Skip to content
EncryptTechBD
Home Industries Our Work Blog Pricing About Contact Discuss Your Project
বাংলা (BN)
Research & Resources 6 min read

Cyber incident response checklist for small businesses

A duty-manager checklist for the first hours of a cyber incident: authority, containment, evidence, recovery, communication and review.

E
EncryptTechBD · 0 comments
Colleagues discussing research at a table
Photo: Direct Media / StockSnap

Most small businesses have an incident response plan in a folder. Few have a checklist that a duty manager can follow at 2 a.m. when a laptop shows a ransom note or a payment dashboard shows an unexplained transfer. The difference is not length. It is knowing who decides, what to preserve, and what to do first.

This checklist is built for a business without a security team. It assumes you have a handful of IT staff or a managed provider, a few critical systems, and no appetite for a six-month consulting project. Work through it before an incident, then keep a printed copy where people can find it when email and chat are unavailable.

1. Prepare the authority, not just the document

  • Name an incident lead and a deputy. The lead coordinates; the technical team investigates. One person must own the decision log.
  • Give explicit authority. Decide in advance who can shut down a system, approve emergency spending, contact the bank and speak to customers.
  • Keep an offline contact sheet. Include the managed service provider, bank relationship manager, payment provider, legal counsel, insurer and an alternate for each.
  • Pre-approve a small emergency budget. A delayed forensic or recovery purchase is often more expensive than the purchase itself.
  • Decide what you will not do. For example, do not negotiate with a ransomware operator without legal advice, and do not wipe a machine before evidence is captured.
  • Review the plan every quarter. A contact who left six months ago is not a contact.

2. Define what counts as an incident

Write down the triggers so people do not wait for certainty. Common triggers include a ransom note, unauthorised administrator access, a large outbound data transfer, a payment destination change, website defacement, a lost laptop with customer data, and a vendor reporting that your account was accessed.

Use a simple severity scale. A critical incident threatens money, customer data or a core service. A medium incident affects one team or one non-critical system. A low incident is contained and has no data or financial impact. The label decides who is called and how often updates are issued.

3. Detect and declare

  • Record the time the alert or report arrived, in one shared timeline.
  • Ask three questions: what system, what data, and what is still happening?
  • Assign a scribe. The incident lead should not be the person taking notes.
  • Declare early. Downgrading a false alarm is cheaper than explaining a late response.
  • Do not post details in a wide chat channel. Use a named incident channel with an explicit member list.

4. Contain without destroying evidence

  • Isolate, do not power off. Disconnect the host from the network or move it to a quarantine VLAN. Memory and temporary files can disappear on shutdown.
  • Disable compromised accounts and revoke sessions. Changing a password is not enough if active tokens remain valid.
  • Rotate credentials from a clean device. Include email, VPN, cloud console, domain administrator, payment portal and backup accounts.
  • Block known indicators at the edge. Add malicious domains and addresses to the firewall, proxy or DNS filter.
  • Preserve logs. Export authentication, endpoint, firewall, email and SaaS audit logs before retention windows roll over.
  • Protect one clean backup. Confirm it predates the compromise and is not reachable with the compromised credentials.

5. Investigate the scope

Establish how the attacker entered, what they touched, and whether they are still present. Check identity logs, endpoint telemetry, network flows, email rules, SaaS integrations and payment beneficiary changes. Keep a chain-of-custody record for any device or image you hand to a specialist. Interview the people who reported the issue separately, and write down what they saw before memories blur.

6. Eradicate and recover

  • Rebuild compromised systems from a trusted image rather than cleaning them in place.
  • Patch the entry point, including the plugin, VPN appliance or exposed service that was missed.
  • Restore data from a backup taken before the compromise, then validate the restore.
  • Reset credentials again after eradication, including service accounts and API keys.
  • Return systems in phases, starting with the least risky, and watch closely for renewed activity.

7. Communicate on a schedule

Appoint one spokesperson. Issue internal updates at fixed times, even when there is nothing new, so rumours do not fill the gap. Tell customers what happened, what they should do, and what you will do next, without speculation about the attacker or the final scope. Involve legal counsel before making regulatory or public statements. In Bangladesh, if payments or personal data are affected, bring in your bank or mobile financial services provider and take advice on sector rules and data protection duties.

8. Run an after-action review

Within two weeks, produce a timeline, a root-cause statement, the controls that failed, and an action list with owners and dates. Keep the review blameless; the goal is a better system, not a culprit. Re-test the restored environment, update the contact sheet, and schedule the next tabletop exercise.

Evidence to keep

  • Incident log with timestamps and decisions.
  • Chain-of-custody forms for devices and images.
  • Communication log, including who was told what and when.
  • Recovery validation record for each restored system.
  • After-action report and the tracked action list.

Common mistakes

  • Reimaging a machine before capturing evidence.
  • Resetting one password while active sessions and API keys remain valid.
  • Restoring from a backup that was already infected.
  • Letting the technical lead also handle customers and the press.
  • Waiting for complete certainty before declaring an incident.
  • Closing the case without fixing the control that failed.

Metrics

  • Time from detection to declaration.
  • Time from declaration to containment.
  • Time to restore each critical service.
  • Percentage of critical systems covered by a tested plan.
  • Overdue actions from after-action reviews.

Where this fits in our work

A checklist is only useful if the underlying controls exist: logging, backups, identity, endpoint protection and a tested recovery path. Our Managed Technology & Cybersecurity division helps put those pieces in place; the pricing page explains how an engagement is scoped. For the budget side of the same problem, read what an incident response plan costs.

Sources

Services that deliver this

Want this implemented properly?

Pricing →

Every engagement starts with a written scope. If a requirement is outside it, we quote it separately before starting.

Responses (0)

No comments yet. Add the first one below.

Leave a response

Your email address will not be published. Required fields are marked *.

Keep reading

More field notes