DMARC Census, November 2026
We looked up the DMARC and SPF records of 1,000,748 domains. For each one, we checked whether it publishes a policy receivers act on and whether its DMARC reports have somewhere to go.
In the November 2026 scan, 54.2% of domains with a valid p=none DMARC record (129,077 of 238,355) publish it with no reporting address that receivers following the DMARC standard send reports to. p=none asks receivers not to act on mail that fails DMARC, so reports are how the owner sees who is sending as the domain.
The November 2026 edition adds reports that go nowhere by country for 38 countries. The percentage of valid DMARC records that are p=none with reports that go nowhere runs from 13.1% under .dk (Denmark) to 43.1% under .kr (South Korea). Only p=none records are counted, so a country's percentage also depends on how many of its domains publish quarantine or reject. Under .dk, 70.8% of valid records are at quarantine or reject, and under .kr, 19.7% are.
We scanned the Tranco top 1 million (list 7462X) and 1,321 US federal domains on 1 Nov 2026. 573 of the federal domains are already in the million, so the total is 1,000,748 domains.
Most percentages on this page are out of the 1,000,438 domains whose DMARC lookup got an answer. We call these the measured domains. Where we use a different group, we say which. We left out the 310 lookups that failed after retries, because a failed lookup can't show whether a record exists.
Jump to key findings, what changed, tiers, rank, countries, reports that go nowhere by country, TLDs, reports that go nowhere, SPF, the federal cohort, collectors, methodology or the data.
Key findings
- Of the top 1,000 domains in the November 2026 scan, 72.3% publish a valid DMARC record and 56.3% are in the enforced tier. Across ranks 100,000 to 1,000,000, those shares are 46.7% and 19.9%.
- 377 of the domains with a usable SPF record in the November 2026 scan end it with +all. That lets any host on the internet send as them.
- Of 1,321 registered US federal .gov domains, 80.2% publish p=reject in the November 2026 scan and 182 publish no DMARC record. That list is wider than the mail-sending executive-branch set that Binding Operational Directive 18-01 has held to p=reject since October 2018, so the gap isn't a count of agencies breaking the directive.
- Between the October 2026 and November 2026 scans, 9,297 domains moved to a stronger policy, 3,926 of them by publishing a first p=none record, and 1,987 moved to a weaker one. 98.7% of the 853,046 domains we measured in both months kept the same policy.
What changed since October 2026
We compare the policy of every domain that was on the list in both months and whose lookup got an answer both times. From weakest to strongest, the policies are no usable record, p=none, p=quarantine and p=reject. A broken record counts as no usable record. If DNS failed in either month, we leave the domain out and don't count it as a change.
853,046 domains were on both lists and got an answer both times. 841,762 of them kept the same policy from October 2026 to November 2026. 9,297 moved to a stronger policy and 1,987 to a weaker one. The stronger count includes 3,926 domains that published their first p=none record. 4,695 reached quarantine or reject from p=none or no usable record, and 2,730 of those had been at p=none. The most common change was no usable record to p=none, on 3,926 domains.
The headline shares, month against month
| Share | October 2026 | November 2026 | Change |
|---|---|---|---|
| Valid DMARC record | 47.1% | 48.0% | +0.9 |
| Enforced tier | 20.7% | 21.4% | +0.7 |
| p=reject | 11.5% | 11.9% | +0.4 |
| p=none and reports go nowhere, as a share of valid records | 27.5% | 26.9% | -0.6 |
| No DMARC record | 52.4% | 51.5% | -0.9 |
| SPF record published | 63.6% | 63.7% | +0.1 |
| SPF permerror, as a share of SPF records | 5.8% | 5.7% | -0.1 |
Change is the November 2026 percentage minus the October 2026 percentage.
The Tranco list changes from month to month. 147,280 domains in the November 2026 scan weren't on the October 2026 list, so part of every change in this table comes from that turnover. The changes below count only domains that were on both lists, so start with those.
