Incident response and recovery
A cyber incident may be happening. Start here.
Something looks wrong. What should you do first? A calm, practical guide for small businesses and IT teams: get help, limit the damage and keep a clear record.
KoshCyber · · · 10 min read
A colleague has clicked a suspicious link. Someone cannot open their files. An unexpected sign-in alert appears. Your first thought might be: is this a cyber incident, and what do we do now?
You do not need all the answers before asking for help. Whether you run a small business, look after IT or are simply the first person to notice something unusual, start with three priorities: get the right people involved, reduce further harm and keep a record of what happened. This guide walks you through those first steps.
Start here: five things to do now
- Stop interacting with the suspicious item. Do not open more attachments, follow links or approve unexpected sign-in prompts.
- Contact someone who can help. Use a known phone number for your IT team, IT provider or incident lead, rather than contact details in the suspicious message.
- Write down what you noticed. Include the time, affected device or account, the message or alert and anything you have already done.
- Agree how to limit further harm. Let an authorised responder guide device isolation or account restrictions; follow your existing emergency procedure if damage is spreading.
- Tell affected colleagues what to do next. Give clear instructions and an agreed time for the next update, without guessing about the cause.
1. Put one person in charge of coordinating
Choose one person to coordinate the response. In a small business, this might be the owner or office manager working with an IT provider. They do not need to do every technical task: their job is to make sure people know who is doing what and which decisions need approval.
Start a simple record in an approved, access-controlled document or incident tool. Note who reported the problem, what they saw, when they saw it and what has already been done. Keep facts separate from guesses. “A colleague received an unexpected sign-in alert” is a fact; “all our accounts have been hacked” is not.
2. Limit the damage without making things worse
It is tempting to switch everything off or reset every password. Pause before making broad changes: they can interrupt essential services and make it harder to understand what happened. Follow your incident plan, if you have one, and ask an authorised responder which action will reduce the immediate risk.
- If email may be affected, use a known phone number or another agreed, trusted way to contact your responders.
- Keep the original suspicious message or alert. Note filenames and times where safe; do not forward risky attachments to colleagues.
- Do not delete files, run cleanup tools or reset systems before agreeing the response with your IT team.
- Write down each action, who approved it and why. Even a short note helps the next person.
For the IT team: choose a targeted action
Containment means limiting further harm. Before disconnecting a device or restricting an account, ask: what could this stop, what essential work could it interrupt and what evidence might be lost? Agree who can approve the action and how you will check whether it worked.
- An affected laptop or computer: consider controlled network isolation, but first check whether it supports an essential or safety-critical service. Network isolation is not the same as powering the device off.
- An affected account: ask the identity or email administrator to review access, active sessions and related activity. Changing a password alone may not remove all unauthorised access.
- Backups or administration systems: involve their owners early and protect recovery options. Do not experiment on the systems you may need to restore the business.
- Damage is still spreading: follow your emergency approval and escalation arrangements. Do not wait to understand the entire attack before limiting harm; preserve what you safely can and record why you acted.
3. Work out what is affected — and what is not yet known
Start small. Which person, account, device or service is involved? What can they no longer do? Is the problem spreading? Ask your IT team to check the evidence rather than assuming the first alert tells the whole story. It is fine to say “we are still checking”.
Useful questions for common warning signs
| What you noticed | What to ask | Why it helps |
|---|---|---|
| Unexpected sign-in or account activity | Which account, service, time and source? Has the account owner confirmed the activity through a trusted channel? Are other accounts involved? | Whether to escalate identity investigation and restrict affected access. |
| Unavailable or apparently encrypted files | Which business services depend on them? Is disruption spreading? Are backup and management services still available? | Whether to activate continuity arrangements and prioritise containment. |
| Possible exposure of customer or staff information | What information might be involved? Is there evidence of access or disclosure, or only a claim? Who owns the privacy assessment? | Which evidence and qualified advice are needed for reporting and communications. |
| An isolated security alert | What generated it? Can the underlying event be retrieved? Is there related activity before or after the alert? | Whether to close with evidence, investigate further or activate the incident team. |
Record both when the suspicious activity happened and when you first learned about it. They may be different. You may discover that more — or fewer — systems are affected than you first thought. Update the record as you learn, with a short explanation of what changed.
4. Get help early — you do not have to handle this alone
Contact your internal IT team or external IT provider. If you have cyber insurance, check the policy’s incident contact and any approval requirements before commissioning outside work. Involve whoever looks after legal or privacy matters if customer or staff information may be affected.
Agree who will investigate, who will keep essential work running and who will update staff or customers. These may be different people. If you do not have an incident plan, start with these roles and a trusted contact list. Check any reporting duties with qualified advisers rather than deciding from an early guess.
