How We Research
Most pet gear sites imply a testing lab they do not have. We would rather tell you what we actually do, so you can judge how much weight to give a ranking.
What goes into a ranking
Specification comparison. Dimensions and weight limits, materials and their safety, fill and foam density, mesh and fabric grades, wattage and thermostat behaviour, filtration rates, battery life, cleaning method. Specifications do not decide a ranking on their own, but they rule out products that cannot do what a listing claims — a bed that is too small for the dog it is sold for, or a heat lamp with no thermostat.
Owner reports at scale. We read verified-purchase reviews in volume, and we weight the ones written months or years after purchase far above the ones written in week one. Durability, chew resistance, service life and failure modes only show up in long-term reports, and that is where most of our reliability judgements come from.
Veterinary and expert guidance. Where a category carries a genuine animal-welfare question — orthopedic support, calming products, reptile heating and UVB, harness fit, dental chews — we follow published veterinary guidance and expert testing rather than paraphrasing it as our own, and we let it override a spec sheet where the two disagree.
Availability verification. Every product is checked as live and in stock before publication, and rechecked on a schedule afterwards. Nothing is worse than a recommendation you cannot buy.
What the score means
Every product carries a whole number out of ten. There is no decimal place, and that is deliberate: the research above can place a product in a tier, but it cannot separate a 9.6 from a 9.2, and a site that prints tenths is claiming a precision it does not have. Two products that share a score are products we could not separate on the evidence. Where we canseparate them, the ranking says so — the rank is the judgement, and the score is the tier.
It is also the only score on the page. We used to print a five-star rating beside the ten-point one, which was the same number twice and implied two independent assessments where there was one.
What this method cannot tell you
Our current rankings are research-based unless a specific article explicitly documents hands-on testing. That means we do not claim how a specific harness feels on a particular dog's chest, or settle a close contest on feel alone. Where two products are genuinely too close to separate on the evidence we have, we say so rather than inventing a winner.
We also do not rank gear that is unavailable through the retailers we link to. Some excellent products are simply not sold in places we can point you at, and their absence from a list is not a judgement on them.
Money
No brand pays for placement, and no brand sees an article before it publishes. The site earns affiliate commission on purchases made through our links, at no extra cost to you — the full terms are on ouraffiliate disclosure page. Commission rates vary between products, and we do not rank on rate. If we did, the cheapest item on every page would somehow never win, and you would notice.