Every policy change, largest first
| From | To | Domains |
|---|---|---|
| no usable record | p=none | 3,926 |
| p=none | p=quarantine | 1,947 |
| no usable record | p=reject | 1,097 |
| no usable record | p=quarantine | 868 |
| p=none | no usable record | 801 |
| p=none | p=reject | 783 |
| p=quarantine | p=reject | 676 |
| p=quarantine | no usable record | 367 |
| p=reject | no usable record | 299 |
| p=quarantine | p=none | 260 |
| p=reject | p=none | 154 |
| p=reject | p=quarantine | 106 |
147,280 domains in the November 2026 scan weren't on the October 2026 list, so there's nothing earlier to compare them with. 422 pairs were on both lists but failed to resolve in one month or the other. They're left out of every count here.
This table compares one month with the month before, so it shows one month of movement.
The quality tiers
We put each domain in one tier, based on whether receivers act on its policy and whether its DMARC reports have somewhere to go.
| Tier | Domains | Share |
|---|---|---|
| No record | 515,371 | 51.5% |
| Broken | 4,694 | 0.5% |
| Reports go nowhere | 129,077 | 12.9% |
| Monitoring | 109,278 | 10.9% |
| Partial enforcement | 27,740 | 2.8% |
| Enforced | 214,278 | 21.4% |
- No record 515,371 domains (51.5%) No DMARC record at the _dmarc name.
- Broken 4,694 domains (0.5%) A record exists and receivers ignore it: more than one record, a bad v= tag, or no usable p=.
- Reports go nowhere 129,077 domains (12.9%) p=none, with no rua address that receivers following the DMARC standard send reports to.
- Monitoring 109,278 domains (10.9%) p=none, with at least one rua address that receivers following the DMARC standard send reports to, or one the scan could not rule out.
- Partial enforcement 27,740 domains (2.8%) quarantine or reject, weakened by pct below 100 or a softer sp=.
- Enforced 214,278 domains (21.4%) quarantine or reject, with no pct or sp tag weakening it.
Two tags weaken a policy without changing it. pct= applies the policy to only that share of failing mail, and sp= sets a separate, often softer, policy for subdomains.
4,694 domains publish a broken record that receivers can't use, including 4,034 with more than one DMARC record. RFC 7489 tells receivers to treat that as no record at all, so we don't count broken records as valid.
Of the 480,373 valid records, 49.6% set p=none, 25.5% set p=quarantine and 24.8% set p=reject.
48.5% of measured domains publish something at the _dmarc name, but that count mixes every tier, broken records included. 214,278 domains (21.4%) reach the enforced tier. 515,371 (51.5%) publish no record. Most of the rest publish p=none, 129,077 with reports that go nowhere and 109,278 in monitoring.
Among domains at quarantine or reject, 10,487 apply the policy to only part of their failing mail with pct below 100, and 18,932 leave subdomains weaker than the organizational domain. Some domains do both, so the two figures add up to more than the 27,740 domains in the partial-enforcement tier.
Percentages are out of the 1,000,438 domains whose DMARC lookup got an answer. We left out the 310 domains whose lookup failed after retries, because a failed lookup can't show whether a record exists.
Valid records and enforcement by rank
The top 1,000 domains publish and enforce at far higher rates than the rest of the million, so don't quote their percentages for the whole list. Both bars are percentages of the domains in that rank range whose DMARC lookup got an answer.
| Rank range | Measured | Valid DMARC | Enforced tier |
|---|---|---|---|
| Top 1k | 1,000 | 72.3% | 56.3% |
| 1k to 10k | 9,000 | 66.6% | 44.5% |
| 10k to 100k | 89,983 | 58.9% | 33.2% |
| 100k to 1M | 899,707 | 46.7% | 19.9% |
Fewer domains publish a record the further down the list you go, and the enforced share falls faster than the valid-record share.
The rank ranges cover the Tranco million only. The 748 federal domains outside the million count toward the overall percentages but not toward any rank range.
Adoption by country
The map shades each country by the percentage of measured domains under its ccTLD that publish a valid DMARC record. Hover over a country to see its numbers, or find it in the table below.
