← All posts

Incident Response Workflow for Irish SMEs: Steps, Roles, Escalation

If a customer, insurer or regulator has asked how you would handle a cyber incident, the short answer is a workflow with five stages: prepare, detect and analyse, contain, eradicate and recover, then review. Each stage needs a named owner, a written trigger and a record. The same five stages are used in the National Cyber Emergency Plan from NCSC Ireland (NCSC-IE), the national cyber security centre.

This guide is for the job of handling an incident inside the regulator's clock. It is written for IT managers, operations directors and compliance leads in businesses of roughly 10 to 250 people who do not have a security team. Accountants and managed service providers advising such businesses will find the roles and escalation sections useful too.

Key Takeaways

  • A workable workflow is five stages: prepare, detect and analyse, contain, eradicate and recover, review. Most of the effort happens before anything goes wrong.
  • Decide in advance who declares an incident, who can take a system offline and who talks to the regulator. Unclear ownership costs more time than any technical fault.
  • If NIS2 applies to you, the clock is 24 hours for an early warning, 72 hours for a fuller notification and one month for a final report, all counted from when you become aware.
  • Keep a log from the first minute. It serves the regulator, your insurer, your board and your next audit.
  • Test the workflow with a short exercise at least once a year and write down what you change.

Table of Contents

Event, incident, emergency: the words matter

An event is any observable occurrence in a network or system: a login, a blocked email, a failed backup. Most events are harmless. NIST defines a cyber incident as an occurrence that actually or imminently jeopardises the confidentiality, integrity or availability of information or a system, or that breaches security policy or the law.

In practice: a spam email is an event. An employee entering their password on a fake sign-in page is an incident. A routine IT fault, such as a printer outage, is a service issue, not a security incident, unless there is a sign it was caused deliberately or exposes data.

Writing this distinction into your plan matters because it decides who gets woken up. Many teams use priority levels, commonly P1 to P4, to express it. A simple version for a small business:

Priority What it looks like Who is told Target response
P1 Ransomware, confirmed data theft, payment fraud in progress Incident lead, managing director, IT provider, insurer Immediately
P2 Compromised account, malware on one device, lost unencrypted laptop Incident lead, IT provider Within hours
P3 Phishing email clicked, nothing entered Incident lead Same working day
P4 Blocked attempt, suspicious but contained Logged only Weekly review

The response times are our suggestion for a small business, not a legal standard. Adjust them to your size and hours of cover.

The five stages of the workflow

There are many versions of the lifecycle: four, five or six phases. They describe the same work. NCSC-IE uses five phases: preparation, detection and analysis, containment, eradication and recovery, and post-event activity. We use them here because they are the ones an Irish regulator will recognise.

Five stages of incident response workflow

Stage 1: Preparation (one to two days, once)

This is the stage that decides how the rest goes. Do it when nothing is wrong.

  1. Name the team. The NCSC-IE Cyber Security Baseline Standards ask for a first responder plus a second person for that post, with every team member listed by name, contact details, title and role.
  2. Line up outside help. The same standards ask for a third-party incident response or forensics provider with documented contacts. Agree terms before you need them.
  3. List your critical systems. If you can only restore three things first, which are they? This is what your business continuity plan is for, and our business continuity planning guide covers it.
  4. Make tested backups. The UK NCSC's Response and Recovery Guide advises regular, tested backups and joint responsibility for cover so nobody is a single point of failure.
  5. Keep a contact sheet outside your email. Hosting provider, IT support, cloud provider, insurer's claims line, NCSC-IE, and your legal adviser. Print it. If Microsoft 365 is down, the sheet in the cloud is no use.
  6. Write trigger points. The UK guidance suggests documenting the points at which decision ownership transfers, for example from IT to the managing director when customer data is involved.

For a fuller template, see our guide to building an incident response plan.

Stage 2: Detect and analyse (first 30 to 60 minutes)

Someone notices something. The workflow starts when that person knows exactly where to report it, and that they will not be blamed for reporting a false alarm.

The UK NCSC lists common signs: slow computers, users locked out of accounts, documents that cannot be opened, ransom demands, odd emails sent from your domain, requests for unauthorised payments and unusual account activity. It advises checking antivirus alerts and server or audit logs to find the specifics.

The incident lead then asks a short list of questions, drawn from the NCSC's ten questions:

  • What was reported, and by whom?
  • Which services or devices are not working?
  • Is there any sign of data loss or a ransom request?
  • When did it first happen, and how far has it spread?
  • Did it start inside the business or through a supplier?
  • What is the likely impact on the business?

