Insights

Threat hunting: what your security provider should investigate

9 November 2026

Understand threat hunting, how it differs from alert monitoring, and what to ask your provider about telemetry, scope, escalation and evidence.

Huntress has published an explainer on threat hunting, outlining proactive investigation and the role of human analysts alongside automated detection. For Australian small and medium businesses, it raises a practical question: does your security service investigate suspicious activity beyond the alerts its tools generate?

The distinction matters when you are comparing providers or reviewing an existing agreement. A dashboard showing no outstanding alerts does not establish that someone has searched for less obvious signs of compromise. Ask what is investigated, what information is available and how the work is documented.

How threat hunting differs from alert monitoring

Alert monitoring starts with a notification. A security tool identifies a condition, such as a suspicious program or unusual sign-in, and an analyst assesses whether it requires action. Good monitoring is essential, but its starting point is something a detection rule or model has already recognised.

Threat hunting starts with a question. An investigator might ask whether an attacker is using legitimate remote-access software on unmanaged devices, or whether compromised accounts have created forwarding rules that existing alerts missed. The investigation searches relevant records, tests explanations and follows evidence across systems.

The activities overlap: an alert can prompt a wider hunt, and a hunt can produce a new detection rule. Neither replaces patching, access controls, backups or incident response.

A useful distinction is the deliverable. Monitoring should produce assessed alerts and response records. A hunt should also produce a documented question, search scope, findings and limitations—even when nothing suspicious is found.

Check the telemetry before buying the promise

Telemetry means the activity records investigators can search. The quality and reach of those records determine which questions they can answer.

Ask your provider to map the available evidence against your actual environment:

| Area | Records worth checking | Limitation to clarify | | --- | --- | --- | | Computers and servers | Process activity, command lines, network connections and security events | Which devices are enrolled and reporting? | | Microsoft 365 and identity | Sign-ins, audit events, mailbox changes and application consent | Which events are collected under your licensing and configuration? | | Network and remote access | Firewall, VPN and DNS records | Are relevant sites and remote connections visible? | | Business applications | Administrative changes, access events and data exports | Can investigators access useful logs from each application? |

Confirm retention for each source, not just a platform-wide headline. Records retained in their original system may not be searchable in the provider's investigation tools. Ask how far back an analyst can search and whether collection failures generate notifications.

Also clarify who can access logs, where they are stored and how sensitive information is protected. More collection is not automatically better: it should serve defined security purposes with appropriate access and retention controls.

Define the investigation scope and method

A service described as including threat hunting may mean different things. It could involve scheduled investigations, searches prompted by new threat intelligence, or additional analysis only after an alert. Ask which activities your agreement actually includes.

Three useful starting points are:

  • A testable hypothesis: Could a compromised finance account have created a hidden mail-forwarding path?
  • Threat intelligence: Does evidence associated with a newly reported attack appear in your environment?
  • An unusual pattern: Why has a previously quiet account started accessing multiple systems outside its normal working pattern?

Searching for known malicious addresses or file identifiers can be useful, but those details change. Behaviour-focused investigations can also look for sequences of actions, such as unexpected remote access followed by credential changes and access to shared files.

Request a written scope covering systems, accounts, time windows, exclusions and investigation frequency. Confirm whether the work examines your specific environment or consists mainly of broad searches across the provider's customer base. Both can contribute, but they are not interchangeable.

What a completed hunt should look like

Consider a hypothetical investigation into unauthorised remote-management software. The objective is not simply to list installed applications. It is to determine whether remote access exists outside approved support arrangements.

A defensible workflow would include:

  • Establish the approved baseline. Identify authorised tools, administrators and support suppliers.
  • Search the available evidence. Examine software inventories, service creation, execution records and relevant network activity.
  • Investigate exceptions. Compare unexpected installations with change records and device ownership.
  • Expand when justified. Check whether the same account, tool or connection appears elsewhere.
  • Record the outcome. Document confirmed explanations, unresolved questions, coverage gaps and recommended actions.

An unfamiliar tool is not automatically malicious. It might be a legitimate supplier installation that was poorly documented. Human investigation matters because business context can distinguish an administrative mistake from an intrusion.

Our remote-access tool checklist covers related controls. For identity investigations, the Microsoft 365 token replay guide explains why device evidence, authentication and mailbox activity need to be considered together.

Agree escalation before an investigation finds trouble

Finding suspicious activity is only useful if someone can act on it. Ask your provider to explain the transition from a hunt to an incident, including the threshold for escalation and who takes ownership.

Your agreement should answer:

  • Who receives urgent notifications, including outside business hours?
  • Who can authorise device isolation, account suspension or session revocation?
  • Which containment actions are pre-authorised, and which require explicit approval?
  • Who investigates business impact and coordinates recovery?
  • What happens if the nominated business contact cannot be reached?

Separate investigation from remediation in the commercial scope. A provider might identify a problem without including the work needed to remove it, restore affected systems or support external reporting.

Agree how evidence will be preserved, including relevant timestamps, exported records and an action history. For preparation beyond the hunt itself, review our ransomware response checklist.

Ask for evidence of completed threat hunting

You should not need to read every technical query, but you should receive enough information to understand what was tested and what remains unknown.

Request a redacted sample report before signing. A useful report includes the investigation date, question, systems and time period searched, evidence sources, findings, limitations and follow-up owners. Where practical, it should reference supporting records or saved searches so another analyst can review the reasoning.

“No evidence found” is a bounded conclusion—not proof that the business is free of compromise. Missing device coverage or expired logs should be stated plainly, rather than buried behind a reassuring summary.

Measure progress through resolved visibility gaps, completed follow-up actions and useful detection improvements. Counting hunts alone says little about their depth or relevance.

Start with a focused security review

Threat hunting is most useful when it answers questions that matter to your business and sits alongside reliable monitoring and response. Before adding a service, establish what your provider can see, what it will investigate and what happens when it finds something concerning.

Tech Engine can help you review your cyber security, including logging coverage, access controls and response responsibilities. Contact sales@techengine.au or call 1300 088 324 to discuss your current arrangements.

Want this applied to your business?

Request an AI Blueprint and we will map the processes worth automating first.

Request an AI Blueprint