Most organisations survive a cyber incident and then quietly move on, relieved it is over. That is the most expensive mistake you can make. The incident already cost you money, downtime and trust; the only way to get a return on that cost is to learn from it properly. A disciplined post-incident review is how a breach becomes the moment your security program actually matured.
This guide sets out how to run a review that produces real change rather than a document nobody reads. It draws on the same approach we use when we are brought in after an incident as part of virtual CISO engagements across Australian government and enterprise.
Why most post-incident reviews fail
The typical review fails for predictable reasons. It is held too late, when half the detail has evaporated. It descends into blame, so people become defensive and stop volunteering information. It produces a long report with no owners and no deadlines, so nothing changes. Or it focuses entirely on the technical root cause and ignores the decision-making, communication and governance failures that usually made the incident worse.
A good review avoids all four traps. It is timely, blameless, action-oriented, and looks at the whole response, not just the malware.
Set the ground rules before you start
Open the session by stating the purpose plainly: we are here to understand what happened and how to make the organisation stronger, not to find someone to punish. This is not a soft nicety. When people fear consequences, they hide the very details that explain the incident, and you end up fixing the wrong thing.
Make the rules explicit:
- Blameless framing. Human error is a starting point for investigation, not a conclusion. Ask why a reasonable person made that choice given the information and pressure they had.
- Facts before opinions. Build the timeline from logs, tickets and messages first, then discuss interpretation.
- Everyone contributes. The most junior responder often saw the earliest signal. Make space for them.
- A neutral facilitator. For significant incidents, have someone who did not work the response chair the session so questioning stays objective.
Reconstruct the timeline
The backbone of any review is a precise, evidence-based timeline. Work from initial compromise through to full recovery, recording each event with a timestamp and a source. Capture not just technical events but human ones: when someone first noticed something odd, when the incident was formally declared, when leadership was briefed, when customers or regulators were notified.
Pay particular attention to the gaps. The time between compromise and detection (dwell time) and between detection and containment are the two numbers that most determine how bad an incident becomes. If your detection took days, that is a Detect capability problem worth far more than arguing about the initial phishing email.
Ask the questions that matter
With the timeline agreed, work through a structured set of questions. The aim is to separate what genuinely worked from what you got lucky on, and to expose the conditions that allowed the incident to escalate.
- How did the attacker get in, and why did that path exist? Trace it back to a missing control, an unpatched system, or an over-privileged account, not just the user who clicked.
- Why did detection take as long as it did? Was there no logging, no alert, an alert nobody saw, or an alert dismissed as noise?
- Did our incident response plan match reality? Plans written in calm rarely survive contact. Note every point where the team improvised because the plan was wrong, missing, or could not be found.
- Were the right people reachable and empowered? Many incidents stall because the person who could authorise shutting down a system was unreachable or unwilling to decide.
- How well did communication work? Internally between responders, upward to executives, and outward to customers, partners and regulators.
- What did we do that worked, and should keep? Reinforcing good practice matters as much as fixing failures.
Find systemic causes, not single points of blame
Resist the urge to stop at the first cause you find. A user clicked a link because the email was convincing, because the filter let it through, because multi-factor authentication was not enforced on that service, because nobody owned the asset inventory that would have flagged it as exposed. Each layer is a control that could have stopped the chain. The discipline is to keep asking why until you reach causes you can actually design out.
This is where mapping findings to a framework helps. Aligning each weakness to a control set such as the Essential Eight or an ISO 27001 control turns a messy list of complaints into a structured remediation backlog. APRA-regulated entities should also test their findings against their obligations under CPS 234, which requires incident response capability to be tested and lessons to be actioned.
A useful technique is to classify each contributing factor by the type of failure it represents: a technical control that was missing, a control that existed but was misconfigured, a process that was not followed, a process that did not exist, or a decision made under pressure with incomplete information. The mix tells you where to invest. If most factors are process gaps, more tools will not help; you need clearer runbooks and rehearsed roles. If they are technical, the spend belongs in controls and monitoring.