Record the answers with times. Then assign a priority from the table above and start the clock log. This is the moment you "become aware", which matters for the regulatory deadlines below.

Stage 3: Contain (hours)

Containment limits the damage while you work out the cause. Typical actions are isolating a device from the network, disabling a compromised account, blocking a malicious address and pausing a payment run.

The UK NCSC's incident management guidance points out that containment can mean hard decisions, such as taking a core business system offline, and that attackers may react to what you do. That is why you decide in advance who is allowed to make that call. Do not leave it to be argued at 2am.

Do not wipe or rebuild machines yet. You need the evidence.

Stage 4: Eradicate and recover (hours to days)

Eradication removes the cause: malware, a rogue account, an exploited vulnerability. Recovery brings systems back to normal operation.

CISA's incident playbooks give a useful sequencing rule, even though they were written for US federal agencies: before eradication begins, account for all means of persistent access, contain the adversary's activity and collect all evidence. Eradication and recovery can then run side by side.

In practice for an SME:

  1. Preserve logs, screenshots and affected disk images, or ask your forensic provider to.
  2. Remove the cause and reset credentials, including service accounts.
  3. Restore from a backup you know pre-dates the compromise.
  4. Monitor closely for a week for signs the attacker has returned.
  5. Confirm with the business owners that each service works, and close the incident formally.

If you are unsure whether the attacker is still in, bring in the outside responder rather than guessing.

Stage 5: Review (within two weeks)

Hold a short, blame-free review. Cover what happened, what you noticed late, which decisions were slow and what changes. The UK NCSC's Cyber Assessment Framework expects lessons from incidents, including near misses, to be captured, prioritised, acted on and reported to senior management.

Feed the findings into your risk register and your plan. Reading how others structure failures helps; this piece on AI incident reporting frameworks maps real failures to standards in a way that is useful when your incident involves an AI system.

Roles: who does what

A small business does not need a large team. It needs every duty to have a name and a deputy.

Role Responsibility Typical person
Incident lead Declares the incident, sets priority, keeps the log IT manager or operations director
Deputy lead Takes over when the lead is away Second IT or operations person
Technical responder Containment, evidence, recovery IT provider or MSP
Decision owner Approves shutdowns, payments, public statements Managing director
Communications Customers, staff, press Managing director or marketing lead
Compliance contact Regulator notifications, record of decisions Compliance lead or DPO
Legal and insurer Advice on liability and cover External

NCSC-IE's draft NIS2 risk management guidance (measure RMM014) expects entities to define roles, responsibilities and procedures for handling incidents, and to make sure people in incident handling roles have the competence and experience for them.

Escalation: when and to whom

Escalation is a rule, not a feeling. Write it as conditions. Examples:

  • Any sign of personal data being accessed: escalate to the compliance contact and decision owner immediately.
  • Any system supporting customers offline for more than an hour: escalate to the decision owner.
  • Any ransom demand or payment request: escalate to the decision owner, insurer and legal adviser before anyone replies.
  • Any incident involving a supplier or customer: tell the person who owns that relationship.

Escalation also works downward. When the evidence shows an incident is smaller than feared, record the de-escalation and why. The UK NCSC guidance says tasks and findings should be tracked throughout, with escalation or de-escalation as needed.

The regulator's clock

Three questions decide which clock applies to you: are you in scope of NIS2, does personal data feature, and are you a financial entity.

NIS2. Under Article 23, essential and important entities must send:

  1. An early warning within 24 hours of becoming aware of a significant incident.
  2. An incident notification within 72 hours, with severity, impact and indicators of compromise.
  3. Intermediate reports on request.
  4. A final report within one month, covering the description, likely root cause, mitigation and any cross-border impact.

NCSC-IE's guidance confirms the same 24-hour, 72-hour and one-month pattern for notifications to the national CSIRT, the computer security incident response team. Our NIS2 incident reporting guide walks through what each report should contain.

Irish transposition. The EU deadline for turning NIS2 into national law was 17 October 2024, and William Fry reported in February 2025 that Ireland had missed it and that the National Cyber Security Bill was still to be drafted. Check the NCSC-IE site and the Oireachtas for the current status. Whatever the legal timing, the clocks are the standard customers, insurers and auditors now expect, so build your workflow to meet them. Whether your business is in scope can depend on legal interpretation; take legal advice if you are unsure. Our NIS2 compliance checklist helps you frame the question.

