A review from this desk should be usable by someone who has not decided whether they care about the subject. It should say what was used, what was done with it, and what the writer is actually claiming.
A few rules, stated now so later pieces can be held to them.
The thing being reviewed has to have been used. A summary of a spec sheet, a trailer, or a press quote is not a review. If the time spent was short, the piece should say so instead of borrowing authority from a longer test that did not happen.
Observation and preference stay distinguishable. A measurement and a dislike are both allowed. They are not the same sentence, and a review that mixes them without warning is harder to use than one that is openly opinionated.
Scores are optional. A score without a scale and a reason is a decoration. This publication will not publish a number just to put a badge on a card. Many pieces will not have a score at all. The sample article includes the score component so the template can be seen. The value there is not a rating.
Comparisons have to name the other thing and why it is the comparison. “Better than the competition” is not a comparison. If the other thing was not used, the piece should not pretend it was.
Corrections will be dated. If a piece is updated because something was wrong, or because the subject changed, the page will say so. A silent rewrite of a published claim is not a correction.
There is no plan to build the archive out of shopping links. A link to a maker or a store, if one appears in a future review, will be there so a reader can find the thing. It will be identifiable as a link. It will not be the point of the piece, and it will not be hidden inside a button that pretends to be a verdict.
These rules are a standard, not a claim that a long run of reviews has already met it. The archive does not contain that run yet.