5. Check it is safe before getting back to normal
A service working again does not necessarily mean the incident is over. With your IT team, check how the problem started as far as the evidence allows, remove unauthorised access and confirm that backups and restored systems are safe to use. Keep monitoring for further activity. Afterwards, agree what needs improving and who will do it.
What could the first hour look like?
Use this as a guide, not a stopwatch. If people are at risk or damage is spreading, get urgent help straight away. With enough people, some tasks can happen at the same time.
| Period | Priority | Useful output |
|---|---|---|
| First 15 minutes | Check the report, choose a coordinator and look for immediate danger. | Someone is leading, the issue is recorded and help has been contacted. |
| Next 15 minutes | Check what is affected and agree actions to limit the damage. | A short list of facts, unanswered questions and who is taking each action. |
| Remainder of the hour | Bring in specialist help, keep essential work going and update people. | A clear update, safely stored evidence and an agreed time to review progress. |
Keep useful evidence, not just screenshots
- Record the original alert or report, affected host or account, source system, timestamp and time zone. Distinguish the event time from the time it was reported.
- Preserve original exports where possible. Record who collected them, when and how, and keep working notes separate from originals.
- Store incident material in an access-controlled location and share it only with authorised responders. Do not paste passwords, tokens or sensitive customer records into general chat.
- If collection itself could disrupt a service or alter evidence, ask the incident lead or a qualified responder before proceeding.
Make it easy for the next person to take over
When another colleague or specialist takes over, tell them who is leading, how the business is affected, what is confirmed and what is still being checked. Share where the evidence is stored and which actions have already been approved or completed. Finish with the next decision needed and the next update time. This helps avoid repeating work or undoing someone else’s actions.
What should you write down?
You do not need a polished report while you are responding. Short, clear notes are enough to start. Use an approved incident tool or a document restricted to the right people. Link to securely stored evidence rather than copying sensitive information into every update.
- Who is involved: incident reference, coordinator, who can approve decisions and how to reach them.
- When it happened: event time, report time, time zone and when you will review the situation.
- What is affected: services, accounts or devices, the effect on the business and what remains uncertain.
- Where the evidence is: who collected it, when, from which system and where it is safely stored.
- What you have done: each action, why it was taken, who approved it and whether it helped.
- What you need next: supplier support, recovery needs, updates to people and any reporting decisions.
Copy this simple update template
Use these prompts for an internal update. Fill in only what you know, mark anything still being checked and get the right approval before sharing outside the response team.
Incident reference and update time:
Who is coordinating and how to contact them:
What we know:
How the business is affected:
What we are still checking:
Actions taken and whether they helped:
Decisions needed, who will make them and by when:
Where evidence is stored and what is still missing:
What staff or service owners should do now:
Next update time:For example: “We are investigating access to one staff account. We have not yet confirmed whether information was accessed. The identity team is reviewing the account and related activity. Staff should report unexpected sign-in prompts through the approved service-desk route. The next update will be at the agreed review time.” This communicates uncertainty without claiming that there has been no data loss.
Do we need to report this?
Not every security alert needs to be reported to a regulator. If personal information may be involved, ask your privacy or legal adviser to assess what happened, the risk to people and whether notification is required. Record the decision and its reasons. Customer contracts, your sector or other laws may introduce additional duties.
For organisations subject to UK GDPR, the ICO states that a notifiable personal data breach must be reported without undue delay and within 72 hours of becoming aware of it, where feasible. Where a breach is likely to result in a high risk to individuals’ rights and freedoms, those individuals must also be informed without undue delay. Keep a record of personal data breaches even when notification is not required.
Putting it together: an unexpected sign-in alert
A colleague reports an unexpected sign-in alert. Start by preserving the original alert and verifying it through the service’s trusted administration portal, rather than following a link in the message. Contact the colleague through a known-good route and ask whether the timing and activity are expected.
If you cannot confirm a legitimate explanation, ask an authorised administrator to investigate. Record the account, relevant times, related activity and any essential services that depend on it, then agree how to restrict access if needed. Check other accounts when the evidence gives you a reason to do so; one alert does not mean every system has been compromised.
When handing over, explain what was confirmed, which access was restricted, what evidence was kept and what still needs checking. If the alert stops, that is useful information — but it does not, on its own, prove that the problem is resolved.
Take the next step with KoshCyber
You do not have to tackle every part of the response at once. Use the links below to choose a response playbook for the situation, follow the Incident Assist stages or find investigation tools for your IT team. Keep sensitive evidence in your organisation’s approved storage, and involve qualified responders when you need them.
Official sources
External guidance may change. Check the source and your own policies before acting.