Operators of essential services. Under the earlier NIS Regulations, Regulation 18 requires notification within 72 hours of becoming aware of an incident with a significant impact on service continuity.

Personal data. If personal data is involved, GDPR has its own 72-hour notification to the Data Protection Commission. See our GDPR breach response guide. Financial entities may also owe reports to the Central Bank of Ireland under DORA; confirm the details with your compliance adviser.

NCSC-IE's emergency plan also asks victim organisations to report promptly to the NCSC operations team, report to other authorities as required (for example the Data Protection Commission, An Garda Síochána and the Central Bank), and bring in a specialist response firm if needed.

A simple flowchart you can copy

Draw this on one page and pin it where people will see it:

  1. Something odd is noticed. Report to the incident lead.
  2. Lead logs the time and asks the six questions. Is it an incident? If no, log and close.
  3. If yes, assign P1 to P4 and start the clock log.
  4. Does it involve personal data, a regulated service or a customer? If yes, tell the compliance contact now.
  5. Contain. Preserve evidence.
  6. Early warning due within 24 hours if NIS2 applies; notification within 72 hours.
  7. Eradicate, recover, verify.
  8. Review within two weeks. Update the plan and the risk register.

Testing it

A workflow nobody has tried will fail at the first step. NCSC-IE's draft guidance (measure RMM014.FA08) expects periodic exercises and scenarios to test incident management. Run a 90-minute tabletop twice a year: choose a scenario such as ransomware on the file server or a fake invoice from a supplier, walk through the flowchart and note every point where people hesitated. The UK NCSC also offers a free exercise product, Exercise in a Box, which many Irish businesses use as a starting point. Pair the exercise with staff training; our security awareness guide explains how.

What to do this week

You can do all of this in a week, in about six hours in total.

  1. Monday (1 hour): Name an incident lead and a deputy. Write both on one page.
  2. Tuesday (1 hour): Build the contact sheet and print it.
  3. Wednesday (1 hour): Agree P1 to P4 and the escalation conditions with your managing director.
  4. Thursday (1 hour): Check that your most recent backup restores. Time it.
  5. Friday (2 hours): Run a tabletop exercise with the flowchart above. Write down three fixes.

Then put the fixes in your calendar for the following month.

Our perspective

We have sat in the CISO chair, and the pattern is consistent: the technical response is rarely what goes wrong. What goes wrong is that nobody knows who decides, the contact list is in the system that is down, and the log of what happened when is reconstructed from memory a week later.

For a business with no security team, the answer is not a bigger tool. It is a short, owned, tested workflow that produces evidence as it goes. Evidence is what a customer questionnaire, a board report, an insurer and a regulator all ask for.

That is why ShieldIQ is built for teams without a security team. It helps you assess, prioritise, document and evidence your position. Certification, where you want it, comes from an accredited body, not from us.

Matthew Lemon

Where ShieldIQ fits

ShieldIQ is an AI-powered governance, risk and compliance platform built in Ireland, with customer data hosted in the EU (Ireland) on AWS. Its incident workflows help you log incidents, track the regulatory clocks and keep the evidence in one place, and its AI policy drafter produces a first version of your incident response policy for you to review. A first baseline assessment takes about 15 minutes, and the NIST CSF 2.0 assessment is free; the other frameworks, including NIS2, GDPR, ISO 27001 and DORA, are on paid plans. See plans and pricing or the NIS2 page.

ShieldIQ

If you want a person alongside the software, our consulting team offers incident response preparation and virtual CISO support.

FAQ

What is the difference between an incident and an event?

An event is any observable occurrence on your systems, and most are harmless. An incident is an occurrence that actually or imminently harms the confidentiality, integrity or availability of information or systems, or breaches policy or law, using the NIST definition. Your plan should say who decides when an event becomes an incident.

How many phases does the incident response lifecycle have?

It depends on the source, with four, five and six phase versions in common use. NCSC-IE uses five: preparation, detection and analysis, containment, eradication and recovery, and post-event activity. The content is the same, so pick one and use it consistently.

Do I have to report every incident to the regulator?

No. NIS2 reporting applies to significant incidents affecting entities in scope, and GDPR applies where personal data is breached and there is a risk to individuals. Whether a particular incident meets those thresholds can need legal interpretation, so involve your legal adviser early.

Who should be on an incident response team in a small business?

At minimum an incident lead and deputy, a technical responder (often your IT provider), a decision owner at management level, a communications contact and a compliance contact. NCSC-IE's baseline standards ask for named people with contact details.

Sources

Recommended