Policy split
Every resolved domain, by the policy it publishes.
- 39.7% No DMARC record
- 21.3% p=none (monitor only)
- 17.1% p=quarantine
- 21.5% p=reject
- 0.4% Malformed or mistyped policy
Reporting and partial enforcement
- 79% of DMARC adopters ask for aggregate reports (rua=).
- 22.1% of adopters set an explicit subdomain policy (sp=).
- 3.9% of quarantine and reject domains apply the policy to only part of their mail (pct below 100).
By rank band
| Tranco rank | Publish DMARC | p=reject |
|---|---|---|
| 1–1,000 | 73.4% | 44.5% |
| 1,001–10,000 | 67.9% | 29.9% |
| 10,001–100,000 | 59.5% | 20.6% |
By top-level domain
TLDs with at least 1,000 resolved domains in the population.
| TLD | Publish DMARC | p=reject |
|---|---|---|
| .com | 60.3% | 22.7% |
| .net | 42.8% | 15.6% |
| .org | 66.7% | 21.4% |
| .ru | 62.7% | 20.8% |
| .de | 76.1% | 23.6% |
| .jp | 68.5% | 10.5% |
| .io | 57.4% | 22.2% |
| .br | 77.2% | 21.6% |
| .uk | 81.9% | 37% |
| .in | 63.1% | 23% |
| .cn | 25.7% | 8.3% |
| .fr | 80.6% | 26.1% |
Trend
The trend line is drawn once 4 weeks are published. This is week 1. New scans run Sundays at 22:37 UTC.
Methodology
The population is Tranco list PY2NJ (2026-11-11), top 100,000, pinned for a year so that a weekly change is a change in what domains publish, not in who is on the list. This week, 96,085 of 100,000 domains resolved (96.1% coverage). Public DNS only, no customer data, aggregates only.
- For every domain on the pinned Tranco list, the scanner asks for the domain's NS records and then the TXT records at _dmarc.<domain>, through public resolvers (1.1.1.1 and 8.8.8.8), with a 3-second timeout and two tries.
- A domain counts as resolved when it answers the NS query. A domain that resolves but whose _dmarc lookup times out or fails is excluded, so resolver trouble never reads as missing DMARC.
- Adoption is the share of resolved domains publishing any record that starts with v=DMARC1. More than one such record, a missing p= tag, or an unknown policy value counts as malformed under the current DMARC standard (RFC 9989).
- Reporting and subdomain-policy shares are of adopters. The partial-enforcement share is of quarantine and reject domains whose pct= tag is below 100.
- The method matches the annual report. The population does not: the annual report scans the top 10,000, the tracker the top 100,000.
Limits
- Tranco ranks web popularity, not mail volume. Many listed domains send no mail, and a parked domain with no DMARC counts the same as a bank with none.
- One vantage point: two public resolvers, from one network. A different resolver can see a different answer for a misconfigured zone.
- Domains that do not resolve are excluded from every percentage, and coverage varies a little from week to week. Week-over-week moves under 0.3 points are within that noise.
- The population is pinned for a year, so the list ages: domains that drop off the live ranking stay in the scan until the yearly re-pin.
- DNS shows what a domain publishes today, not what it published before the first scan. There is no backfill.
Every week
Each week keeps its own page and is never revised.
- November 16, 2026: 60.4% publish DMARC
- November 13, 2026: 60.3% publish DMARC