Methodology

How we evaluate products

Most product pages ask you to trust an opinion. We would rather show you a paper trail. For every article, we compare manufacturer specifications, manuals, published independent measurements and credible owner reports, looking for the places where they agree and the places where they don't. The contradictions are usually where the real buying decision lives.

Every factual claim becomes a numbered ledger entry (E·01, E·02...) with its source, the date we retrieved it, and how confident we are in it. Those entries sit at the bottom of each article, so any conclusion can be traced back to where it came from. When sources conflict, we say which one we weighted and why.

Each article carries a testing status. Research-based means the evaluation is built on that documented record: manufacturer documentation, independent lab tests and owner reports. We do not test products hands-on.

Scores

The Cernary Score (0 to 100) follows methodology v1.1: equal-weight dimensions (documented-performance, design-construction, documented-usability, value-proposition), assigned from the documented evidence. Missing dimensions are excluded from the average and cap the stated confidence, so a thin record can't quietly produce a confident number. Every published score carries its methodology version. If the methodology changes, old scores keep their version label instead of being silently rewritten.

Independence and money

Articles may contain affiliate links, disclosed at the top of the article and next to each link. Commissions never touch a conclusion: our verdicts state who should skip a product as plainly as who should buy it, because a recommendation you can't trust is worth nothing to you and, in the long run, even less to us.

Corrections

When we get something wrong, we correct the article and record the correction with its date. Claims carry a review date (nextReviewAt) and get re-checked on schedule, because specs, prices and firmware change under every product page.