- under 30%
- 30 to 39%
- 40 to 49%
- 50 to 59%
- 60% and up
- under 200 domains
| Country | Measured | Valid DMARC | At enforcement |
|---|---|---|---|
| Norway (.no) | 1,893 | 79.0% | 50.0% |
| Anguilla (.ai, registered worldwide) | 3,823 | 75.5% | 42.4% |
| Denmark (.dk) | 2,350 | 75.2% | 53.2% |
| Australia (.au) | 7,276 | 74.2% | 43.2% |
| Estonia (.ee) | 844 | 73.8% | 43.7% |
| Saudi Arabia (.sa) | 569 | 73.8% | 57.6% |
| Croatia (.hr) | 1,138 | 73.7% | 29.4% |
| Ireland (.ie) | 1,351 | 73.6% | 37.6% |
| Slovenia (.si) | 823 | 73.0% | 32.9% |
| Paraguay (.py) | 297 | 71.7% | 44.1% |
| Romania (.ro) | 4,348 | 71.0% | 28.6% |
| New Zealand (.nz) | 1,530 | 70.7% | 40.2% |
| Belgium (.be) | 3,357 | 70.4% | 39.1% |
| Singapore (.sg) | 832 | 70.1% | 47.7% |
| Algeria (.dz) | 217 | 70.0% | 45.6% |
| Luxembourg (.lu) | 249 | 69.5% | 45.0% |
| Uruguay (.uy) | 478 | 68.6% | 34.5% |
| Argentina (.ar) | 3,220 | 68.6% | 26.0% |
| Lithuania (.lt) | 1,132 | 68.4% | 31.3% |
| Israel (.il) | 1,518 | 68.1% | 23.1% |
| Finland (.fi) | 2,245 | 68.1% | 31.2% |
| United Arab Emirates (.ae) | 736 | 67.8% | 42.3% |
| Ecuador (.ec) | 506 | 67.6% | 46.2% |
| Portugal (.pt) | 2,104 | 67.5% | 36.6% |
| Bulgaria (.bg) | 1,279 | 67.5% | 32.1% |
| Tanzania (.tz) | 232 | 67.2% | 42.2% |
| Bosnia and Herzegovina (.ba) | 302 | 67.2% | 39.4% |
| Slovakia (.sk) | 2,084 | 66.9% | 29.8% |
| South Africa (.za) | 2,644 | 66.7% | 37.3% |
| Greece (.gr) | 3,504 | 66.6% | 26.2% |
| Poland (.pl) | 9,829 | 66.0% | 32.7% |
| Chile (.cl) | 2,469 | 65.9% | 35.3% |
| Costa Rica (.cr) | 234 | 65.0% | 47.4% |
| Switzerland (.ch) | 3,900 | 64.8% | 37.1% |
| Brazil (.br) | 15,965 | 64.5% | 28.6% |
| Spain (.es) | 5,618 | 64.5% | 33.6% |
| Kenya (.ke) | 478 | 64.4% | 33.9% |
| France (.fr) | 10,825 | 64.4% | 29.1% |
| Turkey (.tr) | 2,805 | 64.3% | 38.0% |
| Moldova (.md) | 408 | 64.0% | 32.4% |
| Latvia (.lv) | 897 | 63.5% | 40.8% |
| Sweden (.se) | 3,806 | 63.4% | 35.6% |
| Dominican Republic (.do) | 273 | 63.4% | 36.3% |
| Czechia (.cz) | 5,451 | 63.1% | 26.4% |
| Austria (.at) | 3,093 | 62.9% | 32.1% |
| Hungary (.hu) | 3,037 | 62.6% | 21.6% |
| Guatemala (.gt) | 228 | 62.3% | 39.5% |
| Germany (.de) | 30,112 | 62.2% | 32.4% |
| North Macedonia (.mk) | 282 | 61.3% | 30.5% |
| Bangladesh (.bd) | 624 | 60.9% | 39.6% |
| Peru (.pe) | 1,177 | 60.7% | 34.0% |
| Serbia (.rs) | 1,042 | 60.4% | 22.5% |
| Iceland (.is) | 433 | 58.9% | 32.3% |
| Hong Kong (.hk) | 869 | 58.6% | 35.2% |
| Canada (.ca) | 6,324 | 58.3% | 32.0% |
| Armenia (.am) | 355 | 57.7% | 37.2% |
| Mexico (.mx) | 3,159 | 57.6% | 30.6% |
| Thailand (.th) | 943 | 57.6% | 30.1% |
| Venezuela (.ve) | 256 | 57.4% | 38.7% |
| Japan (.jp) | 16,697 | 57.4% | 13.4% |
| Malaysia (.my) | 1,551 | 56.7% | 36.1% |
| Egypt (.eg) | 332 | 55.7% | 34.6% |
| Niue (.nu, registered worldwide) | 277 | 54.9% | 25.6% |
| Indonesia (.id) | 4,794 | 54.5% | 25.9% |
| Colombia (.co, registered worldwide) | 6,895 | 54.3% | 28.5% |
