By DMARCdrift
Why Your DMARC Reports From Microsoft Are Late or Missing
You published a DMARC record with a rua= address, Google's report showed up the next day, and Microsoft's didn't. Usually nothing is broken. Aggregate reports are batch jobs on the receiver's side: they run on the receiver's schedule, they only cover mail the receiver actually saw, and they can be skipped. This post covers how that pipeline works, the common reasons a Microsoft report is late or absent, and how to tell a normal gap from a real problem.
How aggregate reporting works on the reporter's side
An aggregate report summarizes the mail one receiver saw for your domain over a fixed reporting period. RFC 9990, the DMARCbis aggregate reporting spec, says the period "will encompass a single UTC day, beginning at 0000UTC" in the typical case. The report is generated after that window closes, compressed, and emailed to your rua= address.
Microsoft's DMARC configuration guide for Microsoft 365 says the aggregate report email is "typically sent once per day (the report contains the DMARC results from the previous day)". So a message you send on Monday afternoon lands in a report covering Monday UTC, which is sent at some point after Monday ends. Depending on your time zone, that can feel like two days.
Reports also arrive in batches. A receiver works through its whole queue of domains, so you might get Google's and Yahoo's reports in the morning and Microsoft's hours later. Our guide to when reports arrive puts the normal wait for a first report at 24 to 48 hours, with occasional 48 to 72 hour delays from slower receivers.
If you have never opened one of these files, what is in a DMARC aggregate report walks through the XML.
Why Microsoft reports arrive late or not at all
No mail reached Microsoft
A report only covers mail the receiver handled. If nothing from your domain reached a Microsoft mailbox that day, there is nothing for Microsoft to report on. This is the most common cause for small senders: a side project that emails a few Gmail users a week may simply not touch Microsoft on most days.
To test it, send a real message from the domain to an Outlook.com or Microsoft 365 address you control, then wait for the next UTC day to close and give it another day.
The recipient's mail goes through a filter first
Microsoft's documentation adds a condition that surprises people. Microsoft 365 sends aggregate reports "as long as the MX record for the Microsoft 365 domain points directly to Microsoft 365". If a company routes inbound mail through a third-party filter or an on-premises server before it reaches Microsoft 365, "DMARC Aggregate reports aren't sent", according to the same page.
This is about your recipients' setup, not yours. If most of your Microsoft-hosted recipients are businesses behind a security gateway, you may see few or no Microsoft reports even though they get your mail.
The rua destination is on another domain and isn't authorized
If your rua= address is on a different domain from the one you are monitoring (a reporting service, or your main company domain receiving reports for a side project), the receiving domain has to publish an authorization record. RFC 9990 section 4 defines it: for example.com sending reports to reports@example.net, the reporter looks up a TXT record at example.com._report._dmarc.example.net that starts with v=DMARC1.
Microsoft's guide spells out the consequence: "Without this record, receivers don't deliver DMARC reports to the external address." It also lists "cross-domain auth missing" as a likely cause when aggregate reports aren't received.
The mailbox rejects the report
Reports are ordinary emails with a gzip or XML attachment. A mailbox that is full, an address that was deleted, or a filter that strips compressed attachments will bounce or discard them. RFC 9990 notes that this transport "potentially encounters a problem when feedback data size exceeds maximum allowable attachment sizes", and that skipping compression can push a report past the receiving server's size limit. Most reports are small, but a busy domain sending through several services produces bigger ones.
The DMARC record itself is wrong
Reporters can only send to an address they can parse. Check for:
- Two DMARC records at
_dmarc. Under RFC 9989, if more than one DMARC record is returned they are all discarded, so the domain has no usable policy and norua. - A malformed
ruavalue. It needs themailto:scheme, and multiple addresses are separated by commas. - The record at the wrong name, such as the apex or
dmarc.example.cominstead of_dmarc.example.com.
The DMARC record checker parses the record and shows the rua destinations it finds.
Microsoft didn't send one for that day
Sometimes none of the above applies and a day is just missing. Reporting runs on the receiver's infrastructure, and a backlog or an incident there can delay a run or, at worst, skip a period. You can't fix this from your side; the useful question is whether it happens once or keeps happening.
An example of a reporter-side incident
On Thursday 2026-11-12, a Microsoft 365 service incident began affecting Exchange Online mailboxes, Teams and other services. Microsoft was still restoring service region by region on Friday 2026-11-13 and had not published a resolution or root cause by the time we wrote this on 2026-11-14. SecurityWeek reported that Microsoft tracked it as EX1550543 for Exchange Online and MO1550821 for the wider Microsoft 365 incident. ABC News quoted Microsoft saying the elevated load "isn't driven by an increase in traffic, but by a number of servers being unexpectedly unavailable." During recovery, Microsoft said it had extended the email timeout to 48 hours "to help preserve delayed mail", so messages queued during the incident were held and retried rather than bounced.
This was a large incident, and it is a useful test of the advice in this post. In the reports DMARCdrift received, Microsoft's aggregate reports for 2026-11-11 and 2026-11-12 both arrived, even though Microsoft had to generate and send both while the incident was under way. The first report for each day came in 42 and 26 hours after that UTC day ended. That is the same range we saw before the incident: 42 hours for 2026-11-09 and 26 hours for 2026-11-10. Fewer Microsoft reports arrived for 2026-11-11 and 2026-11-12 than for 2026-11-10. That can mean less mail from those domains reached Microsoft that day, or that a straggler is still on its way. With numbers this small, it is not proof of a dropped report.
So even a multi-day outage at Microsoft did not leave a visible hole in reporting for us. If a Microsoft report is missing for a day around an incident like this:
- Check whether the incident is resolved, and its timeline, in Microsoft 365 Service Health, or in the public coverage if you don't have admin access.
- Wait at least two more days before changing anything. A late report for the affected day may still arrive.
- Compare with other reporters for the same day. If Google and Yahoo reported normally, your record and mailbox are fine.
When to worry
| What you see | What it probably means |
|---|---|
| One missing Microsoft day, other days fine | Normal. Low volume that day, or a reporter-side delay. Wait. |
| Several days with no Microsoft report, Google and Yahoo arriving | Check whether you still send to Microsoft recipients, then the external authorization record if rua is on another domain. |
| No reports from anyone for 48 hours or more | Most likely something on your side: the DMARC record, the rua address or the receiving mailbox. |
| Never received a Microsoft report | Check volume first. Most of your Microsoft recipients may sit behind third-party filters. |
Before you change DNS, run through the record checks above. Editing a working record to fix a reporting gap can introduce a real problem.
Microsoft 365 also doesn't send forensic (ruf) reports. Its guide says so "even if a valid ruf=mailto: address exists", so a lack of Microsoft failure reports is expected. See DMARC failure reports for the receivers that do send them.
How DMARCdrift helps
Each domain gets its own inbound report address on DMARCdrift's domain, and DMARCdrift publishes the external authorization record for it, so you don't add a _report._dmarc record yourself. The domain's reports list shows the reporting organization and date range of every report it receives, so you can see at a glance whether Microsoft reported for a day that Google and Yahoo did.
If a domain that has already received reports gets none from any reporter for 48 hours, DMARCdrift sends a "No DMARC reports received" alert. That alert covers all reporters together; it does not fire when only Microsoft goes quiet. DMARCdrift also alerts when your DMARC record changes, which catches a broken or duplicated record soon after it is published.
If you run Microsoft 365 yourself, the Microsoft 365 DMARC setup guide covers the sending side.
DMARCdrift is free for one domain. Point your rua= address at us and see which reporters show up.
