2.5.1. Automation and Orchestration
💡 First Principle: Automate tasks that are repeatable, rule-based, and don't require contextual judgment — leave tasks that require interpreting ambiguous evidence, weighing competing priorities, or making ethical decisions to human analysts.
Identifying Tasks Suitable for Automation:
The key question: "Could we write a complete decision procedure for this, with no ambiguous cases?" If yes, it's a candidate for automation.
| Suitable for Automation | Requires Human Judgment |
|---|---|
| Blocking a known-bad IP at the firewall | Deciding whether to isolate a production server mid-incident |
| Disabling a user account when impossible travel is detected | Determining if a novel process behavior is malicious or legitimate |
| Querying VirusTotal for a file hash | Scoping an active incident with incomplete information |
| Creating a ticket for a triggered alert | Communicating incident status to executives |
| Enriching an alert with WHOIS and geolocation data | Deciding whether to notify law enforcement |
| Resetting a compromised user's credentials after MFA failure threshold | Evaluating whether an unusual but authorized admin action is suspicious |
SOAR (Security Orchestration, Automation, and Response) executes automated playbooks triggered by SIEM alerts. A typical SOAR workflow:
Orchestrating threat intelligence data:
- Data enrichment — Automatically augmenting alerts with context from threat intel sources, asset databases, vulnerability scanners, and HR systems before presenting to analysts
- Threat feed combination — Aggregating multiple threat intel feeds, deduplicating, and normalizing into a unified IoC database for SIEM ingestion
- Minimizing human engagement — SOAR handles the first 10 steps of an investigation automatically so analysts start with context, not raw data
Measuring Automation and SOC Efficiency:
You cannot prove automation is working without metrics. The core SOC efficiency metrics:
| Metric | Measures | What Improves It |
|---|---|---|
| MTTD (Mean Time to Detect) | Time from compromise to detection | Better monitoring, log coverage, correlation rules |
| MTTR (Mean Time to Respond) | Time from detection to containment/remediation | Automation/SOAR, clear playbooks, practiced response |
| MTBF (Mean Time Between Failures) | Average uptime between system failures | A reliability metric — the outlier here; it measures hardware/service dependability, not SOC speed |
SOAR primarily drives MTTR down by automating the response steps; MTTD depends on detection quality, not response automation. MTBF is a reliability metric that may appear on infrastructure dashboards but does not measure detection or response — a common distractor when it is listed alongside the two SOC-speed metrics.
⚠️ Exam Trap: Team coordination is required to implement automation effectively — automation that blocks a production IP without analyst awareness can cause an outage. Successful SOAR deployments involve stakeholders from security, IT operations, and business units to define acceptable automated actions and their blast radius.
Reflection Question: A SOC implements a SOAR playbook that automatically disables any user account that triggers an impossible travel alert. Within the first week, it disables the CEO's account when they use a corporate VPN that exits in a different country. What does this incident reveal about the automation design, and what control would prevent it?