| Micronesia (.fm, registered worldwide) | 342 | 53.8% | 23.7% |
| European Union (.eu, registered worldwide) | 5,966 | 53.7% | 26.6% |
| Philippines (.ph) | 1,206 | 53.6% | 30.0% |
| British Indian Ocean Territory (.io, registered worldwide) | 10,395 | 53.6% | 31.5% |
| Guernsey (.gg, registered worldwide) | 818 | 52.7% | 26.7% |
| Nigeria (.ng) | 655 | 52.4% | 17.4% |
| Pakistan (.pk) | 1,594 | 52.3% | 19.8% |
| Uzbekistan (.uz) | 639 | 52.0% | 29.0% |
| Morocco (.ma) | 502 | 51.6% | 26.1% |
| Georgia (.ge) | 443 | 51.5% | 23.3% |
| India (.in) | 10,948 | 50.3% | 26.7% |
| Belarus (.by) | 1,250 | 49.1% | 26.6% |
| Ukraine (.ua) | 5,004 | 47.5% | 20.2% |
| Saint Helena (.sh, registered worldwide) | 293 | 46.8% | 30.0% |
| Kyrgyzstan (.kg) | 282 | 45.7% | 31.6% |
| Sri Lanka (.lk) | 486 | 45.1% | 20.8% |
| Nepal (.np) | 247 | 43.7% | 26.3% |
| Italy (.it) | 15,126 | 43.7% | 15.8% |
| Taiwan (.tw) | 2,696 | 43.3% | 20.0% |
| Azerbaijan (.az) | 543 | 42.9% | 27.6% |
| Tunisia (.tn) | 285 | 42.8% | 20.7% |
| Netherlands (.nl) | 17,812 | 42.4% | 23.0% |
| Kazakhstan (.kz) | 1,405 | 41.6% | 23.2% |
| Russia (.ru and .рф) | 34,922 | 41.1% | 21.6% |
| Iran (.ir) | 3,401 | 40.9% | 23.9% |
| Isle of Man (.im, registered worldwide) | 311 | 40.8% | 20.6% |
| United Kingdom (.uk) | 29,526 | 36.9% | 20.1% |
| Vietnam (.vn) | 2,746 | 36.3% | 18.1% |
| Montenegro (.me, registered worldwide) | 2,892 | 35.7% | 19.3% |
| Soviet Union (.su, registered worldwide) | 946 | 35.6% | 18.0% |
| Tuvalu (.tv, registered worldwide) | 2,652 | 35.5% | 16.9% |
| Tonga (.to, registered worldwide) | 852 | 28.1% | 14.0% |
| Laos (.la) | 228 | 27.2% | 13.2% |
| South Korea (.kr) | 4,385 | 25.7% | 5.1% |
| Samoa (.ws, registered worldwide) | 334 | 25.1% | 14.7% |
| Sao Tome and Principe (.st, registered worldwide) | 276 | 21.7% | 14.5% |
| United States (.us) | 5,451 | 21.6% | 12.7% |
| Palau (.pw, registered worldwide) | 296 | 19.9% | 9.1% |
| China (.cn) | 11,365 | 14.5% | 8.0% |
| Cocos Islands (.cc, registered worldwide) | 6,884 | 8.8% | 4.4% |
Show all 105 countriesShow fewer
105 countries have at least 200 measured domains under their ccTLD in this snapshot, and 88 of them are shaded on the map. Of the shaded countries, Norway is highest, with 79.0% of measured .no domains publishing a valid DMARC record. China is the lowest shaded country, at 14.5% of measured .cn domains.
A domain counts toward a country by its TLD alone. Every .de domain counts as Germany, wherever it's hosted or owned. Domains on gTLDs like .com have no country in their name, so they're left off the map. The TLD table below covers them. ccTLDs registered mostly from outside their country, .io, .ai, .co, .tv and .me among them, are measured and listed in the table but not shaded, because shading them would color the wrong country. Countries with fewer than 200 measured domains stay uncolored. A few too small to draw at this size, Singapore and Hong Kong among them, are marked with a dot in the same colors.
