Incident Response Plan: What to Write, Section by Section
Skip to content

Incident Response Plan

An incident response plan is a written document that sets out how an organization will detect, manage, and recover from a cybersecurity incident. It records who decides, who acts, and how people reach each other when normal systems may be unavailable.

This guide goes through a plan section by section. For each one it explains what to write and gives example wording that can be adapted.

What Makes a Plan Work in Practice

Most organizations that suffer a serious incident already had a plan. In many cases nobody opened it. The plans that are used during a real incident share a few features, and they have little to do with length or polish.

They can be found. A plan stored only on the file server is gone when the file server is encrypted. Working plans exist on paper and somewhere outside the corporate environment.

They name people. The CISO will notify the board works until the CISO is on a flight. A plan that gives each role a named person and a named deputy, with mobile numbers, still works that day.

They settle authority. The single most common delay in real incidents is waiting for someone to approve an action that everyone agrees is necessary. A working plan says in advance who may disconnect a server, and who may take the website down.

They match reality. A plan that assumes a 24x7 security team, when the organization has two IT staff, describes a different company. The process has to fit the people and tools that actually exist.

A plan is a record of decisions made in calm conditions, for use when conditions are anything but calm. The rest of this guide is about which decisions to record.

Writing for Different Readers

A plan has several audiences, and each reads a different part.

The person who answers the first call

They need to know whether this is an incident, who to tell, and what to do in the next 15 minutes. They read the activation section and the contact list.

The incident lead

They need the whole process, the authority they hold, and when to escalate.

Executives

They need to know which decisions are theirs, how they will be kept informed, and who speaks for the company. They read the roles, severity, and communication sections.

Legal, privacy, and communications staff

They need to know when they are brought in and what is expected.

External responders

They need to know who their contact is, what access they have, and what they may do without asking.

Auditors, insurers, and regulators

They read the plan afterward, to see whether the organization had one and followed it.

A common and effective layout puts a one-page summary first: how to declare an incident, the severity table, and the key phone numbers.

Plan Outline

A workable plan has few sections, and each answers one question. The table is the whole outline, and it comes to eight to ten pages before the contact list.

SectionQuestion it answersLength
Purpose and scopeWhat counts as an incident, and what does this plan cover?Half a page
Roles and authorityWho does what, and who may decide what?One to two pages
Severity levelsHow serious is this, and who must be told?One page
Reporting and activationHow does an incident get declared?Half a page
CommunicationHow do we talk to each other and to others?One page
Response processWhat happens, in what order?One to two pages
Evidence and recordsWhat do we keep, and how?Half a page
External partiesWho do we call, and when?One page
Recovery handoverWho decides it is safe to restore?Half a page
ReviewHow do we learn from it?Half a page
Contact listHow do we reach everyone?Appendix

Detailed technical steps for specific incident types belong in playbooks, which the plan lists. Keeping them out is what lets the plan stay short enough to read during an incident.

The Australian Signals Directorate publishes guidance for cyber incident response plans that organizations can use as a starting structure.

Purpose and Scope

State what the plan covers in plain terms: which parts of the organization, which systems, and which kinds of event. Define an incident so that staff outside IT can recognize one.

Example wording: A cybersecurity incident is any event that has compromised, or is likely to compromise, the confidentiality, integrity, or availability of our systems or information. This plan applies to all company systems, cloud services, and information held by suppliers on our behalf.

Say what the plan does not replace. The disaster recovery and business continuity plans stay separate documents, and the recovery handover section records where each takes over.

Roles and Authority

List each role with the people who hold it. The most valuable lines in the whole plan are the ones that say who may act without asking.

Incident lead

Runs the response and holds the single current version of the facts.

Technical lead

Directs investigation and containment.

IT operations

Isolates, rebuilds, and restores.

Executive sponsor

Approves actions that affect customers, revenue, or public statements.

Legal counsel

Advises on notification, privilege, and contracts.

Communications

Prepares messages for staff, customers, and media.

Privacy lead

Assesses whether personal information is involved.

Scribe

Keeps the incident log of events, decisions, and actions, each with its time.

Example wording: The incident lead may isolate any workstation or internal server, and disable any standard user account, without further approval. Taking a customer-facing system offline requires the executive sponsor, or the deputy if the sponsor cannot be reached within 15 minutes.

