A well-kept incident record turns scattered alerts, chat messages, and analyst notes into a defensible account of what happened. This guide provides a copy-ready security incident documentation template, evidence log, timeline structure, and severity matrix for investigations, shift handoffs, audits, tabletop exercises, and post-incident reviews.
Overview
During a cybersecurity incident, teams must investigate and act at the same time. That makes documentation easy to postpone. The result is often a timeline reconstructed from memory, evidence stored without context, and decisions that cannot be explained later.
Incident documentation should answer five basic questions:
- What was observed? Record the original alert, report, or anomaly without rewriting it into a conclusion.
- What was affected? Identify accounts, devices, applications, cloud resources, data, locations, and third parties in scope.
- What was done? Log containment, investigation, eradication, recovery, and monitoring actions.
- Why were decisions made? Capture assumptions, confidence levels, approvals, and unresolved questions.
- What changed over time? Keep a chronological record of discovery, impact, communications, and recovery milestones.
Use a dedicated incident record with controlled access rather than relying on a personal notebook or an unstructured chat channel. Restrict sensitive material to people who need it, preserve the original form of key evidence, and separate confirmed facts from working theories. Documentation supports response; it should not delay urgent containment or safety decisions.
What to track
1. The incident log
Start one row for every meaningful event, action, or decision. A copy-ready format is:
Incident ID:
Date opened:
Incident owner:
Current status:
Severity:
Timestamp (include time zone):
Reporter or system:
Event or action:
Affected asset/account:
Evidence reference or location:
Decision, owner, and approval:
Business impact:
Confidence: confirmed / probable / possible
Next step and due time:
Use a consistent timestamp standard, such as UTC, and retain the local time when it helps explain business operations. Avoid vague entries such as “checked logs” or “system fixed.” Prefer “reviewed authentication events for the finance tenant from 09:00–12:00 UTC; found 14 failed logins and no confirmed successful access.” Specific entries are easier to validate and hand off.
2. The breach timeline
The timeline should distinguish between when an event occurred, when it was detected, and when the organization acted. Those are often different moments. Track at least:
- First known suspicious activity or initial access indicator
- Detection time and detection source
- Incident declaration and assignment of an owner
- Accounts, hosts, tokens, services, or network paths contained
- Evidence collected and analysis completed
- Scope changes, including newly affected or cleared assets
- Internal, customer, partner, insurer, legal, or regulator communications, when applicable
- Restoration, validation, heightened monitoring, and closure milestones
Mark unknown times as unknown instead of inventing precision. When a time is estimated, label it as an estimate and record the basis for it, such as a log timestamp, user report, backup record, or interview.
3. The evidence log
An evidence log gives every item enough context to be useful later. Record:
Evidence ID:
Description:
Source system or custodian:
Collection date and time:
Collector:
Original location:
File name, query, or reference:
Hash or integrity check, if used:
Storage location and access controls:
Analysis performed:
Related timeline entries:
Keep original exports separate from working copies. Preserve relevant query text, filters, collection scope, and retention limitations. Screenshots can illustrate an issue, but they may not preserve the underlying metadata; where practical, retain the source logs or system export as well. If legal or investigative requirements apply, follow the organization’s evidence-preservation and chain-of-custody procedures.
4. Business impact and communications
Document operational effects separately from technical findings. Note unavailable services, delayed transactions, affected business processes, data categories potentially involved, recovery dependencies, and decisions about customer or employee communications. Do not state that data was exfiltrated, exposed, or unaffected unless the evidence supports that conclusion. Use careful terms such as “under investigation” or “no evidence identified as of [time]” when the investigation remains open.
For a broader notification workflow, use the cyber incident communications matrix. If a vendor may be involved, document its systems, notifications, requested evidence, response deadlines, and escalation path; the third-party incident response guide provides a related question set.
Cadence and checkpoints
Documentation quality improves when updates happen at defined checkpoints rather than only at the end of an incident.
- At intake: Create the incident ID, preserve the initial alert, record the reporter, and state what is not yet known.
- At every material action: Log containment, credential resets, isolation, block rules, restoration, and changes to monitoring.
- At each shift handoff: Summarize confirmed facts, open questions, current risk, pending actions, evidence locations, and the next decision point.
- At scope changes: Update affected assets, data, users, vendors, and business processes. Record why the scope changed.
- At communication checkpoints: Link each message to its approval, audience, delivery time, and factual basis.
- At closure: Confirm recovery criteria, monitoring ownership, unresolved risks, evidence retention, and follow-up actions.
For recurring readiness, review the template monthly or quarterly in a low-pressure setting. Test whether responders can find the incident owner, reconstruct the last eight hours, locate evidence, and identify the next action. A quarterly tabletop exercise can expose missing fields before a real event; see the incident readiness workbook for scenario-based practice.
How to interpret changes
A severity matrix should guide escalation without pretending to replace judgment. Define thresholds in advance and record which threshold was met. One practical model is:
| Level | Typical indicators | Response implication |
|---|---|---|
| Low | Limited suspicious activity, no confirmed compromise, and no material service impact | Assign an owner, investigate, preserve evidence, and monitor for change |
| Moderate | Confirmed compromise of a contained account or device, repeated indicators, or a meaningful control failure | Coordinate security and system owners; expand scoping and consider leadership notification |
| High | Multiple affected assets, privileged access, material service disruption, suspected data exposure, or active attacker behavior | Activate the full response structure, increase update frequency, and evaluate communications and reporting obligations |
| Critical | Widespread operational impact, safety or essential-service concerns, destructive activity, or rapidly expanding compromise | Escalate immediately to executive decision-makers and relevant response partners while prioritizing containment and continuity |
Severity can move in either direction. A suspected incident may be downgraded when evidence rules out impact, while a small alert may become high severity after privileged access or data exposure is confirmed. Record the reason, time, approver, and evidence for every change. Do not use the absence of evidence as proof that an event did not occur; state the collection limits and remaining uncertainty.
For ransomware, preserve evidence before rebuilding where safely possible and track recovery assumptions explicitly. The ransomware incident checklist covers those decisions. For suspected credential reuse after an external leak, link affected accounts and reset priorities to the evidence record; the credential-stuffing guide can support that assessment.
When to revisit
Revisit this documentation system monthly or quarterly, and immediately after a significant incident, audit finding, tabletop exercise, platform migration, or change to logging and retention. Update field names when teams adopt new systems, cloud services, alert sources, or case-management workflows. Review whether the severity thresholds still lead to the right escalation decisions and whether contact details, evidence locations, and communication paths remain accurate.
At the next review, sample a closed incident and ask:
- Could someone unfamiliar with the event reconstruct the timeline?
- Can every major claim be linked to evidence or clearly labeled as an assumption?
- Are containment and recovery actions tied to owners and completion times?
- Were impact and notification decisions documented without overstating certainty?
- Can the team identify what must happen next if the same signal appears tomorrow?
Make one small improvement after each review: add a missing field, clarify a threshold, automate a timestamp, or move evidence into a controlled location. Then run the revised template through a tabletop exercise. Consistent, modest maintenance keeps an incident record useful when the next alert arrives, when responsibility changes hands, and when decisions must be explained months later.