Reports that go nowhere, by country
Each row is one country, counted by its domain ending (ccTLD), such as .de for Germany. We list countries with at least 2,000 measured domains, lowest percentage first. Share is the percentage of the Valid records column, which counts domains with a valid DMARC record and can be lower than 2,000.
Reports go nowhere counts domains that publish p=none with no reporting address, or with only outside addresses whose owners haven't authorized the reports. Receivers that follow the DMARC standard don't send those domains reports. p=none asks receivers not to act on mail that fails DMARC, so reports are how the owner sees who is sending as the domain.
| Country | Valid records | Reports go nowhere | Share |
|---|---|---|---|
| Denmark (.dk) | 1,768 | 231 | 13.1% |
| Sweden (.se) | 2,414 | 416 | 17.2% |
| Portugal (.pt) | 1,420 | 263 | 18.5% |
| China (.cn) | 1,648 | 322 | 19.5% |
| Turkey (.tr) | 1,803 | 353 | 19.6% |
| United States (.us) | 1,179 | 235 | 19.9% |
| Australia (.au) | 5,400 | 1,092 | 20.2% |
| Vietnam (.vn) | 998 | 205 | 20.5% |
| Switzerland (.ch) | 2,526 | 524 | 20.7% |
| Belgium (.be) | 2,362 | 500 | 21.2% |
| South Africa (.za) | 1,764 | 375 | 21.3% |
| Canada (.ca) | 3,685 | 799 | 21.7% |
| Chile (.cl) | 1,626 | 364 | 22.4% |
| France (.fr) | 6,972 | 1,566 | 22.5% |
| Czechia (.cz) | 3,439 | 804 | 23.4% |
| Finland (.fi) | 1,529 | 378 | 24.7% |
| United Kingdom (.uk) | 10,907 | 2,689 | 24.7% |
| Indonesia (.id) | 2,614 | 646 | 24.7% |
| Austria (.at) | 1,946 | 482 | 24.8% |
| Taiwan (.tw) | 1,167 | 289 | 24.8% |
| Germany (.de) | 18,716 | 4,669 | 24.9% |
| Netherlands (.nl) | 7,549 | 1,928 | 25.5% |
| Mexico (.mx) | 1,820 | 466 | 25.6% |
| Slovakia (.sk) | 1,395 | 390 | 28.0% |
| Brazil (.br) | 10,299 | 2,905 | 28.2% |
| Spain (.es) | 3,621 | 1,025 | 28.3% |
| Iran (.ir) | 1,391 | 395 | 28.4% |
| Poland (.pl) | 6,492 | 1,878 | 28.9% |
| India (.in) | 5,505 | 1,611 | 29.3% |
| Russia (.ru and .рф) | 14,349 | 4,304 | 30.0% |
| Ukraine (.ua) | 2,375 | 763 | 32.1% |
| Romania (.ro) | 3,087 | 1,110 | 36.0% |
| Argentina (.ar) | 2,208 | 828 | 37.5% |
| Italy (.it) | 6,603 | 2,512 | 38.0% |
| Greece (.gr) | 2,335 | 935 | 40.0% |
| Hungary (.hu) | 1,900 | 780 | 41.1% |
| Japan (.jp) | 9,579 | 4,067 | 42.5% |
| South Korea (.kr) | 1,129 | 487 | 43.1% |
We list 38 countries in the November 2026 table, each with at least 2,000 measured domains under its ccTLD. We leave out 7 ccTLDs that clear the same bar in the November 2026 scan, such as .co and .io, because these endings don't show which country a domain belongs to. In November 2026, 13.1% of .dk domains with a valid DMARC record, the lowest percentage of any country listed, publish p=none with no reporting address that receivers following the DMARC standard send reports to. In November 2026, 43.1% of .kr domains with a valid DMARC record, the highest percentage of any country listed, publish p=none with no reporting address that receivers following the DMARC standard send reports to. Across the whole November 2026 scan, 26.9% of domains with a valid DMARC record publish p=none with no reporting address that receivers following the DMARC standard send reports to.
