An incident can lose valuable minutes while everyone decides who owns the next action. One analyst checks the endpoint, another repeats the same search, and an affected account remains active. A rehearsed division of work gives the team a way to start investigating together.
I propose using Immediate Action Drills (IADs) for that initial response. An IAD is a short, predefined procedure triggered by a validated threat. The four-person model below is a starting point to test in an enterprise security operations center (SOC), rather than a claim that every incident needs the same staffing.
Assign work before the incident
The useful part of the military drill analogy is rehearsal. People know their responsibilities and can act without waiting for a complete explanation of the situation. Cyber response still requires judgment about service availability, evidence preservation, and the scope of each responder’s authority.
| Role | Initial responsibility |
|---|---|
| Endpoint lead | Examine affected hosts, collect evidence, and coordinate isolation |
| Network analyst | Review traffic, identify suspicious connections, and trace possible lateral movement |
| Identity defender | Investigate accounts, sessions, authentication events, and privilege changes |
| Response lead | Validate the trigger, coordinate actions, maintain the incident record, and communicate with stakeholders |
In a smaller team, one person may cover several roles. Record that limitation explicitly so that an unassigned task does not disappear from the response.
Define the trigger and authority
Start the drill when triage establishes a credible incident or an authorized responder approves escalation. An alert score alone should not grant unrestricted authority to interrupt business operations.
The response lead announces the incident identifier, affected assets, known evidence, and immediate objective. “Contain suspected credential theft affecting this host and account” gives the team something concrete to work toward. The playbook should specify who may isolate a host, disable an account, or interrupt a production service.
Investigate across the affected systems
The endpoint lead reviews the endpoint detection and response (EDR) timeline, identifies relevant processes, and decides whether isolation is justified. Capture volatile evidence when feasible, but do not let collection requirements prevent urgent containment. Record actions that could change or destroy evidence.
The network analyst reviews packet captures or flow records around the affected host. The immediate questions are whether it contacted external command infrastructure and whether the same activity reached other systems. A suspicious connection is a lead to investigate; it does not establish that every connected system is compromised.
The identity defender examines sign-ins, multifactor authentication events, session activity, and privilege changes. Where compromise is supported, coordinate account restrictions and session revocation. A password reset alone may leave an existing session usable.
These tasks overlap. A process discovered on an endpoint may explain a network connection, while an account event may identify how the attacker reached that endpoint. The response lead keeps those findings connected.
Report observations in a consistent format
Use a dedicated incident channel and record the asset, observation, action, timestamp, and remaining uncertainty. For example: “Host123 isolated at 14:32 UTC; suspicious process confirmed; memory collection pending.”
Choose a check-in interval suited to the incident. Five minutes may work during initial containment, but responders should report consequential findings immediately. If the team uses red, amber, and green status labels, define exactly what each means. A color cannot substitute for the evidence behind it.
Rehearse and measure
A monthly exercise is a reasonable starting proposal. Rotate roles and test different entry points, including phishing, stolen credentials, and lateral movement. Include a scenario in which isolation would interrupt an important service so that the team has to exercise its approval process.
Measure time to assignment, time to justified containment, duplicate work, missed tasks, and evidence lost during response. Review whether the four-person structure helped under the conditions tested.
The purpose of an IAD is to make the first coordinated actions repeatable. Start with one incident type, rehearse it, and change the drill wherever the exercise exposes confusion.
Share this article

yankee0one
Cybersecurity leader specializing in incident response, threat detection, and operations.
Related Articles

AI in the SOC: A Practical Plan for Analyst Collaboration
A phased approach to AI-assisted security operations, with defined responsibilities, research limitations, and measures for evaluating a pilot.
Read more
Leading Through Uncertainty in a Cyber Crisis
How incident leaders can make decisions with incomplete evidence, coordinate responders, and communicate what is known during a cyber crisis.
Read more
Human and AI Collaboration: Using Intuition and Testing the Output
How people can use generative AI to explore ideas and analyze security problems while keeping evidence and judgment in the process.
Read more