Over the course of 2021 the number of valid DMARC policies observed in use rose by 84%, to a total of nearly 5 million unique records, compared to the prior calendar year. This doubles the percentage increase seen in 2020, and reflects an acceleration of growth in the second half of the year. Starting with June, the number of new records seen each month was between three and six times the number seen in the first five months of the year. These findings are based on analysis of DNS request data generously supplied by DomainTools.
The following graph shows the total valid records observed at six month intervals since the end of 2016.
The average number of new records each month was roughly 218,000 over the entire year, but looking at the first half versus the second half the averages are roughly 147,500 and 288,000. More graphs and statistics regarding DMARC are available on this page.
DMARC.org uses DNS transaction data supplied by DomainTools (formerly Farsight Security) that includes anonymized requests for DMARC records, and the corresponding responses, captured by their sensors at various strategic points on the Internet. DMARC.org weeds out invalid and malformed DMARC records, and makes sure the remaining records are still being served by looking them up in DNS.
Since this validation is performed in the month following the end of the most recent quarter in which the data was captured, the result is a reasonable point-in-time figure for the portion of the Internet covered by DomainTool’s sensors. It is unlikely to include all DMARC records published or in use on the Internet, but it is a stable subset that provides a useful picture of adoption and usage over time.
DMARC.org thanks DomainTools for making this data available for the purposes of research and education.