The Reports go nowhere column counts only p=none records, so a country where few domains publish quarantine or reject has more records that can be counted. In November 2026, 19.7% of .kr domains with a valid DMARC record publish quarantine or reject, and 70.8% of .dk domains do. Among .kr domains that publish p=none in November 2026, 53.7% publish it with no reporting address that receivers following the DMARC standard send reports to. Across the whole November 2026 scan, 54.2% of domains that publish p=none publish it with no reporting address that receivers following the DMARC standard send reports to. Among the listed countries with at least 500 p=none records, counting only domains that publish p=none, the November 2026 percentage with reports that go nowhere is highest for .ir domains (68.2%) and lowest for .se domains (39.3%).
A domain counts toward a country by its TLD alone, the same way as on the map. This table needs 2,000 measured domains per country, a higher bar than the map's. Each percentage is out of that country's valid DMARC records, which are fewer. Every percentage in this table is out of at least 500 valid records, which is enough to quote it alone. We leave out ccTLDs that don't show which country a domain belongs to, .io, .ai and .co among them, as we do on the map. Their counts are in the published aggregates. The section on reports that go nowhere explains how we check each reporting address, and what our own test found.
Adoption by TLD
The same measurement grouped by TLD, so it includes generic TLDs like .com that the map can't show. Every TLD with at least 1,000 measured domains, biggest first.
| TLD | Measured | Valid DMARC | At enforcement |
|---|---|---|---|
| .com | 423,738 | 49.2% | 24.8% |
| .net | 47,523 | 36.5% | 18.8% |
| .org | 40,977 | 56.3% | 28.1% |
| .ru | 34,414 | 41.4% | 21.8% |
| .de | 30,112 | 62.2% | 32.4% |
| .uk | 29,526 | 36.9% | 20.1% |
| .nl | 17,812 | 42.4% | 23.0% |
| .jp | 16,697 | 57.4% | 13.4% |
| .br | 15,965 | 64.5% | 28.6% |
| .it | 15,126 | 43.7% | 15.8% |
| .cn | 11,365 | 14.5% | 8.0% |
| .in | 10,948 | 50.3% | 26.7% |
| .fr | 10,825 | 64.4% | 29.1% |
| .io | 10,395 | 53.6% | 31.5% |
| .pl | 9,829 | 66.0% | 32.7% |
| .xyz | 7,301 | 9.6% | 5.2% |
| .au | 7,276 | 74.2% | 43.2% |
| .co | 6,895 | 54.3% | 28.5% |
| .cc | 6,884 | 8.8% | 4.4% |
| .ca | 6,324 | 58.3% | 32.0% |
| .eu | 5,966 | 53.7% | 26.6% |
| .es | 5,618 | 64.5% | 33.6% |
| .cz | 5,451 | 63.1% | 26.4% |
| .us | 5,451 | 21.6% | 12.7% |
| .shop | 5,073 | 34.1% | 3.6% |
.com is the largest registry in the snapshot at 423,738 measured domains, 49.2% of them publishing a valid DMARC record. 58 further TLDs clear the 1,000-domain bar and aren't listed here. The full set is in the published aggregates.
Domains are grouped by the last label of their name, so .co.uk is grouped under .uk. All percentages are based on the domains whose DMARC lookup got an answer, which matches the rest of this report.
Reports that go nowhere
The rua tag on a DMARC record holds the addresses receivers send aggregate reports to. A record with a rua on it looks like monitoring. When the address is on another organization's domain, receivers following the DMARC standard send reports only if that domain has published a second record that authorizes them.
129,077 domains publish a p=none DMARC record with no reporting address, or with only outside addresses whose owners haven't authorized the reports. That's 12.9% of measured domains and 26.9% of every valid record in the scan. We call it monitoring theater. The record passes a presence scan as adoption, but receivers that follow the DMARC standard don't send the domain reports.
We looked at DNS only, so we checked whether a domain has a reporting address that receivers following the DMARC standard send reports to. We didn't check whether the mailbox accepts mail, or whether anyone reads a report. The monitoring tier means the domain has such an address, or one we could not rule out. It doesn't mean anyone reads the reports.
When a rua address sits on another organization's domain, RFC 7489 requires an authorization TXT record at <domain>._report._dmarc.<rua-host>. A receiver that finds no DMARC version tag there must not send the report, and it raises no error to the domain that named the address. 35,261 domains in this scan name at least one outside address whose owner hadn't authorized them when we scanned. Receivers that run this check don't send reports to that address. Some of those domains also name another address, on their own domain or with authorization, and receivers can still send reports there.
