Most security incidents that go badly do not fail on technical skill: they fail because nobody is clearly in charge, decisions drift, and the people doing the analysis are also trying to brief executives and regulators. The incident commander (IC) role exists to fix that. It is a coordination and decision role, distinct from the technical lead, and it can be learnt and rehearsed. This guide covers what the role owns and what it delegates, the command structure around it, the battle rhythm of briefings and situation reports, how to make and record decisions under uncertainty, handover, and how to build a bench of commanders. It goes deeper than our Incident Response Guide on the leadership side, and is written for SOC leads, security managers and anyone who may be asked to run a major cyber incident.
Key takeaways
- The commander coordinates, others operate. PagerDuty's public training is explicit that the IC should not perform remediation, check graphs or read logs; NCSC notes the central coordinator need not be a cyber expert.
- One commander, always named. Everyone on the response must know who holds command at every moment, and transfer of command is announced and acknowledged.
- Rhythm beats heroics. A fixed cycle of briefings, written situation reports and decision points keeps a multi-day incident under control and stops executives pulling analysts off the work.
- Decide with what you have, and write down why. A recorded decision with options, rationale and authority is defensible later; an unrecorded one is not.
- Build a bench before you need it. NCSC advises more than one in-house person able to fill each key role, deputies for critical people and shift patterns for long incidents.
What the incident commander owns
The commander holds every responsibility that has not been delegated. Google's SRE guidance describes the commander as tracking the overall state of the incident, organising the team, assigning work by need and priority, and clearing obstacles for the people doing the operational work. In a cyber incident that translates into five areas.
- Objectives and priorities: set and restate the current objectives in plain words, for example "contain the attacker's access to the finance domain, preserve evidence, keep payroll running". Google's priorities for incidents are a useful default: stop the damage, restore service, preserve evidence.
- Structure and staffing: decide which roles are needed, appoint people, and split work into workstreams with a single lead each when the response grows.
- Decisions: own decisions within delegated authority, and escalate those beyond it (paying a ransom, taking a revenue service offline, notifying a regulator) to the named executive decision maker.
- Information flow: make sure findings reach the people who need them and that one version of the truth exists in the incident record.
- Pace and welfare: watch for overload, call breaks, start shifts and request more help early.
What the commander does not own matters just as much. Technical investigation belongs to the technical lead and analysts; legal privilege and notification judgements belong to counsel and the data protection officer; external statements belong to the communications lead. The commander makes sure each of those happens on time.
Incident commander versus technical lead
| Question | Incident commander | Technical lead |
|---|---|---|
| Primary focus | Whole response: people, decisions, stakeholders, time | Attacker activity, scope, containment and eradication actions |
| Hands on keyboard | No | Directs analysts, may work hands on in small teams |
| Typical output | Objectives, task assignments, situation reports, decision log entries | Findings, indicators, technical options with risks, verification results |
| Decision authority | Final call within delegated authority; escalates the rest | Recommends technical options; decides how to execute approved actions |
| Talks to executives | Yes, through scheduled briefings | Only when the commander brings them in |
| Who can fill it | Trained incident manager, SOC lead, IT service manager, security manager | Senior analyst, DFIR lead, retainer lead |
In small teams one person may have to do both for the first hour. Separate them as soon as a second capable person is available, because the commander who is also investigating stops seeing the whole incident.
The command structure around the commander
Incident command for cyber response borrows from the US Incident Command System (ICS) used in emergency management, which organises response into functional areas: command, operations, planning, logistics, intelligence and investigations, and finance and administration. Most cyber teams use a lighter version.
| Role | Responsibility | When to appoint |
|---|---|---|
| Incident commander | Objectives, decisions, structure, briefings | At declaration |
| Deputy commander | Tracks task timings, chases owners, ready to take over; PagerDuty requires deputies to be as qualified as the IC | Any high-severity incident |
| Scribe | Timeline of key events, actions and owners, decisions (as decisions, not vote counts), follow-ups tagged for the review | At declaration |
| Technical lead (operations) | Investigation, containment, eradication; only this group changes systems, as in Google's model | At declaration |
| Communications lead | Internal updates, stakeholder and media statements, customer messaging | When impact is visible beyond the team |
| Legal and privacy lead | Privilege, notification assessment, regulator and law enforcement contact | Any suspected data exposure or criminal activity |
| Liaison | IR retainer, insurer, suppliers, CSIRTs and law enforcement | When third parties are engaged |
| Business owner representatives | Impact assessment, acceptance of business disruption, workarounds | When containment affects services |
| Logistics | Access, rotas, food, accommodation, out-of-band tools | Incidents expected to run beyond a day |
- Workstreams: when the technical work grows, split it into named workstreams (for example identity, endpoint, cloud, recovery), each with a single lead who reports to the commander. PagerDuty calls these sub-teams with a single point of contact.
- Span of control: keep the number of people reporting directly to the commander small. If more than about five or six leads report in, add a layer.
