We Don't Test Products. Here Is What Our Rankings Actually Are

A lot of "best of" pages imply that someone bought the products and used them. We want to be clear that we did not. Nobody here has held the air fryer, sat in the office chair, or worn the running shoes. If you need a hands-on verdict, you need a publication that tests, and several good ones exist.

What we do is different, and narrower.

What we collect

For each product we track, we record what the store's own listing shows:

• the current price, and the list price if the store displays one
• the star rating and the number of ratings
• whether it is in stock
• the price again every time we check, which over weeks becomes a price history

That is all of it. There is no editorial opinion in the data and no information the store has not published.

What we do with it

We run every listing through one formula: star rating 40%, review volume 20%, price 25%, stock and freshness 10%, and a small fixed weight for the store, 5%. The listing with the highest score gets the "Best Pick" label, and the page shows the reasons that applied to it.

The formula is the same for every product, every store, and every affiliate commission rate. It is published in full on the How we pick products page, and a separate post works through the rating and review part with real numbers.

What this is good for

Sorting. A category page puts the products with the strongest ratings and the most buyers at the top. That replaces a fair amount of scrolling and tab-opening.

Checking a price. The price history shows what a listing has actually sold for while we have been watching. A "40% off" badge measured against a list price nobody was charging looks very different next to that chart. We wrote about this in Is That Deal Actually a Deal?.

Seeing the alternatives. Where we track the same product at more than one store, the page lists them all with prices, so you can see what picking a different one would cost.

What it cannot do

It cannot tell you how a product performs. A rating is an average of other people's experiences. It says nothing specific about noise, build quality, sizing, battery life, or anything else you care about in particular.

It cannot see problems that buyers have not rated yet. A product that fails after a year will hold a high rating for most of that year.

It does not know about recalls or safety notices. If a product has been recalled, a ratings formula can still rank it highly. Check the manufacturer's site or your country's product-safety authority for anything where safety matters, such as a child's car seat or an appliance that heats.

It only covers what we track. A product we have not listed is not worse; we have simply not seen it. Our price history starts on the day we began tracking a listing, so a "lowest price we've recorded" is the lowest in that window and not the lowest ever.

It can be wrong about stock and price. Prices change between our checks. The store's own page is always the authority on what you will be charged.

How we would use it

Use the ranking to get from two hundred listings to three or four. Open those. Read the low-star reviews on the store's own page, because that is where the recurring faults are described. Look at the price history to decide whether today is a reasonable day to buy. Then choose.

We earn a commission from some stores when you buy through our links. It does not change the ranking, and every page on the site says so at the bottom. If that ever stopped being true, this page would be the first thing we would have to rewrite.