edition stringdate 2026-08-01 00:00:00 2026-09-01 00:00:00 | tranco_list_id stringclasses 2
values | metric stringlengths 9 30 | count int64 343 1M | denominator int64 469k 1M | denominator_label stringclasses 4
values | share_percent float64 0.03 100 |
|---|---|---|---|---|---|---|
2026-08 | JZ8PY | domains_scanned | 1,000,766 | 1,000,766 | domains scanned | 100 |
2026-08 | JZ8PY | domains_measured | 1,000,296 | 1,000,766 | domains scanned | 99.95 |
2026-08 | JZ8PY | dmarc_lookup_failed | 470 | 1,000,766 | domains scanned | 0.05 |
2026-08 | JZ8PY | dmarc_record_published | 474,544 | 1,000,296 | measured domains | 47.44 |
2026-08 | JZ8PY | dmarc_record_valid | 469,415 | 1,000,296 | measured domains | 46.93 |
2026-08 | JZ8PY | policy_none | 234,444 | 1,000,296 | measured domains | 23.44 |
2026-08 | JZ8PY | policy_quarantine | 116,806 | 1,000,296 | measured domains | 11.68 |
2026-08 | JZ8PY | policy_reject | 118,165 | 1,000,296 | measured domains | 11.81 |
2026-08 | JZ8PY | tier_none | 525,752 | 1,000,296 | measured domains | 52.56 |
2026-08 | JZ8PY | tier_broken | 5,129 | 1,000,296 | measured domains | 0.51 |
2026-08 | JZ8PY | tier_dead | 128,980 | 1,000,296 | measured domains | 12.89 |
2026-08 | JZ8PY | tier_monitoring | 105,464 | 1,000,296 | measured domains | 10.54 |
2026-08 | JZ8PY | tier_partial | 27,320 | 1,000,296 | measured domains | 2.73 |
2026-08 | JZ8PY | tier_enforced | 207,651 | 1,000,296 | measured domains | 20.76 |
2026-08 | JZ8PY | tier_dead_of_valid_records | 128,980 | 469,415 | valid DMARC records | 27.48 |
2026-08 | JZ8PY | spf_record_published | 639,844 | 1,000,296 | measured domains | 63.97 |
2026-08 | JZ8PY | spf_permerror | 36,735 | 639,844 | published SPF records | 5.74 |
2026-08 | JZ8PY | spf_permerror_too_many_lookups | 16,591 | 639,844 | published SPF records | 2.59 |
2026-08 | JZ8PY | spf_multiple_records | 10,474 | 639,844 | published SPF records | 1.64 |
2026-08 | JZ8PY | mta_sts_published | 9,520 | 1,000,296 | measured domains | 0.95 |
2026-08 | JZ8PY | tls_rpt_published | 11,442 | 1,000,296 | measured domains | 1.14 |
2026-08 | JZ8PY | bimi_published | 17,339 | 1,000,296 | measured domains | 1.73 |
2026-09 | GQJJK | domains_scanned | 1,000,767 | 1,000,767 | domains scanned | 100 |
2026-09 | GQJJK | domains_measured | 1,000,424 | 1,000,767 | domains scanned | 99.97 |
2026-09 | GQJJK | dmarc_lookup_failed | 343 | 1,000,767 | domains scanned | 0.03 |
2026-09 | GQJJK | dmarc_record_published | 477,782 | 1,000,424 | measured domains | 47.76 |
2026-09 | GQJJK | dmarc_record_valid | 472,655 | 1,000,424 | measured domains | 47.25 |
2026-09 | GQJJK | policy_none | 237,064 | 1,000,424 | measured domains | 23.7 |
2026-09 | GQJJK | policy_quarantine | 117,055 | 1,000,424 | measured domains | 11.7 |
2026-09 | GQJJK | policy_reject | 118,536 | 1,000,424 | measured domains | 11.85 |
2026-09 | GQJJK | tier_none | 522,642 | 1,000,424 | measured domains | 52.24 |
2026-09 | GQJJK | tier_broken | 5,127 | 1,000,424 | measured domains | 0.51 |
2026-09 | GQJJK | tier_dead | 131,008 | 1,000,424 | measured domains | 13.1 |
2026-09 | GQJJK | tier_monitoring | 106,056 | 1,000,424 | measured domains | 10.6 |
2026-09 | GQJJK | tier_partial | 27,237 | 1,000,424 | measured domains | 2.72 |
2026-09 | GQJJK | tier_enforced | 208,354 | 1,000,424 | measured domains | 20.83 |
2026-09 | GQJJK | tier_dead_of_valid_records | 131,008 | 472,655 | valid DMARC records | 27.72 |
2026-09 | GQJJK | spf_record_published | 641,215 | 1,000,424 | measured domains | 64.09 |
