How we choose
How we choose, and what we refuse to claim
We have no lab, no sample closet and no credential. This page sets out exactly what we do instead, so you can decide how much weight to give it.

What we do not do
The competing sites in this category say they tested. Some of them did. We have not, and we will not say otherwise, so it is worth being explicit about the whole list of things we do not claim:
- We have not used, held, or owned any product on this site.
- We do not run a test lab, a lux meter or a scale.
- We have not received samples from any brand and have not asked for any.
- We hold no professional credential of any kind — not in lighting, not in bookbinding, not in anything clinical.
- We do not publish first-party ratings, and no product page emits review schema.
That is a smaller claim than the category norm. It is also the one that stays true, and it is why every figure on this site is attributable rather than asserted.
What we do instead
1. A fixed checklist per category
Every category has a checklist of specifications, decided before any product was read and applied identically to everything in it. Book lights are checked on runtime, color temperatures, brightness levels, charging port, clip or wear style, weight and warranty. Reading journals on capacity, trim size, cover material, binding, paper weight, whether it is prompted, and extras. And so on for each.
Fixing the checklist first is what stops the score being a description of whichever seller wrote the best listing. It also means adding an item to a checklist changes every score in that category, which is why we do not edit them casually.
2. Every figure carries its source
Each spec cell shows where the number came from: the seller’s product title, a feature bullet on the listing, the manufacturer’s own page, or a primary source such as a standards or government body. Every one is stamped with the date the listing was read, which is what makes it checkable rather than an argument.
Across this site, 112 of 406 spec cells currently read “Not published”. That is not a gap in our notes. It is the finding.
3. A score with one computed part and four judged ones
Every product carries a gift-fit score out of 10, made of five components worth 20 points each. One of them is computed and four are our judgment, and every product page labels which is which rather than presenting all five as measurement.
The four judged components are opinions held to a published definition. They are not measurements and we do not present them as any. A score you cannot take apart is just an opinion with a decimal point, which is why the breakdown is printed on every product write-up rather than tucked behind a link.
4. Every pick gets a “skip it if”
All 67 products on this site carry a sentence about who should not buy it, including the ones we rank first. 2 carry a stronger statement — a recommendation against, with the reason. A guide where everything is wonderful is a guide nobody believes.
How prices work here
Prices render from one place only: a live retailer data feed with a 48-hour freshness limit. If the feed has not run inside that window, no number renders anywhere on the site and every button reads “Check price”.
We will not type a price from memory, from a screenshot or from a previous visit. A stale price on a gift guide is the single error a reader acts on without checking, and the honest alternative — showing nothing — costs us clicks and costs you nothing. How we are paid covers the commercial side of that.
Corrections
Listings change. A seller who published a wattage last month may not publish it next month, and a figure we quoted may become wrong through no fault of anyone’s. That is why every figure carries the date it was read.
If you find something wrong here, tell us. We will correct it, and where the correction is material we will say on the page that it was corrected and when. Our editorial policy sets out how that works, along with our independence rules.
Why this method rather than a better one
Because it is the best method available to a publisher that does not have a lab, and because the alternative is pretending. A site that says “after weeks of testing” without having tested anything is making the one claim a reader cannot check and a competitor cannot falsify.
What we can do is read every listing in a category with the same checklist, hold every seller to the same standard, show the arithmetic, and be specific about what is missing. That turns out to separate the products in this category more sharply than you would expect — because the sellers who publish the most are, reliably, the ones with the most to publish.