Incident response for SMEs: handling security incidents in a structured way

A security incident never arrives at a good time. If you only work out who decides, who informs and who reports once the emergency is under way, you lose hours. Incident response means a fixed procedure from the first report to the lessons learned, prepared in playbooks and rehearsed around the table. This article walks through the phases and shows how CyberTell maps them for SMEs.

8 min read CyberTell
Key takeaways
  • Incident response follows fixed phases: report, assess, contain, recover, review.
  • One ticket per security incident and a war room for coordination keep decisions traceable.
  • Reporting duties depend on the case: the FDPIC when affected persons face a high risk, the BACS within 24 hours only for operators of critical infrastructure.
  • Playbooks and tabletop exercises prepare the team, and lessons learned improve the playbook after every run.

Incident response for SMEs: why the procedure must be set in advance

In an emergency there is no time for basic questions. An employee reports a suspicious email, shortly afterwards an account is locked, and management wants to know whether customer data is affected. Without an agreed procedure, whoever happens to be available makes the call.

The Swiss economy consists almost entirely of SMEs: 99.7% of companies fall into this category (FSO, STATENT 2023). That makes a procedure that a small team can follow under pressure all the more important.

In 2025 the Federal Office for Cybersecurity (BACS) received 64,733 reports of cyber incidents. More than half concerned fraud, and just under a fifth phishing (BACS semi-annual reports 2025). For an SME it therefore pays to define the procedure before it is needed.

The phases of incident handling

Common incident response models, such as the one from NIST, divide an incident into a few phases. The names differ, but the core stays the same. For an SME, five steps that everyone on the team knows are enough.

Order matters: reinstalling straight away destroys the traces of the attack.

  1. Report and record

    Every suspicion is recorded as a ticket: what happened, when, and who noticed it. Employees report easily and without jargon.

  2. Triage

    Set priority and category and assign an owner. The key question is whether data, accounts or systems are affected.

  3. Contain

    Limit the damage: lock the account, take the device off the network, block the domain. Preserve evidence before anything is deleted.

  4. Eradicate and recover

    Remove the cause, restore systems from verified backups and resume operations in a controlled way.

  5. Lessons learned

    Review the procedure: what worked, what did not, which measures follow. The findings feed into the playbook.

Incident Management Sample data
Overview of incident tickets in CyberTell incident management Overview of incident tickets in CyberTell incident management
Overview of incident tickets in CyberTell incident management

Reporting and triage: one ticket per security incident

The first step shapes everything that follows. An incident that is not reported cannot be handled. Reporting must therefore be simple, and anyone who clicked a link should be able to say so without fear of blame.

In CyberTell incident management, every report becomes a ticket with:

  • status (open, in progress, closed), priority from low to critical, category and damage type
  • assignment to an owner
  • message thread with edit history
  • links to the war room and to MITRE ATT&CK techniques

Employees see their own reports together with the team's replies. For suspicious emails, the phishing analysis helps: you upload the email as an EML file, CyberTell checks SPF, DKIM and DMARC, matches links and attachments against local blocklists and returns an AI assessment with a risk score from 0 to 100. The team makes the decision. On a hit, you can open a ticket directly.

How reliably your employees recognise and report suspicious emails is shown by a phishing simulation.

Incident Management Sample data
Incident ticket with history, priority and assignment Incident ticket with history, priority and assignment
Incident ticket with history, priority and assignment

Contain and investigate

Once it is clear that you are dealing with a real security incident, the focus shifts to limiting the damage. No traces may be lost along the way: log files, suspicious emails and timestamps must be secured before anyone cleans up.

For larger cases, you open an investigation in CyberTell. Its status follows the NIST phases: active, contained, eradicated, closed. Per-scenario guides lead through scope, evidence preservation, reporting and recovery. The investigation is linked to tickets, the war room and the playbook, so everyone involved sees the same picture.

What good containment looks like

  • record every measure with the time and the person responsible
  • limit first, analyse second
  • recover only once the cause is known
Incident Management Sample data
Investigation of a security incident with IoC graph Investigation of a security incident with IoC graph
Investigation of a security incident with IoC graph

Communication in the war room

In a larger incident, IT, management and often external service providers work in parallel. If coordination runs through private chats, phone calls and email chains, in the end nobody knows what was decided and when.

