Every product card on this site carries a score out of 5. A number like that is worthless unless you know how it was produced, so this page explains exactly that — and what it deliberately does not mean.

Where the gear comes from

Everything I test, I buy at retail with my own money. Nothing on this site was loaned by a brand, sent for free, or supplied in exchange for coverage. No manufacturer sees a review before it publishes, and no brand has any input into a score.

That's also the main reason this site grows slowly. Buying the gear is a real constraint on how much gets covered, and I'd rather have a small catalogue I can stand behind than a large one assembled from spec sheets.

Tested, or picked: every card tells you which

Some products here I own; others I've researched but not bought. The difference is on the card itself, in the banner across the top, and I never blur it:

  • "Bought and tested by me" — I own the unit, paid for it, and have used it long enough to have opinions the spec sheet doesn't contain. Every full review on this site is one of these.
  • "My pick" — a selection made from specifications, warranty terms, teardowns and consistent owner reports, without the unit passing through my hands. The strong and weak points are still mine to defend, but they come from research rather than use.

Most of the untested picks are whole-home backup systems costing several thousand dollars each, plus a couple of bundled kits made of components I already own individually. Buying four of those to compare them isn't something a self-funded site can do honestly, so I say what they are instead of implying a test that never happened.

A pick moves to tested when I buy the unit — never the other way round, and never because a brand asked.

The five criteria

Scores come from five criteria, listed here in roughly descending order of how much they move the final number.

1. Does it do the job it's sold for?

The heaviest factor. Judged against the class the product actually competes in, not against larger or more expensive gear doing a different job. A 300 Wh camping unit is scored as a camping unit.

2. Build quality and battery chemistry

Cell chemistry above all, because it determines whether you're buying something for two years or ten. Then the physical things: case, ports, connectors, junction boxes, and whether anything has loosened or corroded with use.

3. Living with it

What changes after the first week. Fan noise and when the fan decides to run, real carry weight rather than listed weight, standby drain over months, recharge behaviour, and whether the companion app helps or gets in the way.

4. Honest value

Cost per usable watt-hour, or per watt actually harvested, measured against the realistic alternatives in the same class — not against the product's own list price, and not against a discount that may not exist next week.

5. Deal-breakers

Any single flaw serious enough to override the other four. Expansion modules priced close to the base unit, an inverter that can't start the loads the product is marketed for, or a solar input so limited that "solar generator" is misleading.

What the numbers mean

Scores on this site cluster between roughly 4 and 5, and that's a property of the sample rather than generosity. I don't spend my own money on gear I expect to be bad, so products I'd score below 4 mostly never get bought — and when something does disappoint, it's more useful to explain the specific failure than to publish a low number.

So the useful signal is the gap between scores, and the weak points listed on every product card. A 4.3 and a 4.8 in the same category are meaningfully different products. Neither score is a recommendation to buy without reading why.

A score is also always relative to a class and a moment. A unit that scored well two years ago may be beaten today by something at the same price, which is what the "Updated" date on each page is for.

What would stop me recommending something

  • Battery chemistry that isn't disclosed clearly in the listing.
  • An inverter that can't run the loads the product is explicitly marketed for.
  • Expansion or accessory pricing that makes the advertised upgrade path unrealistic.
  • A safety concern, or any claim in the marketing I can't reconcile with the product in front of me.

Affiliate links, and what they don't buy

Some links on this site are affiliate links: if you buy through one, I earn a small commission at no extra cost to you. That's what funds buying the gear in the first place.

What that commission does not do is change a score, alter a verdict, or decide what gets covered. Products with no affiliate programme appear here on the same terms as ones that pay, weak points are published whether or not a link is attached, and I'll tell you when the honest answer is to buy nothing at all.

Where a product's price matters to the argument, I describe it in ranges rather than exact figures. Retail prices move constantly, and Amazon's programme terms don't permit displaying prices sourced outside their API — so the retailer's page is always the authority on what something costs today.

Corrections and updates

Pages get revised when something changes that would alter the advice: a newer model supersedes the unit, a fault appears with longer use, or I got something wrong. The "Updated" date reflects the last substantive revision, not a cosmetic edit.

If you think something here is inaccurate, tell me — corrections are made in the text rather than quietly.

Frequently asked questions

Are these scores based on lab measurements?
They are based on buying the product and using it, combined with verifying the manufacturer's stated specifications and weighing them against what other owners consistently report. They are editorial judgements rather than laboratory results, and this site does not claim to run standardised bench tests under controlled conditions.
Do you score products you have not bought?
Yes, and the card says so. A banner reading bought and tested by me means I own the unit; a banner reading my pick means the selection came from specifications, warranty terms and consistent owner reports instead. Most untested picks are whole-home systems costing several thousand dollars each. Every full review on the site is of a product I bought.
Why is nothing rated below 4?
Because the gear I buy is gear I already expect to be good, and I do not deliberately spend my own money on products I expect to be poor. That skews the sample upward. The information worth reading is the weak points section on each product and the gap between one score and another, not the absolute number.
Do affiliate commissions influence the ratings?
No. Commissions fund buying the equipment, and that is the extent of their role. Weak points are published regardless, products without affiliate programmes are covered on the same terms, and no brand has any input into a score or sees a review before publication.
How often are scores revisited?
When something material changes, such as a newer model arriving at the same price or a fault emerging with extended use. Scores are relative to a product class at a moment in time, so the Updated date on each page matters as much as the number itself.