- Executive layer: the commander is not the crisis management team. For major incidents, a separate executive group sets business priorities and takes strategic decisions; the commander briefs it. See the Crisis Management Guide.
- External responders: an IR retainer team brings its own lead. Agree at engagement that your commander keeps command and their lead acts as technical lead or workstream lead, so there is no dual command.
The first hour
- Take and announce command: "I am incident commander for INC-[number] from [time UTC]." Confirm the severity and declaration with the person who declared. See the Incident Severity Classification Matrix Template.
- Appoint a scribe and open the record: start the decision log immediately. See the Incident Log and Decision Record Template.
- Get a first picture: ask the technical lead three questions: what do we know, what do we not know, what is the worst plausible case. Write the answers down as facts and assumptions.
- Check the channel is safe: if the attacker may have access to email, chat or the identity provider, move to out-of-band communications before discussing containment. See the Out-of-Band Communications for Incident Response Guide.
- Protect evidence: confirm volatile evidence and cloud audit logs are being preserved before disruptive containment.
- Set objectives and the first operational period: state objectives for the next two to four hours and when the next briefing is.
- Mobilise: call in legal, privacy, communications, the retainer and insurer as the severity requires, using the call tree. See the Incident Response Contact and Call Tree Template.
- Start regulatory clocks: ask counsel to record the awareness time for any notification regime that may apply. See the Cyber Incident Reporting Requirements Map.
Battle rhythm and situation reports
A battle rhythm is the fixed schedule of meetings and outputs that drives the response. Without it, every stakeholder asks for updates whenever they like, which interrupts the people doing the work.
| Event | Typical cadence (high severity) | Attendees | Output |
|---|---|---|---|
| Tactical stand-up | Every 1 to 2 hours in the acute phase | Commander, technical lead, workstream leads, scribe | Updated task list, blockers, next decisions |
| Situation report (sitrep) | Every 4 hours, or after each stand-up in the acute phase | Written, sent to a fixed distribution | One-page status in a standard format |
| Executive briefing | Twice daily, or as the executive group sets | Commander, executive decision maker, legal, communications | Decisions needed, risks, business impact |
| Legal and notification check | Daily, and on any new finding about data | Commander, counsel, DPO | Notification decisions and their rationale |
| Shift handover | Each shift change | Outgoing and incoming commander and leads | Handover brief, logged |
Sitrep format
- Header: incident reference, severity, time (UTC), sitrep number, commander, next sitrep due.
- Bottom line: two sentences on the current position.
- What changed since the last sitrep: new facts only, with confidence level.
- Impact: services, data, customers, people affected.
- Actions in progress and completed: with owners.
- Decisions taken and decisions needed: who must decide and by when.
- Notifications: made, pending, deadlines.
- Risks and assumptions: what could change the picture.
Keep sitreps short and factual. PagerDuty's guidance to fill quiet periods with a summary of the situation and outstanding actions applies to calls too. Mark sitreps with their handling classification, and assume they may be disclosed in litigation, so avoid speculation and blame.
Making decisions under uncertainty
Containment decisions trade evidence, business continuity and attacker awareness against each other, often with incomplete information. NCSC notes that executives make urgent, critical business decisions with limited information and should know in advance which decisions they may face.
- Gather options with risks: ask the technical lead for at least two options, each with its risk, business impact and reversibility.
- Test for objections: PagerDuty's approach is to ask "are there any strong objections?", note objections and then decide. The commander's decision is final within their authority.
- Prefer reversible actions early: disabling an account or revoking a token can be undone; wiping a host cannot.
- Time-box the decision: waiting for certainty is itself a decision. PagerDuty's training puts it plainly: a wrong decision is better than no decision.
- Assign to a named person with a deadline: never "can someone", which invites everyone to assume someone else will. Get an acknowledgement.
- Record it: options, decision, rationale, authority and time in the decision log.
Decisions to escalate beyond the commander
| Decision | Usually escalated to | Why |
|---|---|---|
| Shutting down a revenue or safety-critical service | Executive decision maker | Business and possibly safety impact |
| Engaging with or paying an extortion demand | Executive group, counsel, insurer | Legal, sanctions and reputational risk |
| Notifying regulators, customers or the market | Counsel, DPO, executive group | Legal obligations and disclosure judgement |
| Involving law enforcement | Counsel and executive decision maker | Evidence, publicity and legal implications |
| Global credential reset or identity provider rebuild | Executive decision maker, CIO | Organisation-wide disruption |
The Colonial Pipeline ransomware attack shows the weight of these calls: the company proactively shut down its pipeline system, and the physical disruption came from that containment decision rather than from the attackers touching pipeline controls. Decisions of that size need a pre-agreed authority and criteria, not improvisation.
Handover and long-running incidents
Major cyber incidents can run for days or weeks. Command must transfer cleanly, and people must rest.
