Breach preparation often fails before the incident begins. The organisation has a policy, a contact list and perhaps a tabletop exercise, but it has not decided who can declare an incident, which logs are retained, how a customer list will be notified or what happens when the person who knows the backup process is on leave. These are not technical gaps. They are risk decisions that were never made.
This guide treats breach response preparation as a risk register rather than a policy document. For each risk it sets out why it matters, a qualitative likelihood and impact, the control that changes the outcome and the residual risk that remains. The ratings are deliberately relative; they should be recalibrated against your own estate.
The risk is in the preparation
A data breach is not a single event. It is a sequence: discovery, containment, evidence preservation, scope assessment, notification decisions, communication, recovery and review. Each step has a deadline and each deadline depends on information that must exist beforehand. Preparation is the work of making that information available under pressure.
Regulators and customers judge the response on what was done, how quickly and how clearly. A slow answer is often caused by missing records, unclear ownership or an approval chain that was never tested.
A simple likelihood and impact method
Use three levels for each. Likelihood reflects exposure and the presence of controls: high if the scenario is plausible within a year, medium within two years, low otherwise. Impact reflects harm to customers, operations, cash, legal obligations and reputation. A single high-impact scenario deserves attention even if its likelihood is medium.
Record the reasoning beside each rating. A rating without a reason cannot be reviewed, and it will be argued about in the incident rather than before it.
The risk register
| Risk | Likelihood | Impact | Primary control | Residual risk |
|---|---|---|---|---|
| Personal data inventory is incomplete, so scope cannot be established | High | High | Maintained record of systems, datasets, owners and retention | Medium; inventory ages without a quarterly review |
| No clear incident owner or decision rights | Medium | High | Named incident lead, deputy and activation threshold | Low if deputies are exercised |
| Logs are overwritten or scattered before evidence is collected | High | Medium | Central logging, defined retention and a preservation procedure | Medium; cost limits retention |
| Notification decision is late, inconsistent or overbroad | Medium | High | Decision matrix with legal, regulator and customer criteria | Medium; facts arrive gradually |
| Rumour fills the communications gap | High | Medium | Pre-approved holding statement, spokesperson and channel plan | Low if reviewed annually |
| Vendor or processor discovers the issue first and reports late | Medium | High | Processor register, contractual notice terms and tested contacts | Medium; depends on third parties |
| Backups are compromised or restore fails during recovery | Medium | High | Immutable copies, offline retention and scheduled restore tests | Low to medium with evidence |
| Credential reset is incomplete after containment | High | High | Identity inventory, session revocation and re-authentication runbook | Medium; legacy accounts persist |
Controls that change the outcome
Know where personal data lives
Build the inventory from systems and business processes, not from a policy template. Include spreadsheets, shared drives, marketing tools, HR records and supplier portals. Record the owner, the retention rule and the lawful basis. Review quarterly. The inventory is the first thing the response team needs and the first thing that is out of date.
Decide who can declare an incident
Name the incident lead, a deputy and the threshold for activation. Give them authority to isolate systems, engage forensics and approve emergency spend up to a stated limit. Without that authority, the first hour becomes a meeting.
Preserve evidence before rebuilding
Define what to capture: logs, disk images, identity records, email headers, cloud audit trails and network flows. Store copies in a separate account or location. Rebuilding a server before preservation destroys the evidence needed to determine scope and notification obligations.
Prepare the notification decision
Work with legal counsel to define the criteria, the decision maker, the regulator contacts and the customer channel. Prepare templates for holding statements, customer notices and internal updates. Do not wait for an incident to discover that the approval chain includes someone who has left.
Test the supplier path
List processors and critical suppliers, record their security contact and contractual notice period, and test one contact each year. A processor that discovers a breach on Friday and tells you on Monday has consumed your response window.
Residual risk and acceptance
Every control leaves something behind. Log retention may be limited by cost. A legacy system may not support modern authentication. A small team may not have forensic skills in-house. Record the residual risk, the owner who accepts it and the date of the next review. Acceptance is a decision, not a failure; unrecorded acceptance is the failure.
Board questions
- Which personal data categories do we hold, where, and who owns each one?
- Who can declare an incident, and who is the deputy?
- How long do logs survive, and when was a restore last tested?
- Which suppliers could discover a breach before we do, and when must they tell us?
- What have we rehearsed in the last twelve months, and what changed afterwards?
- What residual risk are we accepting, and who signed it off?
The first 24 hours
- Activate the incident lead and open a timestamped decision log.
- Contain without destroying evidence; isolate affected systems and preserve logs.
- Establish the initial scope: data categories, systems, accounts and time window.
- Engage legal and communications; start the notification assessment.
- Reset credentials and revoke sessions for affected identities.
- Brief leadership with known facts, unknowns and the next decision point.
Preparing the response capability
If you need help building the inventory, testing the plan or reviewing the evidence trail, start with Managed Technology & Cybersecurity. The pricing page explains how assessment and response support are scoped. For the wider programme, read how SME incident response changes year by year.
Sources
- NIST SP 800-61 Rev. 3, Incident Response Recommendations and Considerations for Cybersecurity Risk Management, accessed 17 November 2026.
- NIST Cybersecurity Framework 2.0, accessed 17 November 2026.
- IBM, Cost of a Data Breach Report 2026, accessed 17 November 2026.
- The Daily Star, Bangladesh Bank issues 17-point cybersecurity directive, accessed 17 November 2026.
- DataGuidance, Bangladesh data protection overview, accessed 17 November 2026.
Responses (0)
No comments yet. Add the first one below.
Leave a response
Your email address will not be published. Required fields are marked *.