Explore our new Cybersecurity Terms Glossary, explained by our experts.

View the Glossary View the Glossary
PRIVIA

Cybersecurity Terms Glossary

How Do You Investigate a Suspicious Email?

Investigating a reported email means acquiring it safely, reading the headers, assessing the links and attachments, and then finding out who else received it and who acted on it.

Illustration for the cybersecurity term Suspicious Email Investigation
Home / Blog / Cybersecurity Terms / How Do You Investigate a Suspicious Email?
17 November 2026

How Do You Investigate a Suspicious Email?

How Do You Investigate a Suspicious Email?

A user reports an email. The clock starts, because if it is a real campaign then other people received it, some of them clicked, and one of them typed their password. The investigation has four jobs in sequence: get the message safely, decide whether it is malicious, find everyone else who got it, and establish what they did about it. The order matters, and most of the damage happens when teams skip straight to the third question.

Acquire It Without Destroying It

A forwarded copy is not evidence. Forwarding rewrites the headers, strips attachments, and may alter the body, which leaves you analysing a different message from the one that arrived. Collect the original as a .eml or .msg file, through the platform’s report button, an attached copy, or an administrative search of the mailbox.

Everything from here happens somewhere that is not your corporate network: an isolated analysis environment, with no access to internal systems and no credentials worth stealing. Opening an attachment on a production workstation to see what it does is how investigations become incidents.

Read the Headers

The headers carry most of the technical verdict. Work through them in a fixed order.

The detail of this stage is covered in email header analysis, and the techniques being detected in email spoofing.

Links and Attachments

Unwrap any link-rewriting applied by your own email security product first, or you will analyse the gateway rather than the attacker. Then check the domain’s registration age, its hosting, and whether anyone else in the estate has ever contacted it. A domain registered four days ago, resolving to shared hosting, contacted by nobody until this morning, needs no reputation feed to be suspicious.

Submit URLs to a sandbox or an online scanner rather than visiting them, and be aware that many phishing kits detect analysis infrastructure and serve a benign page to it. A clean result from an automated scanner is weak evidence, not an acquittal.

For attachments, hash first and check whether the sample is already known. Then triage statically: macros and embedded objects in documents, JavaScript in archives and ISO files, and shortcut files with hidden command lines. Detonate in a sandbox and watch what it does rather than what it is — network callbacks, persistence, and child processes tell you more than any signature.

Scope the Campaign

This is the part that changes the outcome. Search the whole mail platform for related messages, matching on sender address, sending infrastructure, subject, embedded URL, and attachment hash rather than any single one of them, because attackers vary the subject and the display name across recipients.

Then answer two questions for every recipient:

  1. Did they click? Query proxy logs, DNS records, and the email product’s own click-tracking for each URL in the campaign.
  2. Did they authenticate afterwards? Check identity sign-in logs for anyone who clicked: a successful sign-in from unfamiliar infrastructure shortly after the click means credentials were harvested and used, and the incident has changed shape.

For any account that shows signs of compromise, check immediately for a new mailbox forwarding rule, a new OAuth application consent, and a newly registered authentication method. Those three are how access is retained after a password reset.

Contain and Close the Loop

Purge the campaign from every mailbox, block the sender, domain, and URLs at the gateway and the proxy, and revoke sessions and reset credentials for anyone who entered them. Record the indicators somewhere reusable rather than in the case notes alone, and feed the detection gap back into your rule set if the message got through a control that should have stopped it.

Then reply to the person who reported it. A short message telling them what it was and thanking them is the cheapest investment available in future reporting rates, and the volume of reports is what makes the whole process work. UK organisations can also forward phishing messages to the NCSC’s Suspicious Email Reporting Service at [email protected], which has handled many millions of reports since its launch in April 2020 and results in malicious sites being taken down; where money has been lost, the incident should also go to Action Fraud.

Because the steps are so repetitive, this is the first workflow most teams automate with a SOAR playbook. At Privia Security, we help organisations build and test this process through our SOC maturation services.

FAQ – Frequently Asked Questions

Should I Open the Email to Investigate It?

Not on a corporate machine. Use an isolated analysis environment with no access to internal systems or credentials, and examine the message as a file rather than in a mail client where remote content loads automatically. Even opening a message can confirm the address is live and trigger tracking. Attachments should only ever be executed in a sandbox.

What if the Sandbox Says the Link Is Clean?

Treat that as weak evidence. Phishing kits routinely detect analysis infrastructure by IP range, user agent, or geography and serve a harmless page to it, and many are configured to be live for only a few hours. Judge the message on the whole picture: domain age, authentication results, context, and whether the request it makes would ever be legitimate.

How Do I Find Out Who Clicked?

Correlate the campaign URLs against proxy logs, DNS query records, and your email product’s own click telemetry, filtered to the recipient list. Then check identity sign-in logs for those users around and after the click time. A successful authentication from unfamiliar infrastructure shortly after a click is the signal that credentials were captured and used.

What Should We Do After Containing a Phishing Campaign?

Three things. Record the indicators somewhere reusable so the next campaign from the same actor is recognised faster. Establish why the message was not blocked and whether a control needs tuning. Finally, tell the reporting user what happened, because report rates depend entirely on people believing their reports are acted on.

Related Terms

You May Be Interested In These

What Is C2PA Content Provenance?

Cybersecurity Terms

17 November 2026

What Is C2PA Content Provenance?

Read More
What Is Cobalt Strike?

Cybersecurity Terms

17 November 2026

What Is Cobalt Strike?

Read More