days. Simple products such as power banks get three weeks. Anything with a battery, a filter, a foam core or a fabric that wears gets longer, because that's where cheap products give up. Smart rings and sleep trackers get up to 60 days, as a month of sleep data is barely enough to spot a trend.

5. The kit we measure with

First impressions count, but numbers settle arguments. Every tester has the same basic kit, and some categories get specialist tools.

We repeat every key measurement at least three times and take the middle result, so one lucky run never decides anything.

6. Scoring sheets and weightings

Each category has its own scoring sheet, written before testing begins. The criterion tied to the primary goal always carries the heaviest weighting. The weightings add up to 100, and value for money is always included. A product that's 5% better and costs twice as much rarely wins.

Here are six real examples of how the weightings are set.

Cordless and robot vacuums

Kitchen appliances

Sleep

Health and recovery devices

Tech and travel

Outdoor and barbecues

Each criterion is scored from 1 to 10 and multiplied by its weighting, and the total becomes the score out of 10 at the top of every review. We round to one decimal place, and we never round up to make a number look tidier.

7. How we read customer reviews

A few weeks with one unit can't reveal everything. Perhaps we got a good one. So for every product we also read verified customer reviews in bulk, typically 300 to 500 per model across two or three retailers.

We pay little attention to 5-star reviews written the day the parcel arrived. The useful ones are the 1 and 2-star reviews written after months of use, plus the 3-star ones where people explain what they'd change. We group the complaints and count them. One cross review about a dead battery is bad luck. Forty describing the same fault after four months is a pattern.

If a pattern appears, it counts against the score even if our own unit held up. It can cost a product up to a full point, and if it's a safety issue the product is out of the running altogether. We ignore complaints about delivery, late presents and anything that's clearly about the retailer rather than the product.

8. The verdict and the write-up

When a test finishes we set the results against the primary goal and ask one question: which product did the job best for a normal person at a fair price? That's the winner. The tester who lived with it then writes the review in their own words, with the figures, the niggles and the funny moments along the way.

Every review states how long we tested, what we measured and who it suits, and just as plainly who it doesn't. You'll find them all in our reviews.

9. When we retest

Every review shows an "updated" date. We go back to a product when:

• a new model or version launches that changes the comparison
• the price moves by roughly 20% or more, in either direction
• a firmware or app update fixes, or breaks, something we wrote about
• readers report a problem we didn't see in our test
• six months have passed on a popular pick, so we can check long-term wear

Whenever possible we keep the winning unit for exactly these follow-ups. The six-month check is often where the real story lies: a hinge that creaks, a battery now managing 70% of what it did, a pillow that's gone flat on one side. When that happens the verdict changes and we say so on the page.

10. Where the products go afterwards

A test leaves us with a lot of kit. This is what happens to it:

• The winners stay with the tester for long-term follow-up, which is how we can tell you how something is holding up after six months.
• Working products that didn't win are cleaned, reset to factory settings and donated to local schools, family shelters and community centres. Pet products go to animal rescues, which have been delighted with a steady supply of feeders and litter trays.
• Broken products and anything unsafe go to certified electrical recycling, with batteries removed and recycled separately.

We never send opened test units back to the shop for a refund, and we never resell them as new. Nobody should end up with a product we've spent six weeks putting through its paces.

Spotted something we missed?

Our readers are the biggest test panel we've got. If something you bought on our advice let you down, or you think we've got a figure wrong, let us know. Reader reports have sent more than one product back to the bench, and you can see what's on it right now in our tests in progress.