Know what a DMARC policy does
DMARC helps a receiving mail system decide what to do when a message claiming to come from your domain fails authentication. It checks whether SPF or DKIM passes and aligns with the domain in the message’s visible From address. One aligned pass is enough for DMARC to pass. A message can therefore fail SPF and still pass DMARC through aligned DKIM, or the other way around.
A DMARC policy is published as a DNS TXT record at a name such as _dmarc.example.com. The value v=DMARC1; p=none asks receivers to take no policy-based action on failing mail. Changing p to quarantine requests that failing mail be treated as suspicious, often by placing it in spam. Changing p to reject requests that it not be accepted. Receivers make their own handling decisions, so these policies are not guarantees of where every message will go.
The safest rollout starts with p=none, learns which systems send mail using the domain, and fixes legitimate failures before asking receivers to enforce the policy. DMARC is not a substitute for securing mail accounts or reviewing who is allowed to send on behalf of the organization.
Use monitoring to build a sender inventory
Publish one valid DMARC policy for the domain and, if you can process the data, include an aggregate-report destination. For example, v=DMARC1; p=none; rua=mailto:dmarc-reports@example.com requests aggregate reports at that mailbox. Before adding a reporting address, confirm that it can receive the expected volume of reports and that the people reviewing them understand what information they contain. If the address is on another domain, additional DNS authorization may be required.
Aggregate reports can show which sources send mail claiming to use your domain and how their messages perform under SPF, DKIM, and DMARC. Use them to list your regular senders: employee mail, help desks, newsletters, billing tools, website forms, and any service that sends notifications. Compare report patterns with configuration records and with what each team or service owner says they send. A source appearing in a report is not, by itself, proof that it is authorized.
Give the inventory time to capture less frequent mail, such as monthly notices or account-recovery messages. Reports provide a useful view of authentication results, but they do not prove that recipients received or read a message. Test important mail flows separately.
Fix legitimate mail before enforcing
For each authorized sender that fails DMARC, determine whether its SPF or DKIM result can be made to align with the visible From domain. A service may have a passing SPF check for its own return-path domain, yet still fail DMARC because that domain does not align. Where supported, configure a custom return-path domain or aligned DKIM signing domain, following the service’s setup instructions. Check the resulting DNS records and send test messages rather than assuming that a published record is being used.
DKIM is often particularly useful for mail that is forwarded, because SPF commonly evaluates the forwarding server rather than the original sender. Forwarding and mailing lists can still cause failures, especially if a list changes a signed message. Consider those paths when interpreting reports; do not treat every failed result as a reason to authorize an unfamiliar sender. Avoid adding a service to SPF solely because its name appears in a report.
Review the visible From addresses used by each legitimate service. If a system cannot authenticate mail in alignment with your domain, decide whether it should use a different From domain or whether its configuration needs to change. Keep a record of the owner and purpose of each approved sender so future failures are easier to investigate.
Move to quarantine, then consider reject
Once known senders authenticate consistently, change the policy to p=quarantine and watch both reports and real mail flows. Test messages that matter to the organization, including replies, notifications, and less frequent messages when they occur. Have a way for users or service owners to report missing mail. If a legitimate flow breaks, investigate its authentication and alignment before tightening the policy further.
The optional pct tag can request that a policy apply to a portion of failing messages while you test enforcement. It is not a reliable promise that every receiver will handle exactly that percentage, so it should not replace monitoring and delivery checks. When quarantine produces no unexplained problems over a suitable period for your sending patterns, consider p=reject. Reject is the stronger request and leaves less room for a legitimate sender that was overlooked.
Check subdomains as part of the decision. An organizational domain’s DMARC policy can affect subdomains that lack their own DMARC record, and the optional sp tag can specify a separate policy for them. Inventory subdomains used for mail before applying a broad policy. Whether you move in stages or directly to reject, keep reviewing reports and update the sender inventory whenever a mail service changes.