- Plan shifts early: if an incident looks likely to pass 12 hours, set up rotas for the commander, technical lead and scribe. NCSC advises shift patterns for protracted incidents and planning for food, accommodation and building access.
- Brief out of band first: the outgoing commander briefs the incoming one privately (objectives, open decisions, risks, people concerns) before the switch.
- Announce the transfer: on the call and in the record, with the time. Google's guidance is that the outgoing commander stays until the incoming commander explicitly acknowledges taking over.
- Do not hand over to rank: a more senior person arriving does not automatically take command. Transfer only when it helps the response.
- Use deputies for continuity: a deputy who has been on the incident throughout makes a better incoming commander than someone new.
Where identity fits
Many incidents are identity incidents: a stolen session token, a help desk reset abused by a caller, a leaked API key or a compromised identity provider tenant. The commander does not run the identity workstream but should insist it exists from the start, with an inventory of affected human and non-human identities, an owner for each and a rotation plan sequenced so that revoking one credential does not break recovery. Commanders should also check that their own response tooling does not depend on the compromised directory. See the Identity Incident Response Playbook and the Identity Provider Tenant Compromise Playbook.
Building a bench of commanders
- Who to train: SOC leads, IT major incident managers and security managers. Technical depth helps but is not required; calm communication, structured thinking and willingness to decide are.
- How many: at least three people able to command for 24/7 coverage and shift changes; NCSC advises more than one in-house person for each key role, including coordination.
- Training path: shadow as scribe, then deputy, then command a low-severity incident with a mentor, then lead a tabletop. See the Tabletop Exercise Guide.
- Rotate roles: Google recommends rotating people through roles so everyone understands each one.
- Use the process for smaller events: running minor incidents with the same structure keeps it familiar for major ones.
- Small organisations: if you rely on an IR retainer, decide in advance which internal person commands and which decisions the retainer can take. An outsourced technical team still needs an in-house decision maker, as NCSC points out.
Measuring command effectiveness
- Time from declaration to named commander and scribe (target under 15 minutes for high severity).
- Percentage of sitreps issued on schedule.
- Percentage of decision log entries with rationale and authority.
- Number of commanders on the rota who have commanded in an exercise in the last 12 months.
- Post-incident review findings attributed to coordination rather than technical gaps, trended over time.
Framework mapping
| Framework | Reference | Relevance to the commander role |
|---|---|---|
| NIST CSF 2.0 | RS.MA-01 | The incident response plan is executed in coordination with relevant third parties once an incident is declared |
| NIST CSF 2.0 | RS.MA-02, RS.AN-08 | Triage and validation of reports; estimating and validating an incident's magnitude |
| NIST CSF 2.0 | RS.CO-02, RS.CO-03 | Stakeholder notification and information sharing |
| NIST CSF 2.0 | RS.AN-06 | Actions during investigation are recorded with integrity and provenance |
| NIST SP 800-53 Rev. 5 | IR-4, IR-7, IR-8 | Incident handling, incident response assistance, incident response plan (roles) |
| ISO/IEC 27001:2022 Annex A | 5.24, 5.26 | Incident management planning, roles and responsibilities; response to incidents |
| CIS Controls v8.1 | Control 17 | Designating personnel to manage incident handling, contact information, incident response process |
Common failure patterns
- The commander starts investigating: they disappear into logs and nobody is watching the whole incident or the clock on notifications.
- Dual command: an executive, a retainer lead and the SOC manager all give instructions. Name one commander and route all direction through them.
- Briefing by interruption: without a sitrep cadence, executives call analysts directly and the work slows.
- Decisions without authority: the commander is asked to approve shutting down a core service but has no delegated authority, so the decision waits for hours.
- Coordinating on compromised channels: containment is discussed in email or chat the attacker can read.
- No handover: the commander works 30 hours straight, decisions degrade, and nobody else has the context.
Practitioner checklist
- Name at least three trained incident commanders and publish the rota.
- Write delegated authority for commanders into the incident response plan, with the decisions they must escalate.
- Agree a standard sitrep format and distribution list in advance.
- Set default battle rhythm cadences per severity level.
- Train scribes and use the decision log template in every exercise.
- Prepare out-of-band communication channels that do not rely on the production identity provider.
- Agree with your IR retainer that your commander keeps command.
- Include a handover drill in at least one exercise each year.
- Review command performance in every post-incident review, separately from technical findings.
Standards and references
- NCSC: Build a cyber security incident response team
- NCSC: Your cyber incident response processes
- NIST: SP 800-61 Rev. 3, Incident Response Recommendations and Considerations for Cybersecurity Risk Management
- NIST: Cybersecurity Framework 2.0
- FEMA: NIMS components and the Incident Command System
- Google: Site Reliability Engineering, Managing Incidents
- PagerDuty: Incident Response documentation, Incident Commander training
Related NHI Mgmt Group resources: Incident Response Guide · Incident Response Plan Template · Crisis Management Guide · Incident Log and Decision Record Template · Top 10 Incident Response Mistakes