Write an equivalent sentence for each hard decision. Each is difficult to make well at 3 a.m. with incomplete facts, and each can be made on an ordinary afternoon.

Disconnecting a production system

Name who can order it, and say whether the answer changes for systems that customers use directly.

Disabling an executive’s account

Executives are frequent targets, and staff hesitate to act against a senior person without clear authority.

Cutting off a supplier

A managed service provider or software vendor with access may be the route in, and suspending it has contractual and operational consequences.

Calling external responders

A trigger tied to severity level avoids a debate about whether the situation is serious enough.

Emergency spending

Response costs arrive before any budget process can approve them, so the plan sets who can authorize them and up to what amount.

Responding to an extortion demand

This is a board-level decision in most organizations, and the board should have discussed it before it is real.

Running without systems for a week

The plan names who decides on manual processes, alternative suppliers, and which customers are told first.

For each, the plan needs one sentence naming the role, and one naming who decides if that person cannot be reached. The discussion that produces those sentences is often more valuable than the document.

Severity Levels

Define levels by business impact, in words a non-technical manager can apply in the middle of the night.

LevelDefinitionExampleNotifyMeet
CriticalOperations stopped, or sensitive data confirmed stolen.Ransomware across the network.Executives, board, legal, insurer, external responders.Immediately, then every two hours.
HighA key system or privileged account compromised, with spread likely.An administrator account in attacker hands.Executive sponsor, legal, external responders.Within the hour, then every four hours.
MediumA contained compromise with limited impact.One staff mailbox accessed, sessions revoked.Incident lead, IT management, privacy lead.Daily until closed.
LowA blocked or minor event needing follow-up.A phishing email reported and removed.Security team.No meeting.

Add the rule for changing a level. Example wording: The incident lead reviews the severity level whenever new facts arrive, and raises it at once if personal information, a privileged account, or more than one system is found to be involved.

Reporting and Activation

Give staff one way to report a suspected incident, and say what happens next.

Example wording: Report any suspected incident to the service desk by phone. The service desk passes the report to the on-call security contact within 15 minutes. The on-call contact assesses it, assigns a severity level, and declares an incident where the level is medium or higher.

Declaration matters because it is the trigger. It starts the notifications, the meeting rhythm, and the incident log for that level. A plan that leaves declaration vague produces incidents that are handled informally for a day before anyone is told.

Communication

Assume email and chat may be unavailable or read by the attacker.

A separate channel

Name the messaging or conferencing service the team will use, hosted outside the corporate environment and tested in advance.

One source of facts

The incident lead keeps the current summary. Everyone else refers to it.

Update rhythm

Set by severity level, as in the table above.

Who speaks externally

Name the people authorized to talk to customers, regulators, and media.

Holding statements

Keep short pre-approved wording for staff and customers, to be completed with facts on the day.

Example wording: No member of staff other than the communications lead or the executive sponsor makes statements about an incident to customers, suppliers, or the media.

Example holding statement for staff: We are responding to a cybersecurity incident that affects some of our systems. The response team is working on it now. Updates will come through [channel], and the next one is due at [time]. Pass any question from a customer, supplier, or journalist to [name] without answering it yourself.

A holding statement says what is known, when the next update comes, and where questions go. It leaves out the cause, the scale, and any promise that data is safe, because none of those is established in the first hours.

Response Process

Describe the flow in a page or two, so everyone shares one picture. The cyber incident response guide covers each phase in depth.

1
TriageConfirm the event is real and gather the basic facts.
2
DeclareAssign a level and start the incident log.
3
ContainLimit the damage while preserving evidence.
4
InvestigateEstablish how it started, what was affected, and whether data was taken.
5
EradicateRemove the attacker’s access and close the entry point.
6
RecoverRestore systems once they are confirmed clean.
7
Close and reviewRecord the outcome and hold the review.

State the principles that guide the technical work: preserve evidence before changing systems, contain at once where harm is spreading and otherwise establish scope first, and record every action with its time.

Example wording: The technical lead may begin containment before the investigation is complete where an attacker is active or damage is spreading. In all other cases the technical lead confirms the scope with the incident lead before accounts are disabled or systems are isolated.

Evidence and Records

Incident log

One timed record of events, decisions, and actions, kept from declaration.

