We ran real SPF, DKIM, and DMARC DNS checks against the world's 1,000 most-visited domains. Here's what we found.
Email authentication adoption gets discussed constantly, but hard numbers on how widespread it actually is are rare. So we checked: using the same methodology as this site's own scanner (SPF at the domain apex, DMARC at _dmarc., DKIM probed across 14 common selector names), we scanned the Tranco list's top 1,000 most-visited domains.
Across 1,000 of the world's highest-traffic domains:
| Mechanism | Present | Share |
|---|---|---|
| SPF (any policy) | 760 / 1,000 | 76.0% |
| DMARC (any policy) | 725 / 1,000 | 72.5% |
| DKIM (common selector found) | 467 / 1,000 | 46.7% |
| All three present | 460 / 1,000 | 46.0% |
| None of the three detectable | 212 / 1,000 | 21.2% |
Of the domains that do have DMARC, a meaningful chunk aren't actually enforcing it: 110 domains (11.0% of the full sample) publish DMARC in monitor-only mode (p=none), which collects reports but does nothing to stop spoofed email claiming to be from that domain. Only 444 domains (44.4%) are at full p=reject enforcement.
A quarter of the internet's busiest sites publish no SPF record. Over a quarter publish no DMARC record. And one in five has none of the three mechanisms detectable at all — meaning anyone could craft a spoofed email appearing to come from that domain, and the receiving mail server would have no DNS-based way to catch it.
High traffic and email-security hygiene are not the same thing. Even well-resourced, high-profile domains leave this gap open.
Run the free scanner above to get your domain's real SPF, DKIM, DMARC, and MX status with an A–F grade and a fix plan.
The complete paper (abstract, methodology, limitations — including an honest note on why the DKIM figure is a floor, not a ceiling) and the raw 1,000-row dataset are published openly on Zenodo under CC-BY-4.0, with a permanent DOI:
Suggested citation: EmailSecurityGrade (2026). Email Authentication Adoption Among the World's Most-Visited Websites: A 2026 DNS-Based Snapshot. Zenodo. https://doi.org/10.5281/zenodo.23878889