How to inventory sending sources
To inventory your sending sources, collect DMARC aggregate reports while you're at p=none, group the rows by DKIM domain and selector, SPF domain, and IP, work out which service each group is, and confirm an owner for each. Then make every legitimate source pass DMARC with aligned DKIM (and aligned SPF where it makes sense) before you move to quarantine or reject. Reports show IP addresses, domains, and results. They don't name services, so the inventory is the work of turning that data into a list of real systems and the people responsible for them.
It's easy to send from more places than you realize: the mailbox platform, a marketing tool, a CRM, a help desk, invoicing, web forms, scanners, a partner. Domains start at monitoring for exactly this reason: a domain owner can overlook a server here or be unaware of a third-party sending agreement there.
Before you start
- Reports must be arriving. You need a DMARC record with a
ruatag, and reports landing somewhere you can read them. If the mailbox is empty, see Why DMARC reports do not arrive. - Give it time. Depending on how often you send mail, it may take many months of consuming DMARC aggregate reports before you're sure all your mail is authenticated. A quarterly invoice run or an annual renewal campaign only shows up when it sends.
- Cover every domain you send from. That includes subdomains and parked domains. A subdomain with its own DMARC record gets its own reports, at its own
rua, so check those too.
Step 1: Collect and group the data
Each record in an aggregate report gives you the source_ip, a message count, the header_from domain, often the envelope_from (SPF) domain, the DKIM d= domain and selector, the raw SPF and DKIM results, and the aligned DMARC results. The fields are explained in A field-by-field walk-through of a DMARC aggregate report.
Group rows by the most stable identifier first:
- DKIM
d=domain plus selector. The current report format requires the selector, and selectors can point to the platform that signed the mail. envelope_from/ SPF domain. Some vendors use their own bounce domain.- IP ranges. Use these last. Many services share cloud infrastructure, so an IP on its own is weak evidence.
That order is our method, not a formal procedure. Whatever you use, don't do it by hand for long. Reports are meant to be machine-parsed. Many third-party services exist that will process DMARC aggregate reports, or you can create your own tools. DMARC analysis tools exist to do this grouping for you.
Step 2: Identify each source
Use several checks together. These are practitioner methods, not guarantees:
- The DKIM domain and selector. Is it a vendor's own domain, or a selector named after a platform?
- The SPF domain. A vendor bounce domain, like
bounces.vendor.example, tells you who's sending. - Reverse DNS and ownership lookups for the IP.
- Your own records. Finance knows which SaaS tools you pay for. Marketing, sales ops, the help desk, and IT know what they've signed up for. Your DNS history shows which SPF includes and DKIM records were added, and when.
- Ask. The goal is to find the mail created by an agent of the domain owner. Someone in the organization signed up for each legitimate sender. Find them. Who usually owns what is in Who usually owns SPF, DKIM, and DMARC.
Step 3: Sort each source into a bucket
- Legitimate and aligned. Done. Keep watching it.
- Legitimate but not aligned. These must be fixed before enforcement. Those shortcomings must be addressed before any attempt to publish an enforcement policy. The usual causes are a vendor using its own bounce domain, so SPF doesn't align, or signing with its own
d=, so DKIM doesn't align (Third-party mail fails alignment). - Indirect flows. Forwarders, aliases, and mailing lists. SPF failures are expected here. DKIM may survive (Forwarded mail fails DMARC).
- Unknown. Possibly spoofing, possibly a legitimate sender nobody has claimed yet. Keep investigating.
How to tell these apart from the report data is in Legitimate mail vs spoofing in reports, and what to do about rows you can't place is in Reports show sources we do not recognize.
Step 4: Authorize the legitimate ones properly
- Prefer aligned DKIM from each vendor, with
d=set to your domain or a subdomain of it. For domains heading to reject: they must not rely solely on SPF and must apply valid DKIM signatures (SPF include vs DKIM from the vendor, DKIM selectors and DNS records). - Add SPF only where it's needed. Avoid extremely broad records containing many extensive network ranges, because overly permissive SPF can let mail you didn't send pass DMARC. Watch the 10-lookup limit (The 10-lookup limit and SPF permerror).
- Consider a subdomain per stream, such as
mail.example.comfor marketing. Each stream then gets its own policy and its own reports (Shared domains and multi-service setups).
Vendor setup itself is covered in How to add a third-party sender without breaking DMARC. Alignment basics are in Identifier alignment for SPF and DKIM.
Step 5: Keep the register alive
Keep a living list: source, owner, purpose, how it authenticates (SPF, DKIM, or both), whether it aligns, when it was last seen, and roughly how much it sends. That's a working practice, not a standard, but it's what makes the enforcement decision defensible.
New senders appear whenever someone signs up for a tool, so keep reports flowing after enforcement. Microsoft's guidance is to "Review aggregate reports weekly during initial deployment; monthly once stable." You're close to ready for the next step when your legitimate, high-volume sources pass aligned and the remaining failures are understood (When you are ready for p=quarantine). The full sequence is in The right order: SPF, DKIM, DMARC, then BIMI.
What the inventory will miss
Even a careful inventory has gaps you should know about:
- Mail to receivers that don't send reports. Sending reports is a SHOULD, and some receivers opt out. Microsoft 365, for example, only sends aggregate reports when the receiving domain's MX points directly to Microsoft 365.
- Mail rejected early. Messages rejected early in the SMTP transaction, for example by SPF
-all, will never appear in aggregate DMARC reports. - The original source behind a forwarder. You see the forwarder's IP.
So "no failures in the reports" doesn't prove the inventory is complete.
Mistakes worth avoiding
- Only looking at failures. Passing sources belong on the list too. They can break later.
- Grouping by IP alone, so one shared cloud range gets pinned on the wrong service.
- Forgetting low-frequency senders, such as quarterly invoices and annual renewals.
- Forgetting devices and apps that send directly: printers, scanners, website forms, monitoring alerts.
- Checking the main domain but not subdomains that have their own records.
- Adding every unknown IP to SPF to make the failures go away. That can authorize a spoofer.
- Moving to enforcement with "unknown but probably fine" high-volume sources.
Free DMARC account
Turn raw DMARC reports into a readable dashboard
Point your rua address to our platform and read the results by sender and source instead of opening XML files.
- Results grouped by sending source
- SPF, DKIM and alignment results per sender
- The original XML reports are still there to download
FAQ
Who is sending email from my domain?
Your aggregate reports show every source that sent mail as your domain to receivers that report, as IPs and domains. Group them, identify each service, and confirm an owner. Expect gaps for receivers that don't report.
How long should we collect reports before enforcing?
Long enough to see every sending cycle you have. It may take many months, depending on how often you send.
A report shows an IP I don't recognize. Is it an attacker?
Not necessarily. Check the DKIM domain and selector, the SPF domain, and ask around before deciding. See Legitimate mail vs spoofing in reports and Reports show sources we do not recognize.
Should I add every vendor to SPF?
Prefer aligned DKIM from each vendor, add SPF only where it's needed, and keep the record tight.
Do we need to inventory subdomains?
Yes, every domain you send from. Subdomains with their own DMARC record report separately.