Growth in DMARC policy records from 2020 through 2025
It has been several years since the statistics for DMARC policy records had been updated on this page. The main reason for this was surges of up to millions of records each quarter, if not each month. They would show up, demanding analysis and explanation, and then in many cases most of them would diappear over the following quarters. We’ve described this situation in some of our presentations, like the last nine slides of this deck.
In some cases a domain registrar or large domain owner/operator might publish policies for all of their domains, which is great news. In other cases, a large operator might have published a wildcard record that applied to tens of thousands of seemingly-random DNS labels, like routers and network switches. And in some cases, we’ve seen what looked like algorithmically generated DNS labels (and wondered what they were being used for). While the latter might be obvious when seen individually, they can be harder to identify in the nearly 4 billion records of recursive DNS lookups that DomainTools has generously shared to support this work.
For the time being, the DMARC statistics on this page still reflect every validly formatted DMARC policy record captured by DomainTools over the past 15 years or so. Hopefully we’ll be able to resume regular updates as we refactor our tools, and update the pages on BIMI and DKIM as well.