2026-09 | GQJJK | spf_permerror | 36,598 | 641,215 | published SPF records | 5.71 |
2026-09 | GQJJK | spf_permerror_too_many_lookups | 16,482 | 641,215 | published SPF records | 2.57 |
2026-09 | GQJJK | spf_multiple_records | 10,448 | 641,215 | published SPF records | 1.63 |
2026-09 | GQJJK | mta_sts_published | 9,563 | 1,000,424 | measured domains | 0.96 |
2026-09 | GQJJK | tls_rpt_published | 11,542 | 1,000,424 | measured domains | 1.15 |
2026-09 | GQJJK | bimi_published | 17,428 | 1,000,424 | measured domains | 1.74 |
DMARC Census
A monthly DNS measurement of DMARC and SPF across the Tranco top 1 million domains, plus a cohort of US federal domains. Each edition is one scan. This repository holds the aggregate results of every edition.
The report for each edition, with its methodology, is at https://domaincanary.com/research/dmarc-census.
What we measured
We looked up the DMARC and SPF records of every domain on one pinned Tranco list, once a month. The September 2026 edition scanned 1,000,767 domains on 1 September 2026. 1,000,424 of them answered the DMARC lookup, and we call those the measured domains. 343 lookups failed after retries. Those domains could not be measured and we left them out. We never counted them as publishing no record.
In September 2026, 47.2% of measured domains published a valid DMARC record. 20.8% of measured domains published quarantine or reject with no pct or sp tag weakening it. 131,008 domains published p=none with no reporting address that could receive a report, which is 13.1% of measured domains and 27.7% of valid DMARC records.
Files
| File | Contents |
|---|---|
headline.csv |
The headline counts for every edition, one row per metric. Each row has its count, its denominator and what the denominator is. |
<edition>/aggregates.json |
Every count and share published in that edition: adoption, policy, record quality, reporting addresses, SPF, mail providers and report vendors, by country, by TLD and for the US federal cohort. |
<edition>/manifest.json |
What you need to repeat the scan: the Tranco list id, the Public Suffix List hash, the resolvers, the scanner version and the scan times. |
SHA256SUMS |
Checksums of the data files. |
CITATION.cff |
How to cite the dataset. |
The data is aggregate. It has counts and shares, and no record of how any single domain was configured.
Reading a share
A share in this dataset never stands alone. headline.csv gives the denominator for every row, because the same count is a different percentage of measured domains, of valid DMARC records and of published SPF records. Quote the share with its denominator and its edition.
One month of change is one data point. Draw a trend only across three or more editions.
Editions
| Edition | Tranco list | Scan date | Domains scanned |
|---|---|---|---|
| 2026-09 | GQJJK | 2026-09-01 | 1,000,767 |
| 2026-08 | JZ8PY | 2026-08-20 | 1,000,766 |
Source and citation
The population is the Tranco list: Victor Le Pochat, Tom Van Goethem, Samaneh Tajalizadehkhoob, Maciej Korczyński and Wouter Joosen, "Tranco: A Research-Oriented Top Sites Ranking Hardened Against Manipulation", NDSS 2019, https://doi.org/10.14722/ndss.2019.23386.
Editions from September 2026 are also archived on Zenodo. The series DOI is https://doi.org/10.5281/zenodo.22849900 and it resolves to the newest edition.
Suggested citation: Purvis, J. DMARC Census. DomainCanary. https://doi.org/10.5281/zenodo.22849900
License
CC BY 4.0. Reuse the data, including commercially, with attribution.
- Downloads last month
- 74