30,359 domains name only outside addresses without authorization. They're the domains in the unauthorized group with no other address to fall back on. Neither group covers every p=none domain whose reports go nowhere, because a p=none record without a reporting address counts too. It gives receivers nowhere to send a report.
When we tested the authorization rule on our own domain in October 2026, Google and Microsoft sent reports to a Gmail address without an authorization record, and Yahoo did not. So a domain that names only outside addresses without authorization may still get reports from Google and Microsoft. Of the 129,077 p=none domains whose reports go nowhere, at most 30,359 are counted because they name only such addresses, and at least 98,718 don't name a reporting address at all.
197,142 domains name at least one outside reporting address. 137 of them name at least one address the scan could not resolve to a verdict, which counts as neither authorized nor unauthorized. A domain is counted once in each category it has an address in, so the categories overlap. At least 2,145 domains name more than one kind, which is why the counts sum past the total.
SPF
DMARC passes on SPF or DKIM alignment. When receivers can't evaluate a domain's SPF record, its mail can pass DMARC only on DKIM.
637,257 of the 1,000,438 measured domains publish an SPF record. 36,551 of those records (5.7%) return permerror for at least some of the mail checked against them. 17,640 exceed the ten-lookup limit, where receivers fail any sender they need an eleventh lookup to find. 9,385 publish more than one record, which receivers refuse to evaluate at all. The scan also flags an include that points at a name with no SPF record, and an include or redirect loop. It doesn't check for other syntax errors, so the true permerror count can only be higher.
How the records end
| Ending | Domains | Share |
|---|---|---|
| -all (fail) | 233,853 | 38.9% |
| ~all (softfail) | 333,180 | 55.5% |
| ?all (neutral) | 15,834 | 2.6% |
| +all (pass everything) | 377 | 0.1% |
| no all mechanism | 17,462 | 2.9% |
Percentages are out of the 600,706 domains with a usable SPF record. Records that already permerror are left out, because receivers never reach their final mechanism.
US federal .gov domains
This is every registered federal .gov zone in CISA's public registry, across the executive, legislative and judicial branches. CISA's Binding Operational Directive 18-01 required executive-branch domains that send mail to reach p=reject by October 2018. This list is wider than that, and its parked and redirect names have no mail to protect, so its reject share is not a compliance rate.
| Tier | Domains | Share |
|---|---|---|
| No record | 182 | 13.8% |
| Broken | 6 | 0.5% |
| Reports go nowhere | 24 | 1.8% |
| Monitoring | 19 | 1.4% |
| Partial enforcement | 8 | 0.6% |
| Enforced | 1,081 | 81.8% |
The enforced count, 1,081, sits above the reject count, 1,059, because 30 quarantine domains join the tier and 8 weakened domains leave it.
The federal reject share is 80.2% (1,059 domains). Across the whole scan it's 11.9% (119,327 domains), and that figure already includes this cohort. 182 federal domains (13.8%) publish no DMARC record, and 24 publish p=none with reports that go nowhere. Some of those are parked names that send no mail. The registry doesn't say which.
1,321 domains in the cohort. 80.2% publish p=reject and 82.4% publish quarantine or reject.
Who is named to receive the reports
The rua tag lists where a domain wants its reports sent, so counting rua hosts shows which platforms domains pick to receive them. A rua entry doesn't prove a report arrives, and these counts aren't the platforms' customer numbers. The largest named collector is Cloudflare, on 30,785 domains, then Brevo at 17,019 and Valimail at 15,048.
A p=none domain whose reports go nowhere moves to monitoring once it adds a reporting address on its own domain, or an outside one whose owner has authorized the reports. With a free collector, an operator can get a first report without signing a contract. Cloudflare, the largest collector in the scan, includes collection with the domain. Postmark's free DMARC Digests is the fifth largest collector.
Each bar counts the domains whose rua list includes a destination on that platform, once per domain. A domain naming two platforms counts for both.
155,379 matches across 53 platforms. The largest destination matching no platform in the list is gmail.com, named by 2,399 domains.
MTA-STS, TLS-RPT and BIMI
Far fewer domains publish these three records than publish DMARC.
