How Do You Prevent Business Email Compromise From Lookalike Domains?
Published 2026-11-14 · Reviewed 2026-11-14 · 7 min read · By the Ingenious AI team
Is someone impersonating your domain right now?
Run the free DomainGuard scan: lookalike registrations, MX-armed variants, certificate-transparency and CAA checks for your domain, in one report.
Run my free DomainGuard scanThe short answer
You prevent business email compromise (BEC) from lookalike domains by removing the attacker's two advantages: invisibility and obedience. Invisibility ends with DMARC enforcement on your own domain, external-sender banners, gateway flagging of lookalike senders, and monitoring that finds cousin domains when they register. Obedience ends with process: no bank-detail change or urgent payment is ever actioned from email alone, and every such request is verified by a callback to a number you already hold. Technology shrinks the number of attacks that arrive; the callback rule stops the ones that do.
Why BEC works: the anatomy of the con
A BEC attack is a short play in three acts. Act one is registration: a domain one letter away from yours, or from your supplier's, costing a few dollars. Act two is patience: mailboxes are set up on the lookalike, and the attacker studies real invoice formats, signatures, and payment runs — often from a compromised mailbox or from documents the business itself published. Act three is timing: a plausible message lands at a busy moment, asking for a bank-detail change or an urgent payment, and it only needs one tired person to succeed.
Notice what is absent: malware, hacking, and technical wizardry. The entire attack rides on resemblance and routine. That is why purely technical defenses underperform against it, and why the businesses that stop BEC reliably are the ones that changed the routine — the moment a payment instruction must leave the email channel and be confirmed elsewhere, the attacker's perfect forgery becomes worthless paper.
Close the exact-domain door with DMARC enforcement
Half the BEC problem is spoofing of your real domain, and that half is fully solvable. With DMARC at p=reject, backed by aligned SPF and DKIM, a stranger cannot send mail as your exact domain and have it delivered by major receivers. Your own staff also gain protection: a forged 'CEO' message from your real domain stops arriving, so the internal-urgency variant of the con loses its best costume.
Enforcement does nothing about cousin domains, and pretending otherwise creates false confidence. The lookalike's mail passes its own SPF, DKIM, and DMARC because the attacker owns its DNS. So the correct mental model is two problems with two toolkits: exact-domain spoofing is a DNS project with a finish line, and lookalike deception is a detection-plus-process project that never fully ends. Budget and attention should be split accordingly, not poured entirely into the half that feels more technical.
Make lookalikes visible in the inbox
Your mail gateway is the one place every lookalike message must cross, so make it speak up. External-sender banners on all outside mail set a baseline of caution. Better gateways add lookalike awareness: flagging or prepending warnings when an external sender's domain is a near-match of your own or of a frequent correspondent's, and when a display name matches an executive but the address does not. These flags work because they interrupt the glance — the half-second of reading an inbox preview — where BEC is won and lost.
Pair visibility with detection upstream. Monitoring registrations and certificates for your brand's variants finds many lookalikes before the first fraudulent message, and MX records appearing on a fresh lookalike are the single best early warning that a BEC run is being prepared. DomainGuard alerts on exactly that combination — registered lookalike plus mail infrastructure — because in practice it is the smoke that precedes this specific fire by days.
The callback rule: the control that never fails
Every control above can be bypassed by a sufficiently patient attacker with a fresh domain. The callback rule cannot. Write it into policy in one sentence: bank details and payment instructions received by email are never actioned until confirmed by phone, on a number already on file — never a number in the message itself. Add a two-person rule above a sensible amount threshold, so no single employee can move serious money alone, however convincing the request.
Make compliance easy or it will not survive busy seasons. Give staff a short script for the call, a named list of who may approve changes, and explicit backing that delaying a payment for verification is always defensible while skipping verification never is. Then audit gently: a quarterly sample of bank-detail changes, each traced to its verification, keeps the habit alive better than any poster. Attackers test routines; routines that are actually practised are the ones that hold.
Rehearse the failure, not just the policy
Run one realistic simulation a year: a well-crafted lookalike message to your finance team, built with the same techniques a real attacker would use, followed by a blameless debrief. You are not testing whether individuals are clever; you are testing whether the system — banners, flags, callback habit — catches a good forgery on a normal Tuesday. Every gap the simulation finds in peacetime is a wire that does not leave in wartime.
Also prepare the response before you need it. If a fraudulent payment does go out, speed beats everything: banks can attempt recalls within hours that become impossible within days. Your incident plan should name who calls the bank, who preserves the email and headers, who notifies insurers, and who files with law enforcement — in that order, on one page, reachable when the person who wrote it is on holiday.
Five actions to take today
- Enforce DMARC at p=reject with aligned SPF and DKIM so your exact domain cannot be spoofed.
- Turn on external-sender banners and gateway flagging for sender domains that near-match yours.
- Monitor lookalike registrations and alert on any that publish MX records — the BEC preparation signal.
- Adopt the callback rule: bank-detail changes are confirmed by phone on a number already on file, never one from the message.
- Add a two-person approval rule above your payment threshold and rehearse the whole flow with one simulation a year.
Worked example: Mesa Verde Property Group stops a $212,000 diversion
Mesa Verde Property Group, a fictional property firm monitoring 8 domains, received a polished email from a one-letter lookalike of its main contractor, announcing new bank details ahead of a $212,000 progress payment. The message passed every authentication check — for the attacker's own domain. It failed the callback rule: the accounts clerk rang the contractor's number on file and heard, in one sentence, that no change had been requested. The lookalike had been registered 11 days earlier and had published MX records 6 days before the attempt — a signal the firm's monitoring had flagged and its gateway had bannered, giving the clerk two warnings before process delivered the third.
FAQ
Does DMARC stop BEC?
It stops the exact-domain half of it completely once enforced at p=reject. Lookalike-domain BEC passes the attacker's own authentication, so it additionally needs gateway flagging, monitoring, and the callback rule.
What is the single most effective BEC control?
The callback rule for bank-detail changes, verified on contact details you already hold. It defeats every variant of the con regardless of how convincing the email is, because it removes the decision from the compromised channel.
How do attackers choose which supplier to imitate?
From compromised mailboxes, published documents, social posts about partnerships, and ordinary observation of who a business pays. Any firm whose name appears on your invoices is a candidate costume.
Should we banner all external email?
Yes as a baseline, but banners alone fade into background noise. Add specific lookalike-domain flagging so the warning escalates exactly when the sender resembles you or your regular correspondents.
A payment already went out. What matters most?
Hours. Call the sending bank immediately to request a recall, preserve the message and headers untouched, then notify your insurer and file a law-enforcement report. Recall success drops steeply after the first day.