Blue Team · 9 min read · Log Academy training team
What Is Incident Response? The Phases Explained
Incident response is a process prepared in advance for when a cyber incident happens: detect it, stop it from spreading, remove the threat from the system, restore operations and learn from the outcome. The most widely used models are the NIST and SANS frameworks. The goal is to act on a plan rather than in a panic and keep the damage to a minimum.
- Incident response is not an improvised reaction but a process with defined phases.
- The SANS model has six phases: preparation, identification, containment, eradication, recovery and lessons learned.
- In the first hour, the priority is to preserve evidence and stop the spread, not to wipe the system.
- Every incident is closed with a written report, and that is where preparation for the next one begins.
- Incident response is one of the core skills of a Blue Team member and a SOC analyst.
What is incident response and whose job is it?
A cyber incident is an actual security breach or a suspected one: a malware infection, a login with a stolen account, a data leak, an attack that starts with a phishing email. Not every alert is an incident; an alert becomes an incident once it has been investigated and confirmed.
In a small company, a single system administrator may handle this work. Large organizations have a dedicated team: SOC analysts detect and triage the incident, incident response specialists run the in-depth investigation, and management and the legal department step in for decisions and communication.
What are the phases of incident response?
| Phase | Goal | Example work |
|---|---|---|
| 1. Preparation | Have tools, a plan and roles ready before an incident | Log collection, contact list, playbooks |
| 2. Identification | Confirm that an incident has occurred | Investigate a SIEM alert, determine the scope of impact |
| 3. Containment | Stop the spread | Disconnect the machine from the network, block the account |
| 4. Eradication | Remove the threat from the system | Delete the malicious file, close the vulnerability |
| 5. Recovery | Return to normal operations | Restore from a clean backup, increased monitoring |
| 6. Lessons learned | Prevent a recurrence | Report, updates to rules and procedures |
Revision 2 of NIST SP 800-61 splits the same work into four phases: preparation; detection and analysis; containment, eradication and recovery; and post-incident activity. Revision 3, published in 2025, maps incident response to the functions of the NIST Cybersecurity Framework 2.0. The models differ, but the logic is the same.
What should you do in the first hour of an incident?
- Stay calm and note the time: when, who and what was seen.
- Check whether the alert is real: does the same behavior show up in other sources?
- Determine the scope: which machines, which accounts, which data.
- Isolate the infected machine from the network, but do not shut it down: evidence in memory may be lost.
- Change the passwords of suspicious accounts and end their active sessions.
- Notify the people listed in the plan and management.
- Log every step you take, along with the time.
The most common mistake
Formatting or shutting down the suspicious machine right away. This destroys the evidence of how the attacker got in, and a second attack comes through the same gap.
What is the difference between containment and eradication?
Containment is not putting out the fire; it is stopping it from spreading. For example, the machine is isolated, the account is blocked, and the suspicious IP address is cut off at the firewall. These steps should be fast and reversible.
Eradication removes the root cause: deleting malicious files and autostart entries, patching the exploited vulnerability, and removing accounts the attacker created. Moving to eradication before the root cause is found often ends in reinfection.
How are incidents classified by severity?
It is impossible to respond to every incident at the same speed. That is why severity levels are defined in advance, during the preparation phase. The level is assigned based on the incident's impact, how fast it spreads and what data it touches.
| Level | Example | Response |
|---|---|---|
| Low | A blocked phishing email; nobody clicked the link | Log it and review in the normal queue |
| Medium | Malware detected on one computer and quarantined | Investigation and cleanup within the shift |
| High | A server was accessed with a stolen account | Immediate containment and escalation |
| Critical | Ransomware is spreading or a data leak is confirmed | Immediate involvement of the whole team and management |
The breakdown varies from company to company. What matters is that for each level, who gets notified and the response time are defined in writing. Making these decisions for the first time in the middle of a crisis costs time and invites mistakes.
What should an incident response plan include?
- Roles and responsibilities: who manages the incident, who does the technical work, who briefs management.
- Contact list: people to call outside working hours, plus external partners.
- Playbooks: step-by-step instructions for typical scenarios such as phishing, malware and a stolen account.
- Evidence collection rules: which logs and memory images to collect, and where and how they are stored.
- Communication rules: when and how to inform employees, customers and, if needed, government agencies.
- Exercise schedule: regular testing of the plan through tabletop exercises.
A plan that is never exercised rarely works in a real incident. In a tabletop exercise, the team walks through a realistic scenario: "A ransomware alert comes in at 2:00 a.m. Who does what in the first ten minutes?" Exercises like this expose gaps in the plan before an incident does.
How do you write an incident report?
The report should be clear even to someone who did not follow the incident. Technical details can go in an appendix, but the main part should be short and structured.
- Executive summary: what happened, when it was detected, what the impact was.
- Timeline: key events from the first trace of the attack to closure.
- Root cause: how the attacker got in.
- Actions taken: containment, eradication and recovery steps.
- Indicators of compromise (IOCs): IP addresses, file hashes, domains.
- Recommendations: what needs to change to prevent a recurrence.
How we do it in our lab
In the Log Academy Blue Team program, incident response is taught as a separate module: triage, escalation, containment steps and writing the incident report. The course ends with a capstone project in which each student investigates a simulated incident from scratch in a personal lab and prepares a full defense report. In the lab scenario, the student uses ELK or Wazuh to find a series of failed logins followed by a successful login and the creation of a new user, tags it with a MITRE ATT&CK technique, and documents the incident in the sequence triage → escalation → containment.
Real-world scenario: how do you respond to a compromised account?
Late at night, the SIEM raises an alert: after several failed attempts from abroad, someone has logged in successfully to an accounting employee's account. A few minutes later, a new user is created under that account. The steps below show how the SANS model plays out in this scenario.
- Identification: the analyst checks where and when the employee usually logs in. The login does not match normal behavior, so the alert is confirmed.
- Containment: the account is blocked, active sessions are terminated, and the newly created user is disabled.
- Investigation: the logs are traced to see which machines the attacker accessed, which files they viewed and which commands they ran.
- Eradication: accounts and autostart entries created by the attacker are removed, the password is reset, and two-factor authentication (2FA) is enabled.
- Recovery: the employee returns to work with a new password, and the account stays under increased monitoring for a while.
- Lessons learned: the team determines how the password was stolen, for example through a phishing email, and plans awareness training for employees.
What skills do you need to learn incident response?
- Networking fundamentals: to tell normal traffic from anomalous traffic.
- Linux and Windows logs: to find traces of an incident in system logs.
- Working with a SIEM: search, correlation and alert analysis.
- MITRE ATT&CK: to describe attacker behavior in a shared language.
- Writing skills: producing a clear, fact-based report.
Frequently asked questions
- How would you define incident response in one sentence?
- A planned process for detecting a cyber incident, stopping its spread, removing the threat, restoring the system and learning from the outcome.
- Is every alert an incident?
- No. An alert is only a signal. It is recorded as an incident once an investigation confirms it is a real threat.
- What should a company without an incident response plan do?
- At a minimum, write down on one page who makes decisions, who to call, where logs are stored and how to isolate a machine. The plan can be improved later.
- Are incident response and digital forensics the same thing?
- No, but they are closely linked. Forensics is the collection and in-depth analysis of evidence, while incident response is the management of the incident as a whole, and forensics can be part of it.
- Where can you learn incident response?
- In a hands-on lab. The 6-month Log Academy Blue Team program covers incident response, SIEM and detection rules in both offline and online formats. The first lesson is free.