How we test and rank

Every score on VettedTen comes out of our confidence-weighted ranking model. The exact weightings stay private, because the moment they're public they're gameable. The principles don't change from page to page, and here they are.

The evidence base

Our main input is verified owner reviews: thousands per product, covering the whole time it has been on sale. A week of hands-on testing tells you how a product feels out of the box. Only long-term owners can tell you whether the drawer runner snaps in month eleven. We weight their verdicts by:

• Volume. A 4.5★ average across 40,000 owners beats 5.0★ from a dozen. It isn't close.
• Recency. Recent reviews reflect current stock and firmware, so they count for more than reviews of an old revision.
• Reviewer reliability. Patterns that look like incentivised or fake reviewing drag a score down. Credible reliability concerns get a product flagged, or thrown out of the list entirely.

Confidence weighting

New products with thin review counts get pulled toward the category average until real evidence stacks up. Hype isn't evidence. A just-launched product with a perfect early rating won't leapfrog a proven best-seller here.

Value and badges

Each guide carries at most four badges: Best Choice, Best Premium, Value for Money and Best Budget. The model assigns them, not an editor with a favourite. A product we've flagged for reliability can still rank, but it never wears a badge.

What we don't do

• No payment, product loans or hospitality in exchange for placement. Ever.
• Editors never see affiliate earnings data while scores are set.
• We don't pull a product from a ranking because a brand asked. Nicely or otherwise.

Questions about the method? Email hello@vettedten.com.