Blazalek.com

How to read a DMARC aggregate report XML, record by record

Reading a DMARC aggregate report comes down to identifying which systems sent the mail and finding who manages them. Grab your XML report, your list of approved sending services, and server logs for that window. If an IP or domain is unfamiliar, investigate who set it up before assuming it’s rogue.

By

On this page

Reading a DMARC aggregate report comes down to identifying which systems sent the mail and finding who manages them. Grab your XML report, your list of approved sending services (ESP accounts, SaaS tools, transactional gateways), and server logs for that window. If an IP or domain is unfamiliar, investigate who set it up before assuming it’s rogue. If you are examining aggregate telemetry for the first time, start with the fundamentals of reading aggregate (RUA) reports.

Keep the scope attached before you count anything

Run the report through your established processing path. Set aside malformed payloads for inspection. Valid formatting on its own doesn’t make the reporting truthful; RFC 9990 addresses both questionable formats and false report information. If your challenge is collection or pipeline backlog rather than reading files, consult the aggregate-report processing runbook.

Keep each extracted record linked to its reporter, report ID, UTC date window, and observed policy. Before combining totals across files, check for duplicate and overlapping reports: partial overlap may make precise deduplication impossible, a limit explained in the RFC 9990 section on duplicate handling. When a policy rollout occurs, track policy transitions, since a single report can reflect one published configuration while containing dispositions evaluated under an earlier one (policy transitions).

The aggregate report XML, record by record

The fixture below is an original synthetic scenario. The domains use reserved .example names and the IP addresses come from documentation address ranges.

This walkthrough uses strict settings for illustration. The fixture follows the RFC 9990 schema with quarantine active, testing disabled, and strict alignment. Those are exercise assumptions, not advice for a production configuration and not evidence that any receiver has adopted them. Choosing treewalk as the discovery method doesn’t make psl invalid; the format keeps the latter to describe legacy discovery (discovery-method values).

<?xml version="1.0" encoding="UTF-8"?>
<feedback xmlns="urn:ietf:params:xml:ns:dmarc-2.0">
  <version>1.0</version>
  <report_metadata>
    <org_name>Harbor Receiver Example</org_name>
    <email>reports@harbor.example</email>
    <report_id>harbor-20260912-cedar-01</report_id>
    <date_range>
      <begin>1789171200</begin>
      <end>1789257599</end>
    </date_range>
  </report_metadata>
  <policy_published>
    <domain>cedar.example</domain>
    <discovery_method>treewalk</discovery_method>
    <p>quarantine</p>
    <adkim>s</adkim>
    <aspf>s</aspf>
    <testing>n</testing>
  </policy_published>
  <record>
    <row>
      <source_ip>192.0.2.46</source_ip>
      <count>84</count>
      <policy_evaluated>
        <disposition>pass</disposition><dkim>pass</dkim><spf>fail</spf>
      </policy_evaluated>
    </row>
    <identifiers>
      <header_from>cedar.example</header_from>
      <envelope_from>returns.dispatch.example</envelope_from>
    </identifiers>
    <auth_results>
      <dkim><domain>cedar.example</domain><selector>cedar26</selector><result>pass</result></dkim>
      <spf><domain>returns.dispatch.example</domain><scope>mfrom</scope><result>pass</result></spf>
    </auth_results>
  </record>
  <record>
    <row>
      <source_ip>198.51.100.72</source_ip>
      <count>19</count>
      <policy_evaluated>
        <disposition>quarantine</disposition><dkim>fail</dkim><spf>fail</spf>
      </policy_evaluated>
    </row>
    <identifiers>
      <header_from>cedar.example</header_from>
      <envelope_from>returns.notice.example</envelope_from>
    </identifiers>
    <auth_results>
      <dkim><domain>notice.example</domain><selector>notice26</selector><result>pass</result></dkim>
      <spf><domain>returns.notice.example</domain><scope>mfrom</scope><result>pass</result></spf>
    </auth_results>
  </record>
  <record>
    <row>
      <source_ip>203.0.113.91</source_ip>
      <count>7</count>
      <policy_evaluated>
        <disposition>quarantine</disposition><dkim>fail</dkim><spf>fail</spf>
      </policy_evaluated>
    </row>
    <identifiers>
      <header_from>cedar.example</header_from>
      <envelope_from>cedar.example</envelope_from>
    </identifiers>
    <auth_results>
      <dkim><domain>cedar.example</domain><selector>unassigned26</selector><result>fail</result></dkim>
      <spf><domain>cedar.example</domain><scope>mfrom</scope><result>fail</result></spf>
    </auth_results>
  </record>
</feedback>

The reporting window covers 12 September 2026, from 00:00:00 to 23:59:59 UTC. These timestamps mark the collection boundaries of the report rather than timestamps of specific messages (date-range semantics). For this walkthrough, label the three records below as A, B, and C in order of appearance, retaining the report ID alongside each label.

Treat auth_results as raw authentication results, while interpreting policy_evaluated/dkim and /spf as DMARC alignment outcomes. Under strict alignment parameters, sending identifiers must be identical to the author domain, and passing DMARC requires one passing aligned path.

