Product specifications, compared against the category

A specification means very little on its own. Every site here reads a whole product category and works out what an ordinary unit in each class actually looks like, so you can see where any particular one sits against it.

Most buying advice tells you which product is best. These sites tell you what the market looks like, so you can see where any particular unit sits in it. A 7 cu ft chest freezer sounds generous until you know the median in its class, and a 1,200 MB/s SSD sounds quick until you know what its own interface is capable of. Knowing that is the difference between a decision and a guess.

The sites

Each site owns one category and holds its own catalogue, its own baselines and its own finder. They are separate because their specifications are not comparable: nothing useful is shared between a freezer's capacity and a mini split's SEER rating, and a baseline that mixed them would describe neither.

• EV Charger Data

  Home charging hardware, by amperage and installation type.

• Freezer Data

  Chest, upright, compact, convertible, garage-ready and portable 12 V freezers.

• Mini Split Data

  Ductless heating and cooling, by capacity and efficiency class.

• Power Station Data

  Portable power stations, their real capacity and what they will actually run.

• Robot Mower Data

  Robotic mowers, by coverage, slope and boundary technology.

• SSD Data

  NVMe, SATA, portable and PS5 solid-state drives, read against the bus they actually run on.

• Hard Drive Data

  NAS, surveillance, enterprise, desktop and external spinning drives, and the price per terabyte behind them.

What a baseline is, and why it is the whole point

A single average across a whole category would be meaningless. Baselines are computed per class and per size band, so a buyer shopping at one end of a category is not handed a number pulled toward the other end. Where a class holds fewer than eight products, no baseline is published — below that a median describes the sample rather than the market, and would move if one more listing appeared.

What we do not do

We do not physically test products, and every site says so on every page. There is no lab here. What we have is a whole catalogue read consistently, which is a different kind of useful and an honest one.

We do not publish star ratings or review counts, and we do not print exact prices — prices change daily, and a figure captured on a scrape date is misleading by the time you read it. Where a listing does not state a specification, we print not stated rather than a plausible default.

The methodology page sets out how a corpus is cleaned, how baselines are computed and where the process fails.

How this is paid for

Every site in the network earns a commission on qualifying purchases made through its links to Amazon. That pays for the work; it does not decide what the numbers say. A product appears in a comparison because its listing publishes the figures the comparison needs, and for no other reason. The full disclosure is here.