Incident response

Writing an incident response plan your team will actually use

A practical incident response plan for small businesses: roles, contacts, a simple decision tree and a one-hour tabletop exercise to test it.

Somewhere on your shared drive there may be a document called "Incident response plan". It's 40 pages long, it was written for an audit, and the IT contact in it left two years ago. On the day your server shows a ransom note, nobody will open it. And if the server is encrypted, nobody will be able to.

A plan that works in a crisis looks very different. It fits on a few pages, it names real people, it tells them which decisions are theirs, and it has been practised at least once around a table. This article shows how to write that version and how to test it in an hour.

What the plan is for

An incident response plan answers three questions before you need them: who decides, who do we call, and what do we do first. Everything else is detail.

That matters because the first hours of an incident are full of decisions under pressure: disconnect the server or not, call the insurer first or the IT provider, tell customers now or later. Our first-hour ransomware checklist shows how many of those choices come up in the first 60 minutes. A plan means they've been thought through on a calm day.

It's also increasingly expected. The NIS2 Directive lists incident handling, business continuity and crisis management among the security measures in-scope organisations must take, and Belgium's CyberFundamentals framework includes a requirement that a response plan is maintained, tested and executed. Customers sending security questionnaires ask about it too.

Keep it short and keep it reachable

The CCB and the Cyber Security Coalition's Cyber Security Incident Management Guide is a good free reference for the full picture. It also makes a point that's easy to skip: top management validates the plan and stays involved at every step. A plan that only IT has seen isn't a company plan.

For a business of 10 to 250 people, the working version should be short enough to read in ten minutes. Put background and policy in an appendix. The front pages are for action.

Store the plan where you can reach it when your systems can't: a printed copy at the office and at home for the key people, and a copy outside your own Microsoft 365 tenant.

Email, Teams and the shared drive may all be down or under the attacker's control. Plan for a phone-only first hour.

Roles: names, not job titles

Give each role a named person and a named backup. In a small company, one person may hold two roles. That's fine, as long as everybody knows it.

  • Incident lead: coordinates the response and keeps the timeline. Often the operations manager or the IT contact, not the CEO.
  • Decision maker: a director who can approve spending, shutting down systems, calling in outside experts and talking to the press. The UK's National Cyber Security Centre recommends deciding in advance at what point senior management gets involved.
  • Technical contact: your internal IT person or your IT provider, who contains and investigates.
  • Communications: one person who speaks to staff, customers and suppliers, so the message stays consistent.
  • Privacy contact: your DPO or whoever handles GDPR, who judges whether personal data is involved.

What to do:

  • Write the backup name next to every role. Incidents start on Friday evenings and during holidays.
  • Agree who may authorise an emergency spend, and up to what limit, so nobody waits for a signature.

The contact list

This is the page people will actually use. The NCSC's small business guide suggests listing the external people who can help you, such as your IT support provider, and checking the details every couple of months.

Include mobile numbers, not just switchboards, for:

  • your IT provider and any incident response partner, with the contract or customer number
  • your cyber insurer's incident line and policy number (many policies expect you to call them before hiring anyone)
  • your bank's fraud line, for payment fraud
  • your lawyer
  • the CCB, which handles incident reports at cert@cert.be, and its NIS2 notification platform notif.safeonweb.be if you're in scope
  • your local police, for a criminal complaint
  • the people in your company who can sign in to the Data Protection Authority's notification portal

A decision tree, not a manual

Most small business incidents fall into a handful of types. For each, write the first three or four actions on half a page. A starting set:

Situation First actions
Ransom note or encrypted files Disconnect affected devices, keep them running, protect backups, call IT provider and insurer
A mailbox has been taken over Block sign-in, revoke sessions, reset password, check forwarding rules and sent items
A payment went to a fraudster's account Call your bank immediately, then the police, then check the mailbox involved
Lost or stolen laptop or phone Check it's encrypted and managed, remotely wipe or retire it, reset the user's password
Customer data sent to the wrong person or leaked Contain it, record what was exposed, involve the privacy contact

Then add the questions that start legal clocks. Is personal data affected? If yes, GDPR may require a notification to the Data Protection Authority within 72 hours. Is the organisation in scope of NIS2 and is the incident significant? Then an early warning to the CCB is due within 24 hours. Our NIS2 guide explains who is in scope, and the ransomware article covers both deadlines in more detail.

Finally, give the plan a simple severity scale. Something like: low (one user, no data at risk, IT handles it), medium (several users or possible data exposure, incident lead takes over) and high (business stopped or data stolen, decision maker and outside help immediately). The scale's job is to tell people when to escalate.

Test it with a one-hour tabletop exercise

A tabletop exercise is a meeting where you walk through a made-up incident and talk about what you'd do. Nothing is switched off. It's the cheapest way to find the gaps, and it's where most plans change.

You don't need to write a scenario from scratch. The NCSC's Exercise in a Box is a free online tool for practising your response to a cyber attack, and you don't need to be an expert to use it. The US cybersecurity agency CISA publishes tabletop exercise packages with template scenarios and discussion questions.

A simple format:

  1. Invite the people named in your roles, including a director. Book an hour.
  2. Set the scene: "It's Monday at 7.40. The accounts team can't open the shared drive and there's a ransom note on the server."
  3. Add an update every ten minutes: the backup drive turns out to be encrypted too, a journalist calls, a customer asks whether their data is safe.
  4. At each step ask: who decides, who do we call, and where is that written down?
  5. Write down every gap and give each one an owner and a date.

Plans fail in predictable ways: nobody has the insurer's number, the only admin account belongs to someone on holiday, nobody knows who can sign in to the regulator's portal. An exercise finds these in a meeting room instead of during a real attack.

Repeat it once a year and after any big change, such as a new IT provider or a move to a new system. The US standards body NIST's updated incident response guidance, SP 800-61 Revision 3, also treats lessons learned from incidents as input for improving your security, not just for closing a ticket. Every real incident, however small, is a reason to update the plan.

Where to start

  1. Name the roles this week: incident lead, decision maker, technical contact, communications and privacy, each with a backup.
  2. Build the contact list and print it. Check that your insurer's incident line and policy number are on it.
  3. Write half a page for each of the five common situations in the table above.
  4. Have a director sign it off, so the plan carries authority on a bad day.
  5. Book a one-hour tabletop exercise within the next three months, and update the plan afterwards.

The takeaway

The plan that helps in a crisis is short, names real people and lives on paper as well as on the server. You can write a first version in an afternoon and test it in an hour around a table. Start with the contact list: if that's all you have when something goes wrong, you're already ahead.

This article is general information, not legal advice. Check the CCB's incident management guidance and the Data Protection Authority for the latest official guidance.

← All articles

Not sure where you stand?

Find your gaps before an attacker does

A free, no-obligation assessment maps your risks and gives you a prioritised plan, whether or not you work with us.