Group / countRaw evidence retainedReconstructed result
A / 84DKIM cedar.example, selector cedar26: pass. SPF returns.dispatch.example: pass.DKIM matches cedar.example; SPF doesn’t. DMARC passes.
B / 19DKIM notice.example, selector notice26: pass. SPF returns.notice.example: pass.Neither domain matches cedar.example. DMARC fails even though both raw checks passed.
C / 7DKIM cedar.example, selector unassigned26: fail. SPF cedar.example: fail.Matching names can’t supply successful authentication. Both evaluated paths fail.

The three records account for 110 messages: 84 passing and 26 failing. With quarantine active, group A receives disposition=pass while groups B and C carry quarantine, which is consistent with their evaluated failures; the RFC 9990 schema defines these as distinct action values.

Keep individual groups distinct even when an IP address recurs across multiple lines, because each record binds an IP to specific authentication results and identifiers (record grouping). These 110 observations reflect the perspective of a single reporting receiver during one defined time window. They do not prove that 110 messages reached user inboxes, as local delivery handling remains at the discretion of the receiving provider (final handling).

Who owns each source, and what each owner does next

Connecting an XML report to infrastructure requires external business records. For this exercise, assume an approved sender inventory, SaaS dashboards, and message logs match group A to the Dispatch account cedar-orders, and group B to the Notice account cedar-billing. Both applications send with cedar.example in the From header, matching their configured authentication domains in the XML. The order operations lead owns group A, the billing operations lead owns group B, and group C has no matching records.

Inventory entryEvidence and confidenceAssigned next step
A: order confirmationsCorroborated service, account and business owner; the report shows aligned DKIM. Delivery not established.Order operations lead: keep the evidence references with group A; no authentication repair follows from this group.
B: billing noticesCorroborated legitimate stream; the raw domains show an alignment problem. Customer impact not established.Billing operations lead: assign the Notice administrator to investigate the account’s authenticated domains. Due 15 September 2026.
C: unassignedSeven reported failures; owner, cause and business impact unknown.Messaging operations lead: reconcile service and account records and ask internal owners for matching sending evidence. Due 16 September 2026.

The owners and due dates are an example workflow, not protocol requirements. For row B (billing notifications due 15 September 2026), assign the Notice administrator to investigate the account’s authenticated domains, and close the task only once a verified configuration correction and comparable later observations show a passing aligned path. Establishing delivery recovery or customer impact requires separate investigation through server logs, delivery status notifications, or direct recipient inquiries. For row C (unassigned, due 16 September 2026), trace the origin before closing the ticket.

For group C, attach the raw report ID and record details to an investigation ticket. Reconcile service and account records and ask internal owners for matching sending evidence. A failure alone can’t classify legitimacy: legitimate mail can fail DMARC. Keep ownership unconfirmed until corroborating evidence identifies the account, its purpose and the responsible person, or a documented investigation reaches another supported conclusion. To understand the diagnostic boundaries of reporting data, review the runbook analysis on detecting spoofing with DMARC reports.

Prioritize investigation by business risk and critical mail flows, factoring in message volume and failure frequency. Nineteen billing failures justify investigating a known stream. Seven mystery failures still need an investigation owner. Neither number is a universal urgency threshold.

When telemetry conflicts, pause enforcement conclusions and reconcile raw data with the reporting receiver or service owner. The absence of error reports does not confirm healthy delivery: receiving providers can opt out of reporting, and a rejection before DATA can leave no record.

Do not advance DMARC enforcement while legitimate business traffic continues to fail authentication. RFC 9989 requires remediation before enforcement, and readiness also depends on sending cadence. Once you have cataloged your senders and fixed failing streams, you can decide whether you’re ready to move from quarantine to reject.

The natural next step is to reconcile remaining authorized services using the sender inventory procedure.

Frequently asked questions

Does a DMARC pass in the report mean the message reached the inbox?

No. The 110 observations in this report confirm only authentication outcomes during the evaluated window. Final handling stays the receiver’s decision.

Can both SPF and DKIM pass in auth_results while DMARC fails?

Yes. In group B both raw cryptographic checks succeeded, but neither domain aligned with cedar.example. Under strict alignment rules, any domain mismatch causes DMARC to fail.

Does a failing record prove the source is spoofing?

No. A failure alone can’t classify legitimacy: legitimate mail can fail DMARC. Ownership remains unconfirmed until you verify the sending service, credentials, and operational owner.

Does a missing report mean the problem is fixed?

No. A lack of aggregate records does not prove that delivery has recovered. Receiving mail servers can opt out of generating feedback, and a rejection before DATA can leave no record.

Can you tighten the DMARC policy while groups are unresolved?

Policy escalation waits while legitimate authentication problems remain open. RFC 9989 requires remediation before enforcement, and readiness also depends on sending cadence; a policy change needs its own decision.

Reconcile your sender inventory

Email deliverability consultant

Get a second opinion on your email setup

Tell me what your customers are seeing or what you’re planning to change. In 30 minutes, we’ll work out what your team can check first and whether a deliverability audit would help.

I led deliverability at GetResponse, designed a dedicated newsletter system for Easy.Tools and helped Youmoney migrate its automated emails in stages. Read the case studies

Book free call

Email deliverability audit

See what I check and read a sample report with recommendations for your team.