Evidence handling

Who may collect evidence, where it is stored, and how access to it is recorded.

Log retention

Which logs exist, how long they are kept, and where copies are held outside the production environment.

Legal privilege

When counsel is engaged, so that advice and some investigation records can be protected where the law allows.

Example wording: The incident lead appoints a scribe at declaration. The scribe records each decision, who made it, and the time, in the incident log.

External Parties

List each outside party, the reason to contact them, and who makes the call: the incident response provider, the cyber insurer, legal counsel, key IT suppliers, the bank, the national cyber agency, police, and regulators.

For the response provider, record the steps to invoke the incident response retainer. For regulators, record who assesses whether a notification duty applies and the timeframes involved. Where personal information is involved, the data breach response plan carries that assessment.

Example wording: The incident lead calls the incident response provider for any incident rated high or critical. The executive sponsor notifies the insurer within the period the policy sets. Legal counsel assesses whether a regulator must be notified and advises the executive sponsor, who approves the notification.

Recovery Handover

Incident response, disaster recovery, and business continuity plans answer different questions, and a cyber incident can trigger all of them.

PlanQuestion it answersTypical trigger
Incident response planHow do we stop the attacker and establish what happened?A security incident.
Disaster recovery planHow do we restore systems and data?A system or site failure, from any cause.
Business continuity planHow does the business keep operating meanwhile?Any major disruption.

Disaster recovery plans usually assume backups and identity systems can be trusted. After an intrusion that may be false. The incident response plan should name who confirms that restoration is safe.

Example wording: No system is restored to production until the technical lead confirms that the entry point is closed and the backup predates the intrusion.

Review

Say who calls the review, how soon, and what happens to its findings. The cyber incident response guide sets out what a post-incident review covers. The plan only has to make sure one is held and acted on.

Example wording: The incident lead holds a review within two weeks of closing any incident rated medium or higher. The review records what delayed the response, which parts of this plan were used, and what should change. Each action has an owner and a date, and the plan owner reports open actions to the executive sponsor each month until they are closed.

Include the plan itself in the review. If the team worked around a section, the section is wrong, and the review is the moment to rewrite it.

Contact List

The contact list is the part of the plan used most and trusted least. It sits in an appendix so it can be replaced without reissuing the plan.

Internal roles

Give each role its named person and deputy, with a mobile number and a contact route that does not depend on corporate email.

External parties

Record the response provider’s incident line, the insurer’s claims line with the policy number, legal counsel, the bank’s fraud team, and out-of-hours numbers for key suppliers.

Verification date

Print the date of the last check on the list, and check every number each quarter.

Copies

Keep the list with the printed plan and on the phones of the people named in it.

Example entry: Incident lead: [name], mobile [number], personal email [address]. Deputy: [name], mobile [number]. Last verified: [date].

A Version for a Small Organization

An organization with one or two IT staff, or an outsourced provider, does not need every section above. It needs the same decisions on fewer pages.

A workable small-organization plan fits on two or three pages and covers:

1
What to do firstDisconnect the affected computer from the network, leave it switched on, and call the number below.
2
Who to callThe IT provider, the incident response provider or insurer hotline, and the owner or managing director, with mobile numbers.
3
Who decidesOne named person with authority to shut systems down and spend money, and one deputy.
4
How we talkA group chat on personal phones, set up in advance.
5
Where the backups areWhat is backed up, where, how to restore, and when it was last tested.
6
Who we may have to tellThe insurer, the bank, customers, and the privacy regulator, with a note on who makes that call.
7
Key accounts and suppliersInternet provider, email provider, domain registrar, accounting system, with support numbers.

Printed, signed by the owner, and kept where staff can find it, this covers most of what a longer plan achieves.

Outside Readers of the Plan

A plan is increasingly something outsiders ask to see.

Cyber insurers ask about incident response plans in applications and at renewal. A policy may require prompt notification, use of approved response firms, and consent before certain costs are incurred. The plan should carry the policy number, the insurer’s notification route, and those conditions, so they are followed under pressure. The incident response retainer guide covers getting a provider approved by the insurer in advance.

Regulators look at preparation when they assess an incident. Whether an organization had a documented, tested plan is relevant to whether it took reasonable steps. Sector rules for financial services and critical infrastructure go further and expect a maintained and exercised capability.

