Incident vs Event: Defining the Threshold
You'll learn: How to draw the line between routine security events and actual incidents that require escalation and response.
Understanding Events vs Incidents
In a SOC, your SIEM generates thousands or millions of events daily—logins, file accesses, firewall blocks, DNS queries. Most are normal business activity. An incident is a subset of events that indicates actual harm, policy violation, or imminent threat requiring coordinated response.
Think of it like a hospital emergency department: every patient arrival is an "event," but only those needing urgent intervention become "incidents" requiring a trauma team.
Defining the Threshold
The key question: When does an event cross into incident territory?
Common incident criteria:
- Confirmed malicious activity (not just suspicious)
- Policy violation with business impact
- Data confidentiality, integrity, or availability compromise
- Successful exploitation or unauthorized access
- Regulatory trigger (breach notification thresholds)
Examples:
-
Event: Failed login attempt from unusual location
-
Incident: Successful login followed by data exfiltration
-
Event: Antivirus blocks malware download
-
Incident: Ransomware actively encrypting files
Classification Criteria
Your organization needs clear escalation thresholds documented in playbooks:
- Severity scoring (critical/high/medium/low based on CIA impact)
- Affected asset criticality (production server vs. test laptop)
- Scope (single user vs. enterprise-wide)
- Confirmation level (true positive validation)
Without defined criteria, you'll either escalate everything (overwhelming responders) or miss real incidents buried in noise.
Key Takeaway: An event is an observation; an incident is a confirmed security problem requiring response. Clear classification criteria prevent both alert fatigue and missed threats.