Network Segmentation: Splitting an Organization into Zones
16 November 2026Multi-Site Network: Connecting Several Offices with the Same Rules Everywhere
16 November 2026An email that looks like it comes from your organization does not necessarily come from it. The Canadian Centre for
Cyber Security puts it simply: "remember, addresses can be spoofed"
(ITSAP.00.101). Three records
in your domain’s DNS zone let the servers that receive your email tell the real messages from the fake ones.
What each one does
According to Microsoft
(How email authentication works in Microsoft 365):
- SPF "specifies the source email servers that are authorized to send mail for the domain";
- DKIM "uses a domain to digitally sign important elements of the message to ensure the message remains unaltered
in transit"; - DMARC "specifies the action for messages that fail SPF or DKIM checks for senders in the domain, and specifies
where to send the DMARC results (reporting)".
Microsoft is blunt: "Anything less than all of the email authentication methods results in substandard protection."
In which order
The two sources do not give the same order, and each has its reason:
- Microsoft: "Before you configure DMARC for custom domains or subdomains, you need to create SPF TXT records and
configure DKIM signing for all custom domains and subdomains you use to send email in Microsoft 365"
(Set up DMARC). - The Canadian Centre: "By deploying an initial DMARC record with a policy of none, you can immediately begin to
receive DMARC reports, even before deploying SPF or DKIM"; then deploy SPF and DKIM; then enforce
(ITSP.40.065).
Both agree on what matters: DMARC starts at "none", and nothing is blocked before the reports have been read.
From none to reject
"The goal is to get to a p=reject DMARC policy for all of your custom domains and subdomains", says Microsoft
(Set up DMARC). The
Canadian Centre asks organizations to "gradually increase the strictness of enforcement for a domain from a baseline of
none, through quarantine, to an end state of reject", and notes that "only a policy of reject at 100% will prevent all
illegitimate messages from being delivered"
(ITSP.40.065).
Between the steps, you read the reports.
The reports: what they show
"DMARC aggregate reports are produced by receiving systems and are typically sent to domain owners daily", says the
Canadian Centre. They show where the email carrying your domain comes from, and help "identify infrastructure used by
malicious actors to impersonate your domains"
(ITSP.40.065). Microsoft asks to
review them regularly "to check for unintentional DMARC failures (false positives)", which would block real email
(Set up DMARC).
False positives almost always come from a forgotten sender: the billing software, the newsletter tool, the copier that
emails scans. Each one has to be added to SPF or sign with DKIM before moving to reject.
Domains that send nothing
A domain bought to protect a name, or kept from an old brand, can be used to send fake email. For these parked domains,
Microsoft says to "configure the DMARC TXT records in those domains to specify no email should ever come from those
domains" (Set up DMARC).
The Canadian Centre adds: "The recommended SPF record for non-mail domains is: v=spf1 -all"
(ITSP.40.065).
What Gmail and Yahoo already require
"Starting February 1, 2024, email senders who send more than 5,000 messages per day to Gmail accounts must meet the
requirements", Google says, including DMARC
(Google). Yahoo applies similar rules from the same month
(Yahoo). An organization that sends less is not covered by those
thresholds, but its email gets through better when all three are in place.
Frequently asked questions
Are SPF, DKIM and DMARC all needed, or is one enough?
All three. According to Microsoft, they work together, and anything less than all of them results in substandard
protection.
Can a DMARC reject policy block our own email?
Yes, if a service that sends on your behalf is not listed in SPF or does not sign with DKIM. That is why you start at
none and read the reports before moving to quarantine, then reject.
What should be done with a domain that never sends email?
Publish a DMARC reject policy and the SPF record v=spf1 -all, as Microsoft and the Canadian Centre recommend, so
nobody can use it as a sender.
To go further
Our Cloud and Microsoft 365 page presents what we configure. A fake email is
often after a password: our article on mandatory multifactor authentication
explains the second barrier, and the one on the first hour of a ransomware attack
says what to do when an attack succeeds.
What now?
MPJM’s monthly security review reads your DMARC reports and your
Microsoft 365 accounts, and the monthly block of hours reserves the time
to move from none to reject without blocking your real email. To start, the free network analysis
looks at your Microsoft 365 accounts and what protects your domain. It is offered to organizations with 40 or more
employees.
P.S. Sixteen questions, eight domains, a clear picture of your strengths and blind spots: start the assessment.
