Methodology

How we test dog collars

We do not run a laboratory. What we run is a filter, and these are the five things it checks before anything reaches this catalog.

Our criteria

1. The measurements come from the maker, converted inward. We take the manufacturer's own neck table and convert it to inches, rounding toward the smaller number at every boundary. A dog measured at the top of a range is genuinely at the limit, so we send you up a size rather than let you discover it at home.
2. A color is listed only where it really exists. The collar comes in ten colors and the matching set in six. We could show ten on both and sort it out later; instead the selector hides what cannot be ordered, so nobody pays for a combination we cannot ship.
3. Every listed variant has a photo of that variant. Not a swatch, not a recolored render — a photograph of the item in that color, with its own lining visible. Every color in the selector has one.
4. Ratings are rebuilt from individual scores. We paginate the full set of ratings rather than reading a headline average, drop everything below four stars, and recompute the mean on what remains. The number you see is that computation, not a figure copied from somewhere else.
5. When two sources disagree, the less flattering one wins. This one costs us. Three of our five neck ranges were listed a half inch wider elsewhere than the maker's table allowed. We publish the maker's numbers. It makes the catalog look tighter than the competition and it keeps collars on dogs.

What happens after you order

The plate is engraved once your order is in, with the exact text you enter at checkout, which is why the collar takes longer to reach you than something pulled off a shelf. Every order carries a 30-day money-back guarantee, so the risk is on us, not you.