Beta testing is running. New organisations join the beta and are not billed during it - sign up for the beta.

DMARC failure explanations

DMARC failure explanations: the likely cause and what to do next.

Every domain in DMARCHub has a 'Why mail is failing' section. It reads your DMARC aggregate reports, groups the failing mail by likely cause, and gives you plain steps to fix it, with the evidence alongside.

For each cause

What every explanation shows you.

Explanations cover the last 30 days of reports, or a longer period you choose within the domain's data retention, and sit at the top of each domain's Messages tab.

The likely cause

Failing mail is grouped by cause and by the email service sending it, so a hundred sources with the same problem read as one explanation, not a hundred rows.

How sure we are

Each explanation is marked as a clear, likely or possible cause, depending on how strong the evidence in your reports is.

What receivers did

How many messages and sources are affected, and how many the receivers delivered, quarantined or rejected under your policy.

What to do

Plain steps for that cause, worded to match how sure we are. A possible cause gets steps to check first, not changes to make.

The evidence

The busiest sources behind each explanation, exactly as receivers reported them, with the DKIM and SPF domains that were checked.

Who reported it

How many of the receivers reporting it we could confirm as genuine, because anyone can send a forged DMARC report.

Causes

The common reasons DMARC fails.

Most failures come down to SPF alignment, DKIM alignment, or mail that is not yours. These are the causes the explanations recognise.

A service signs with its own domain

A marketing, CRM or helpdesk platform passes SPF or DKIM for its own domain rather than yours, so DMARC alignment fails.

A sender is missing from SPF

Mail comes from a server your SPF record does not list, or your SPF record returns an error.

A DKIM signature for your domain failed

The message was signed with your domain but the signature did not verify. The key may be missing or out of date, or the message may have been changed after signing.

Strict alignment is too strict

SPF or DKIM passes for a subdomain of yours, but your DMARC record asks for an exact match.

Forwarding and mailing lists

Forwarding and mailing lists commonly break SPF and sometimes DKIM. Where reports cannot tell forwarding from spoofing, we flag it as possible forwarding or spoofing.

No authentication, or an unknown source

Mail with no SPF or DKIM result for your domain, or from sources we cannot match to a known service. This can be someone else using your domain.

When the reports do not say enough to tell, the explanation says the cause could not be determined rather than guessing. For the background on alignment, read how DMARC, DKIM and SPF fit together or our guide to why DMARC fails and how to fix it.

Trust

Careful about who to trust.

Anyone can send a DMARC report to a public reporting address, so a report on its own is not proof of anything. We only suggest approving a sending service when we recognise the service from its confirmed server name and at least two receivers we have confirmed as genuine reported it. Shared mailbox providers such as Microsoft 365 and Google Workspace are never assumed to be yours, because their servers also carry other organisations' mail.

Where failing mail could be someone else sending as your domain, the explanation says so and the steps tell you how to check, rather than pointing you towards adding it to your SPF record.

Optional

Plain-English summaries, written by AI.

On eligible plans, an ‘Explain in plain English’ button sits above the explanations. It asks Microsoft Azure OpenAI for a short summary of the findings, written for someone who does not work with DNS every day. It runs only when you click it, and the summary is clearly labelled as written by AI.

The causes, confidence and steps always come from our own fixed rules. The summary only restates them, so the steps in each explanation remain the ones to follow.

What the summary sees

  • Sent: cause categories, message counts, the names of well-known email services, your domain name and our own explanatory text.
  • Never sent: IP addresses, server names, report contents or anyone's personal details.
  • Processed within the EU, in Microsoft's EU Data Zone, and not used to train models.
  • The summary is checked before display and shown as plain text.

The detail is in our privacy notice and on the security page.

FAQ

Questions about DMARC failures.

Where do I find the failure explanations?

At the top of each domain's Messages tab, in the 'Why mail is failing' section. They cover the last 30 days of reports, or a longer period you choose within the domain's data retention, and update as new DMARC reports arrive. For the background on why DMARC fails, read our guide to the common causes and fixes.

What is a DMARC failure report?

People use the phrase for two things. Aggregate (RUA) reports are daily summaries from each receiver that include every failing source. Forensic (RUF) reports describe a single failed message. DMARCHub reads both, and the failure explanations are built from the aggregate reports.

How sure are the explanations?

Each explanation is marked as a clear cause, a likely cause or a possible cause, depending on how strong the evidence in your reports is. Where the reports do not say enough to tell, we say the cause could not be determined rather than guess.

Will DMARCHub tell me to approve a sender I do not recognise?

Only when the evidence backs it up. We suggest approving a sending service only when we recognise the service from its confirmed server name and at least two receivers we have confirmed as genuine reported it. Otherwise the steps tell you how to check first, because failing mail can be someone else using your domain.

Is the AI summary used to work out the cause?

No. The causes, confidence and steps come from fixed rules applied to your reports. The optional AI summary, available on eligible plans, only restates those findings in plain English. The steps in each explanation remain the ones to follow.

What data does the AI summary use, and where is it processed?

Only the findings for that domain: cause categories, message counts, the names of well-known email services, your domain name and our own explanatory text. No IP addresses, server names, report contents or personal details are sent. It is processed within the EU by Microsoft Azure OpenAI and is not used to train models.

See everything DMARCHub does, including managed SPF, hosted MTA-STS and DNS drift alerts, or learn how to read a DMARC aggregate report.

Beta

See why your mail is failing.

DMARCHub beta testing is running. Sign up with your work Microsoft or Google account to join it. Add your domain, and the explanations appear as your DMARC reports arrive.