Methodology

How we score website trust

Every SiteReviewChecker report produces a trust score from 0 to 100, built from publicly available signals and weighted toward the factors that most reliably distinguish legitimate websites from scams. We use no paid third-party reputation APIs — every signal is independently verifiable.

The four scoring categories

Domain Reputation carries the most weight because domain age is the single strongest statistical predictor of fraud — most scam sites are only days or weeks old.

Domain Reputation

35 pts
  • Domain age (up to 18 pts) — via RDAP registration date
  • TLD reputation (8 pts) — established vs. abuse-prone extensions
  • Typosquatting (6 pts) — Levenshtein distance to major brands
  • Registrar transparency (3 pts)

Connection Security

25 pts
  • Valid HTTPS connection (10 pts)
  • Trusted, unexpired TLS certificate (6 pts)
  • HTTPS enforcement (4 pts)
  • HSTS and security headers (5 pts)

Email & DNS

20 pts
  • DNS resolution (5 pts)
  • MX records (6 pts)
  • SPF policy (3 pts)
  • DMARC policy (3 pts)
  • DKIM signing (3 pts)

Transparency & Content

20 pts
  • Live, substantive content (7 pts)
  • Contact information (4 pts)
  • Privacy, terms & about pages (6 pts)
  • Structured data & metadata (3 pts)

Verdict bands

ScoreVerdict
86–100 Highly Trusted
71–85 Probably Legitimate
51–70 Exercise Caution
31–50 Suspicious
0–30 Likely Scam

Safeguards against false signals

Scammers can trivially obtain free HTTPS certificates and publish realistic-looking content, so those signals alone never produce a high score. We apply hard ceilings: a domain less than 30 days old cannot exceed a "Suspicious" rating; a newly registered domain on an abuse-prone TLD is capped in the "Likely Scam" range; and a name that closely imitates a major brand (typosquatting) is capped regardless of its other signals.

Conversely, a domain that does not resolve in DNS and returns no HTTP response receives no score at all — it is reported as "Could Not Analyze" rather than being given a misleading number.

This report is automated and provided for educational purposes. It is not a professional security audit and does not guarantee the safety or legitimacy of any website.

Research & standards behind each check

Every signal we score is defined by a public standard or grounded in published fraud research, so any report can be reproduced with open tools.

Read more in our guides on how trust scores work and how AI systems rate websites, or start with the complete website safety guide.