CertiMonitor

Blog

DMARC with p=none doesn't stop spoofing

Jaime Guerrero · · Updated · 3 min read

In February 2024 Google and Yahoo started requiring bulk senders to publish a DMARC record. A lot of domains added one at that point, and most chose p=none, the policy that meets the requirement while changing nothing about delivery. A message pretending to come from your domain still lands in the inbox.

You can look up any domain's DMARC record with dig. example.com, for instance, publishes an enforced policy:

$ dig +short TXT _dmarc.example.com
"v=DMARC1;p=reject;sp=reject;adkim=s;aspf=s"

Run it against your own domain. If the record says p=none, as in v=DMARC1; p=none; rua=mailto:dmarc@yourdomain.com, receivers are only reporting on spoofed mail, not stopping it.

SPF, DKIM and DMARC

Email has no built-in way to prove who sent a message. These three DNS records add one.

RecordWhat it doesWhere it lives
SPFLists the servers allowed to send mail for the domainTXT record on the domain, starts with v=spf1
DKIMPublishes a key that receivers use to verify each message's signatureTXT record at selector._domainkey.example.com
DMARCTells receivers what to do when a message fails both, and where to send reportsTXT record at _dmarc.example.com, starts with v=DMARC1

SPF and DKIM give the receiving server something to check. DMARC adds the rule that the domain passing SPF or DKIM has to match the domain in the From address, which is the one people read, and it says what happens when a message fails.

The three policies

p=none means deliver the message as usual and send me a report. p=quarantine asks the receiver to treat failing mail as suspicious, which usually means the spam folder. p=reject asks the receiver to refuse it.

p=none was meant as the first step, a way to find every service that sends mail as your domain before you start enforcing. Plenty of domains never take the second step.

Getting to reject

Switching straight to p=reject tends to block your own invoices, password resets and newsletters, because some tool you forgot about sends mail without proper authentication. This order works:

  1. Add a reporting address to the record: rua=mailto:dmarc@yourdomain.com. Receivers send daily XML reports listing every server that sent mail as your domain. A DMARC report viewer makes them readable.
  2. Go through the senders in the reports: your mailbox provider, website, CRM, helpdesk, billing tool. Set up DKIM signing with your domain for each one, and add them to SPF.
  3. Wait a couple of weeks until the reports show your real mail passing.
  4. Change to p=quarantine and watch the reports for a couple of weeks.
  5. Change to p=reject; t=y first. The t=y test flag asks receivers to apply one level less, so they still quarantine instead of rejecting while you confirm nothing legitimate fails.
  6. Remove t=y to enforce p=reject fully.

Older guides use pct= to roll out gradually. The May 2026 DMARC revision, RFC 9989, removed it in favor of t=. More on what changed in how SPF, DKIM and DMARC actually decide.

Subdomains follow p= unless you set sp=, so an enforced main policy already protects billing.yourdomain.com. If you do set a looser sp= while a subdomain's senders are being fixed, add np=reject: it applies only to subdomains that don't exist at all, a favorite target for spoofing because nothing legitimate ever sends from them. Then tighten sp= later.

SPF has a 10-lookup limit

Each include: in an SPF record counts as a DNS lookup, and so do the includes inside it. Past 10, SPF fails for all of your mail. Remove services you no longer use before adding new ones.

You can see a domain's SPF, DKIM and DMARC setup here, including whether DMARC is enforced: