An email arrives from your conveyancer, your supplier or your own accounts department: the banking details have changed, please pay the new account. It looks right. The sender's address looks right. The money goes to a fraudster. Payment-redirection fraud targets the moment money moves, which is why firms that handle client funds, such as attorneys, conveyancers, estate agents and accountants, are attractive targets.
How payment redirection works
There are three common ways to send an email that seems to come from your firm:
- Spoofing your exact domain. The From line shows yourfirm.co.za, but the message came from a server you do not control. SPF, DKIM and DMARC are designed to stop this.
- A lookalike domain, such as yourfirm-za.co.za or a name with one letter swapped. Your DNS records cannot stop this. Careful staff and clients can.
- A real, compromised mailbox. The fraudster has a staff member's password and sends from the genuine account. Multi-factor sign-in and a payment verification rule are the defence.
The FBI's Internet Crime Complaint Center describes business email compromise as a scam that targets people who perform legitimate transfer-of-funds requests, and advises using a second channel to verify any request to change account information. That one habit protects against all three routes.
SPF: who may send
SPF (Sender Policy Framework) is a text record in your domain's DNS that lists the servers allowed to send email for your domain. That includes your email provider and any other service that sends as you, such as accounting software that emails invoices or your website's contact form. A receiving server checks the sending server against the list.
Two rules catch firms out. A domain must have only one SPF record, so every service goes into the same record. And the standard caps an SPF check at 10 DNS lookups, so a record with too many "include" entries fails. SPF also has a gap: it checks the hidden envelope sender, not the From address a person sees.
DKIM: a signature on every message
DKIM (DomainKeys Identified Mail) adds a digital signature to each message your mail server sends. The matching public key is published in your DNS. The receiving server checks the signature, which shows that the message was signed for your domain and was not changed on the way. Your email provider generates the keys. You publish the record they give you and switch signing on, for each service that sends as you.
DMARC: the instruction and the report
DMARC ties the two together. It requires that SPF or DKIM passes for the same domain that appears in the visible From line, which closes SPF's gap. Your DMARC record, published at _dmarc.yourfirm.co.za, tells receiving servers what to do with mail that fails:
- p=none: deliver as normal, but report.
- p=quarantine: treat it as suspicious, usually the spam folder.
- p=reject: refuse it.
DMARC also asks receivers to send reports to an address you choose, so you can see every server that sends mail claiming to be you.
Check your own records
You can look without signing in to anything. In a Windows command prompt, or a Mac or Linux terminal, run:
nslookup -type=txt yourfirm.co.zaand look for a line starting v=spf1.nslookup -type=txt _dmarc.yourfirm.co.zaand look for v=DMARC1 and the p= value.
The DKIM record's name depends on your email provider, so the simplest check is the provider's admin console, which shows whether DKIM signing is on. Warning signs:
- No SPF record, or two of them.
- An SPF record containing +all, which lets any server pass.
- No DMARC record, or p=none left in place for years with nobody reading the reports.
- DKIM switched off, or set up for your mailboxes but not for the software that sends your invoices.
A safe rollout: monitor, quarantine, reject
Jumping straight to reject can block your own invoices if a sending service was missed. Google's recommended rollout and the UK National Cyber Security Centre's guidance both move in stages:
- List every service that sends email as your domain. Ask accounts, reception and marketing, not only whoever manages IT.
- Publish one SPF record that covers them all, and switch on DKIM for each service that supports it. Google recommends doing this at least 48 hours before adding DMARC.
- Publish DMARC at p=none with a reporting address, and read the reports. Google suggests at least a week. We prefer a full month, so a month-end invoicing run is included.
- Fix any legitimate service that fails. Then move to p=quarantine for a small percentage of failing mail, using the pct tag, and raise it over time.
- When the reports show that only mail you did not send is failing, move to p=reject.
- Keep reading the reports. Add any new service to SPF and DKIM before it starts sending.
What this does not fix
DMARC at reject stops exact-domain spoofing at receiving servers that check it. It does not stop lookalike domains or a fraudster inside a real mailbox. Keep three habits alongside it: multi-factor sign-in on every mailbox, a standing rule that any change to banking details is confirmed by phone on a number you already had, and a line on invoices and engagement letters telling clients that your firm will never change its banking details by email.
Next step: a free Exposure Check looks at your domain's email set-up from the outside, along with what is public about your partners. If your records need work, a Website Security audit checks whether someone can send email that looks like yours, and hardening puts the protection in place, then retests.
Sources and further reading
- Business Email Compromise (public service announcement I-091124-PSA, 11 September 2024), FBI Internet Crime Complaint Center. Describes business email compromise and advises verifying account-change requests through a second channel.
- RFC 7208: Sender Policy Framework (SPF), IETF. The SPF standard, including the single-record rule, the 10-lookup limit and what +all does.
- RFC 6376: DomainKeys Identified Mail (DKIM) Signatures, IETF. The DKIM signing standard.
- RFC 7489: Domain-based Message Authentication, Reporting, and Conformance (DMARC), IETF. DMARC policies, reporting, and alignment with the visible From domain.
- Recommended DMARC rollout, Google Workspace Admin Help. The staged move from none to quarantine to reject, the 48-hour lead for SPF and DKIM, and the one-week monitoring period.
- Email security and anti-spoofing, National Cyber Security Centre (UK). Plain-language summary of SPF, DKIM and DMARC, and the staged path from a none policy to rejecting spoofed email.
- Set up DMARC to validate the From address domain, Microsoft Learn. What none, quarantine and reject mean in practice, the pct tag and the value of reading aggregate reports.