Customers include response and notification obligations in contracts. Timeframes for telling a customer about an incident affecting their data are often shorter than any legal requirement. A list of those obligations belongs in the plan’s appendix.

Auditors and certification bodies check for a plan, evidence that it is reviewed, and evidence that it is tested.

Keeping a simple record serves all of them: the plan’s version history, the dates of exercises, who attended, and what changed afterward.

Keeping the Plan Alive

One owner

A named person is responsible for keeping the plan current.

Version and date

Every copy shows its version and date, so nobody works from an old one.

Review on change

The owner updates the plan after new systems, new suppliers, staff changes in named roles, and every incident.

An annual exercise

The people named in the plan, including executives, rehearse it once a year. The cyber tabletop exercise guide has a script that can be run as written.

Briefing for new staff

Anyone given a role is briefed on it before they need it.

Incident Response Planning With ThreatDefence

ThreatDefence reviews incident response plans as part of our incident response retainer and our readiness review. We check roles, escalation contacts, logging, and containment authority, and report what would slow a response today.

The retainer also includes incident-specific runbooks and a tabletop exercise.

Under our 24x7 SOC service, escalation contacts and pre-approved containment actions are agreed at onboarding, so the plan and the monitoring service work from the same decisions.

Response actions in the ThreatDefence portal: revoke sessions, reset password, isolate endpoint and block address, each with its operating mode, status and rollback

Experiencing a breach?

Our incident response team answers 24x7, with or without a retainer.

Talk to Our Experts

Ask about an incident response readiness review, including a review of your plan, contacts, logging, and containment authority.

    I'm an MSP, MSSP, or reseller

    By submitting this form, you accept the terms in our privacy policy.

    Frequently Asked Questions

    What should an incident response plan include?

    A plan covers purpose and scope, roles and authority, severity levels, reporting and activation, communication, the response process, evidence and records, external parties, recovery handover, review, and a contact list. The lines that matter most say who may act without asking, such as who can disconnect a production system. Technical steps for specific incident types sit in separate playbooks that the plan lists.

    How long should an incident response plan be?

    About eight to ten pages plus a contact list appendix for a mid-sized organization, and two or three pages for a small one. Roles and authority and the response process take one to two pages each, and most other sections take a page or less. Put a one-page summary first, with how to declare an incident, the severity table, and the key phone numbers.

    Who is responsible for the incident response plan?

    One named owner, usually the security lead or IT manager, maintains the incident response plan. Executives approve it because it assigns decision rights and spending authority, while legal, communications, and privacy leads review their own responsibilities. The owner also updates the plan after every incident and exercise, and reissues the contact list whenever it changes.

    How often should an incident response plan be updated?

    Review the whole plan once a year, and straight after any incident, exercise, new system, new supplier, or change of person in a named role. Check the contact list every quarter, because phone numbers go out of date faster than anything else in the document. Record each review in the version history, which auditors and insurers ask to see.

    What is the difference between an incident response plan and a disaster recovery plan?

    An incident response plan covers stopping the attacker and establishing what happened. A disaster recovery plan covers restoring systems and data after any failure, and it usually assumes that backups and identity systems can be trusted. After an intrusion that assumption may be false, so the incident response plan names who confirms that restoration is safe before disaster recovery begins.

    Does a small organization need an incident response plan?

    Yes. Two or three pages are enough: what to do first, who to call, who decides, how the team talks if email is down, where the backups are, who may have to be told, and support numbers for key suppliers. Name one person with authority to shut systems down and spend money, and one deputy. Print it and keep it where staff can find it.

    Is an incident response plan required by law?

    It depends on the sector. In Australia, APRA’s prudential standard CPS 234 requires regulated banks, insurers, and superannuation funds to maintain information security response plans and to review and test them every year. Other organizations meet the expectation indirectly. Privacy regulators look at whether reasonable steps were taken, and cyber insurers and customer contracts commonly ask for a documented and tested plan.

    What is the difference between an incident response plan and a playbook?

    The plan sets roles, decision authority, severity levels, and communication for any incident, and it changes when the organization changes. A playbook gives the steps for one incident type, such as ransomware or a compromised mailbox, from trigger to closure. The plan lists the playbooks and stays short by leaving their detail out.