Phishing Case: answer key
Every question with its answer and the reasoning. Work the drill first, or the key only tells you what, not how.
-
Read the authentication results and pull out the three values a verdict turns on.
Authentication-Results: mx.example.org; spf=pass (sender IP is 203.0.113.47) smtp.mailfrom=bounce.marketing-host.example.net; dkim=pass header.d=marketing-host.example.net; dmarc=fail action=quarantine header.from=your-bank.example.com Return-Path: <bounce-4471@bounce.marketing-host.example.net> From: "Your Bank Security" <security@your-bank.example.com> Reply-To: security-team@mail-verify.example.netAnswer: SPF result: pass · DKIM signing domain (header.d): marketing-host.example.net · DMARC result: fail
SPF passed and DKIM passed, and the message is still failing DMARC. That is not a contradiction, it is the entire reason DMARC exists. SPF checks the envelope sender (smtp.mailfrom), DKIM checks whoever signed the message, and neither of them looks at the From: header the recipient actually sees. DMARC is the check that requires one of the other two to ALIGN with the visible From: domain, and here both authenticate marketing-host.example.net while the From: says your-bank.example.com. "SPF passed so it is legitimate" is the single most common misreading of an email header.
Read more: Email Header Analyzer
-
What does an SPF pass actually establish?
Answer: That the sending IP is authorized to send for the envelope sender domain, which the recipient never sees
SPF is a check on the MAIL FROM envelope address and the connecting IP. The address it validates is in the Return-Path, not the From: header shown in the mail client, and an attacker controls both independently. This is why a message can pass SPF cleanly while displaying any From: address it likes. Integrity of the content is DKIM's job; tying either result to the visible From: is DMARC's.
Read more: Email Header Analyzer
-
Trace the Received chain. Pull out the address the message actually entered your infrastructure from, and the hop that is lying.
Received: from mx.example.org (mx.example.org [198.51.100.10]) by inbound.example.org with ESMTPS; Tue, 3 Mar 2026 09:14:22 +0000 Received: from smtp-out.sender-host.example.net (smtp-out.sender-host.example.net [203.0.113.88]) by mx.example.org with ESMTPS; Tue, 3 Mar 2026 09:14:19 +0000 Received: from your-bank.example.com (unknown [192.0.2.201]) by smtp-out.sender-host.example.net with SMTP; Tue, 3 Mar 2026 09:14:11 +0000Answer: IP of the hop that handed the message to mx.example.org: 203.0.113.88 · IP claiming to be your-bank.example.com: 192.0.2.201 · Reverse-DNS result for that claiming host: unknown
Received headers are prepended, so the chain reads bottom-to-top in chronological order and top-to-bottom in order of trustworthiness. Only the hops your own infrastructure added can be trusted. Everything below them was written by systems you do not control, and an attacker can forge as many as they like. The bottom hop here announces itself as your-bank.example.com while its reverse DNS came back unknown, and the receiving server recorded both. That gap between the claimed name and the resolved one is the finding.
Read more: Email Header Analyzer
-
Pull out the three addresses a reply would actually involve.
From: "IT Service Desk" <helpdesk@example.com> Reply-To: it-support-desk@secure-helpdesk.example.net Return-Path: <noreply-7712@bulk.sender-host.example.net> To: staff-list@example.com Subject: Action required: your mailbox will be disabled
Answer: Address shown to the reader: helpdesk@example.com · Address a reply is sent to: it-support-desk@secure-helpdesk.example.net · Address bounces go to: noreply-7712@bulk.sender-host.example.net
Three headers, three different domains, and a mail client shows the reader only the first one, often only the display name ("IT Service Desk") with the address hidden behind it. Reply-To redirects the conversation to infrastructure the attacker controls without needing to compromise anything, and it costs nothing. A Reply-To on a different domain from the From: is not proof of anything on its own, because legitimate bulk senders do it constantly, but combined with an urgency subject it is worth a second look.
Read more: Email Header Analyzer
-
Match each question to the header that answers it.
Answer: Did the sending IP have permission to send for the envelope domain? → Authentication-Results · Which IP handed the message to my own mail server? → Received · Where do bounces go? → Return-Path · Can I find this exact message in another mailbox? → Message-ID · How long did the message spend in transit? → Received · Did anything the recipient sees actually authenticate? → Authentication-Results
Message-ID is the underused one. It is generated by the sending system and travels with the message, so it is the key for finding the same message across every mailbox that received it. That is how you scope a campaign rather than investigating one report. It is also attacker-controlled, so a shared Message-ID across recipients is a strong clustering signal and a weak authenticity one. Transit time comes from the timestamps on consecutive Received hops, and a large gap sometimes means a message sat in a queue on infrastructure worth looking at.
Read more: Email Header Analyzer
-
A message from a supplier passes SPF, DKIM and DMARC, all aligned to the supplier's real domain. It asks to change the bank details on an invoice. What is the most likely explanation?
Answer: The supplier's mailbox is compromised, and the message is genuinely from their infrastructure
This is the case email authentication cannot help with, and it is what most real business email compromise looks like. SPF, DKIM and DMARC establish that a message came from where it claims to have come from. They say nothing about whether the person at that keyboard is the account owner. When everything passes and the request is still wrong, the header analysis is finished and the investigation moves to the sender's sign-in logs, mailbox rules and forwarding configuration.
Read more: Email Header Analyzer
-
Pull out what DMARC decided, and what the domain owner asked to happen.
Authentication-Results: mx.example.org; spf=softfail smtp.mailfrom=example.com; dkim=none; dmarc=fail (p=none sp=none dis=none) header.from=example.com Received-SPF: softfail (example.com: domain of transitioning example.com does not designate 192.0.2.77 as permitted sender)Answer: DMARC result: fail · Published DMARC policy: none · Disposition actually applied: none · IP that failed SPF: 192.0.2.77
The check failed and the message was delivered anyway, because the domain owner published p=none. That policy means "report to me, do not act", and a large share of domains never move past it. So a DMARC failure in a header is not evidence the mail system did anything about it, and the message landing in an inbox is not evidence it passed. Read the result and the disposition as two separate facts. The softfail wording is the other tell: the SPF record ends in ~all rather than -all, which is the same reluctance to enforce.
Read more: Email Header Analyzer
-
Header analysis is done and the message is malicious. What is the highest-value next step for scoping?
Answer: Search every mailbox for the Message-ID and the sending IP, to find who else received it
All four are reasonable actions and three of them are containment or enrichment. Scoping comes first, because the answer changes what containment even needs to cover: one recipient is an incident, two hundred is a campaign, and finding out which takes one search. Blocking before scoping also destroys the search, since a blocked message may never have been delivered to the mailboxes you were about to look in.
Read more: Email Header Analyzer