Turn findings into accountable actions
A review that ends with insights but no commitments has failed. Every material finding becomes an action item with three things: a named owner, a due date, and a way to verify it is done. Prioritise ruthlessly. You will surface twenty improvements; pick the handful that would have meaningfully changed the outcome and fund those first.
Distinguish between quick fixes and structural changes. Enabling MFA on a forgotten service is a quick fix. Building an asset inventory, standing up centralised logging, or rewriting an incident response plan are structural changes that need budget and time. Both belong in the plan, but be honest about which is which so the structural work does not get starved.
Then close the loop. Schedule a follow-up 30 to 60 days later to confirm the actions actually landed. Untracked action items are how organisations suffer the same breach twice.
Be wary of two failure modes in this step. The first is the action item that quietly becomes a project with no end date; if a finding needs a year of work, break it into a phase you can deliver in the next quarter and a backlog item for the rest, so progress is visible. The second is the action that gets marked complete because a document was written, when the actual behaviour has not changed. Verification should test the outcome, not the artefact: confirm MFA is enforced by checking the configuration, not by reading a policy that says it should be.
Capture the wider lessons, not just the technical ones
The richest findings from a serious incident are rarely purely technical. Reviews consistently surface decision and communication failures that mattered more than the malware itself. Common examples include leadership being briefed too late to make timely calls, no clear authority to take a critical system offline, legal and communications being engaged only after a misstep, and staff not knowing how to report what they were seeing.
Treat these as first-class findings. A breach that was technically modest can become a public crisis purely because the communication response was slow or inconsistent. Equally, capture the decisions that went well. If someone made a fast, correct call to isolate a network segment, document why they were able to, because that authority and confidence is something you want to make repeatable rather than dependent on one individual happening to be available.
Feed the lessons back into resilience
The final step lifts the review from tactical clean-up to genuine resilience. Update your incident response plan with what you learned. Adjust your detection rules so the same attack is caught faster next time. Run a tabletop exercise based on the real scenario so the next response is sharper. Brief the board honestly on what happened and what it cost, because that is often what unlocks the funding to fix the structural gaps.
Done well, the post-incident review is the single highest-leverage activity in security. It converts the worst day of your year into a permanent uplift in capability. If you would value an independent facilitator to run the review or help you build the remediation plan, CISO Advisory works with Australian government and private sector organisations on exactly this. Call 07 2112 8502 or get in touch and we can usually be on site the same day or next business day.
Frequently asked questions
What is a post-incident review?
A structured session held after a cyber incident is contained, where the response team reconstructs what happened, identifies what worked and what failed, and agrees specific improvements. It is sometimes called a post-mortem, lessons-learned review, or after-action review, and its goal is organisational learning, not blame.
How soon after an incident should you run the review?
Hold it within one to two weeks of containment, while memories are fresh but the firefighting has stopped. Waiting too long means detail is lost; running it mid-crisis means people are too stressed to reflect. Collect timeline data during the incident so the review itself is faster.
What is a blameless post-incident review?
A review that treats human error as a symptom of system weaknesses rather than the root cause. People speak openly only when they know they will not be punished, so blameless reviews surface far more useful detail. You still hold the organisation accountable for fixing the underlying conditions.
Who should attend the post-incident review?
Everyone involved in detecting, responding to, and deciding during the incident: technical responders, the incident manager, communications, legal, and an executive sponsor. For serious incidents, include a neutral facilitator who did not work the response so questioning stays objective.
Do we need a post-incident review for a small incident?
Yes, but scale it. A near-miss or minor incident may need only a 30-minute discussion and a few action items. Small incidents are often early warnings of the control gaps that later enable a major breach, so reviewing them cheaply is high value.
Need this handled for your organisation?
Confidential and no obligation. We respond the same business day — on-site same day / next business day, or remote, Australia-wide. Prefer to talk now? Call us 24/7 on 07 2112 8502.