Frequently asked

No. A high score means the public evidence we could find is consistent with an established business. It is not a safety guarantee, a certification, or a promise that you will receive what you order. Established companies still have bad service, data breaches and disputes.

No, and this matters. A low score most often means we found little public information. A new bakery with one page and no Trustpilot profile scores low for exactly the same reasons a throwaway scam shop does. Read the individual signals rather than fixating on the number.

Usually one of: the domain is under a year old, no company or VAT number is published, there is no postal address or phone number, no terms or refund policy exists, or no social profiles are linked. Publishing those things helps real customers trust you too, not just our checker.

RDAP (the modern replacement for WHOIS) for registration data, a direct TLS handshake for the certificate, the site's own homepage for business details, Trustpilot's public review page, Reddit's public search, and the Internet Archive for history. Every source is linked on the report so you can verify it yourself.

Yes, but only the public report, not who checked it. Every completed check saves the URL, score, verdict, and gathered signals to our database and lists it in the directory. There are no accounts and no personal history. See the privacy policy for the full detail.

No. The score is computed deterministically from weighted signals, so identical evidence always yields the same number. The model only writes the plain language explanation of what was found, and it is instructed to reason strictly from the gathered evidence.

Common reasons: the registry restricts RDAP data (very common for country domains), the site blocks automated visitors, Trustpilot has no profile for that domain, or an upstream source timed out. Failed checks are listed on the report rather than silently ignored, and they lower the stated confidence.

The report is generated from public sources and saved in the directory. If a source we read is wrong, such as an outdated registry record or a mistaken Reddit thread, correct it at that source and run the check again to update the saved report. Nothing here is a statement of fact about any business.

Yes, within reason. There is an IP based rate limit of a few checks per minute so upstream sources are not hammered. If you hit it, wait a moment and try again.