The war room brings crisis communication together in one place:

  • team chat with file attachments, linked to the incident tickets
  • participant management per war room
  • status active, paused or closed
  • push and email notification for unread messages
  • chronological history as an audit trail
  • access through the CyberTell apps, also on the move

When a playbook runs in the war room, its start and completion appear in the chat. Individual steps can be assigned to people and, if needed, continued as a separate ticket. The history records who decided what and when, the basis for the review and for any reports.

Incident Management Sample data
Active war room with team chat and a running playbook Active war room with team chat and a running playbook
Active war room with team chat and a running playbook

Reporting duties for a security incident in Switzerland

Whether and to whom you must report an incident depends on the case. The main rules at a glance:

  • Data Protection Act (Art. 24 FADP): If a data security breach is likely to result in a high risk for the persons concerned, it must be reported to the FDPIC as soon as possible. The law sets no fixed deadline in hours.
  • Information Security Act (Art. 74a ff. ISA): Since 1 April 2025, operators of critical infrastructure must report cyberattacks to the BACS within 24 hours of discovery. Since 1 October 2025, violations can be fined.
  • GDPR (Art. 33): If the GDPR applies, for instance because you have customers in the EU, a 72-hour deadline applies towards the supervisory authority.

Even without an obligation, an SME can report an incident to the BACS voluntarily. In the second half of 2025, the BACS received 145 incidents subject to reporting, including 57 ransomware incidents (BACS semi-annual report 2025/2).

In CyberTell playbooks, these reports are fixed building blocks, not AI text. A case check per incident shows what applies: FDPIC, informing affected persons, GDPR, BACS, insurance or criminal complaint. The texts are guidance and do not replace legal advice.

Preparing with playbooks and tabletop exercises

A procedure that exists only in one person's head is of little help in an emergency. Playbooks record what to do for which kind of attack, with tickable steps, roles and time marks.

Playbooks

Under AI playbooks & tabletop you choose a scenario, for example ransomware, phishing or a data leak. The AI writes the steps as a draft in compact, standard or detailed scope, tailored to your company profile and industry. You review and edit every step in the editor. Roles can be internal or external, such as the IT service provider or the insurer. If the portal is unavailable, the playbook is ready as a PDF, as a Word file and as a form to tick off on paper.

Tabletop exercises

In a tabletop exercise, the team plays through an incident around the table without touching real systems. CyberTell generates scenarios from eleven types at three difficulty levels, with decision points and open questions. The AI rates free-text answers on three levels, and the review is available as a debrief PDF.

Lessons learned

After every playbook run, whether a dry run or a real incident, CyberTell creates a post-incident review as a draft, with a timeline of the completed steps. From the findings, the AI proposes changes to the playbook. Only what you select is applied.

Whether your company's addresses have already appeared in data leaks is shown by the free threat analysis, no registration needed.

Incident Management Sample data
Incident response playbook with case check and tickable steps Incident response playbook with case check and tickable steps
Incident response playbook with case check and tickable steps

Frequently asked questions

What does incident response mean?
Incident response is the orderly reaction to a security incident. It covers reporting, assessment, containment, recovery and review. The aim is to limit the damage and document decisions in a traceable way.
Does an SME have to report a cyberattack?
It depends on the case. If personal data is affected and there is likely a high risk for the persons concerned, the Data Protection Act requires a report to the FDPIC as soon as possible. The 24-hour reporting duty to the BACS has applied to operators of critical infrastructure since April 2025. A voluntary report to the BACS is always possible.
What belongs in an incident response playbook?
A playbook describes, for one scenario, the steps in order, who is responsible and by when. It includes the contacts of the IT service provider and the insurer, the reporting channels and the criteria for resuming operations. It should also be available offline.
How often should an SME run a tabletop exercise?
There is no fixed rule. A fixed date each year has proven useful, plus an additional exercise after major changes to the team, systems or service providers. Every exercise should end with concrete improvements to the playbook.
What is a war room?
A war room is the shared crisis room for the duration of an incident. Everyone involved communicates there instead of through scattered chats and emails. The history documents decisions and later serves as the basis for the lessons learned.
Topics