DMARC is easy to enable and easy to break.

The safe path is not “set p=reject on Friday.” It is a staged rollout: observe first, quarantine the small failures, then reject only when the data is boring.

Start at p=none

p=none does not enforce anything. It tells receivers to send reports while you learn which senders use your domain.

Add a reporting address:

v=DMARC1; p=none; rua=mailto:dmarc@example.com; fo=1 

Wait until you have enough reports to see every legitimate sender. This can take days or weeks depending on volume.

Fix alignment before enforcement

For each legitimate source, confirm:

  • SPF passes and aligns with the From domain
  • DKIM passes and aligns with the From domain
  • the Return-Path domain is expected
  • third-party senders are authorized

A sender can pass SPF but fail DMARC if the domains do not align. Do not confuse “SPF pass” with “DMARC pass.”

Move to quarantine

Start with a small percentage:

v=DMARC1; p=quarantine; pct=10; rua=mailto:dmarc@example.com 

Watch the reports for legitimate mail that fails. Increase pct gradually.

Do not jump to 100% just because a test message passed. Reports show the long tail.

Move to reject

When failures are only from unauthorized or unwanted sources, move to reject:

v=DMARC1; p=reject; rua=mailto:dmarc@example.com; fo=1 

Keep reporting enabled. Reject does not mean you can stop watching.

Test the policy before publishing

A policy simulator lets you check what would happen to a message under none, quarantine, and reject before changing DNS:

https://mailers.email/tools/dmarc-policy-simulator/

Common mistakes

  • enabling reject without reports
  • forgetting a legitimate ESP
  • using a subdomain policy that is stricter than the parent
  • ignoring SPF lookup limits
  • not monitoring the reporting mailbox

The rollout rule

Observe, fix, quarantine, reject. In that order.