Out of the 1,000,438 measured domains, 1.0% publish MTA-STS, 1.3% TLS-RPT and 1.8% BIMI.
Methodology
| Population | Tranco list 7462X, 1,000,000 domains, plus the cohort frames below. We scanned 1,000,748 domains in total. Tranco builds the list from the Crux, Farsight, Majestic, Radar and Umbrella rankings over the 30 days before the list date, combined with the Dowdall rule. Reference: 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, doi:10.14722/ndss.2019.23386. |
|---|---|
| Frames | tranco:
1,000,000 domains from https://tranco-list.eu/list/7462X cohort:us-federal-gov: 1,321 domains from research/cohorts/us-federal-gov.csv 573 cohort domains also rank in the Tranco million and were scanned once, which is why the frames add up to more than the number of domains scanned. |
| Collection window | 2026-11-01T08:26:48.247Z to 2026-11-01T09:58:05.578Z. Aggregated 2026-11-16T11:22:56.199Z. |
| Resolver | Queries went to unbound:5353, with retries against 1.1.1.1, 8.8.8.8. |
| Error rate | 310 DMARC lookups failed after retries, which is 0.03% of the domains scanned. We left those domains out of every percentage on this page. |
| Public Suffix List | Whether a reporting address is external is decided by comparing organizational domains against the Public Suffix List fetched 2026-11-01T08:26:47.582Z, sha256 617d1b276660e5a0. The list is pinned per snapshot so a re-run classifies destinations the same way. The list is used with its private section, so two customers of one platform that publishes a private rule count as external to each other, and a report address on a sibling tenant needs its own authorization record. |
| Multiple records | More than one v=DMARC1 record at _dmarc is invalid per RFC 7489. This scan scores those 4,034 domains as broken. |
| Vendor classification | Each rua destination is classified by the host that receives the mail, so a platform reselling another platform's collection counts under the receiving host. Hosts we don't recognize stay unmatched. We don't guess. |
| Absence versus failure | NXDOMAIN or an empty TXT answer is absence. A timeout or SERVFAIL after retries is unknown. Failures are never counted as absence. |
| Month over month | We compare each domain's policy with the one we stored for it in the October 2026 snapshot. If a lookup failed in either month, we drop the domain and count it separately, so a DNS failure never shows up as a policy change. |
| Independent comparison | The adoption tracker at dmarcdkim.com/dmarc-adoption updates daily. It covers 1.35 million domains it has collected over time. When we checked it in August 2026 it reported 11.4% at full reject, where this scan finds 11.9% (119,327 domains) at p=reject. This scan's count includes 1,831 rejects weakened by pct and 7,835 with a weaker sp, which the tier framework scores as partial. The two measurements agree closely on policy rates. Their adoption headlines are not directly comparable, because the tracker measures an accumulated set of domains and this scan measures one pinned Tranco snapshot. We haven't compared the two populations. |
| Scanner | Version 2cf1b67a-dirty. |
Limitations
The scan reads what domains publish in DNS. Nothing here is a delivered-mail pass rate or an actual rejection.
A domain whose reporting address couldn't be resolved to a verdict counts as neither authorized nor unauthorized. At p=none that puts it in monitoring, so the true count of domains whose reports go nowhere can only be higher.
- With 4 editions there are 3 months of change. This page compares only the latest month with the one before.
- Some DNS lookups fail even after retries. 310 ended in error and are excluded rather than scored as no DMARC. If those lookups had succeeded, any count on this page could have moved by up to that many domains.
- We checked rua authorization at scan time only. If a destination's owner added or removed an authorization after the scan, this snapshot doesn't show it.
- So far the only named cohort is the 1,321 US federal domains. Other planned cohorts aren't in this edition.
- We haven't matched every rua host to a platform yet. As we add more, some unmatched destinations will move into named platforms.
The data
- aggregates.json: every count and percentage on this page, as published. Dates, list ids and the scanner version are in the manifest next to it.
- Data repository: the per-snapshot files.
- Hugging Face dataset: the same files for every edition, with the headline counts as one CSV.
- This edition isn't archived on Zenodo. The archived editions are at doi:10.5281/zenodo.22849900, which resolves to the newest one.
Citing and reusing the data
We publish the aggregate data under CC BY 4.0. Reuse it, including commercially, with attribution. Suggested citation: