Explore our new Cybersecurity Terms Glossary, explained by our experts.
Cybersecurity Terms Glossary
Root cause analysis works backwards from an incident to the conditions that allowed it, distinguishing the proximate trigger from the systemic weaknesses that made it effective.

What Is Root Cause Analysis?
Root cause analysis is the structured examination of why an incident was possible, as distinct from how it was carried out. The investigation establishes the mechanism; root cause analysis establishes the conditions. An organisation that closes an incident with “an employee clicked a link” has described the trigger and learned nothing it can act on.
The test of a stated root cause is simple. If you removed it, would the incident have been prevented or materially limited? If removing it still leaves an identical outcome, you have not finished.
Take a typical intrusion: a finance employee opens an attachment, credentials are harvested, the attacker authenticates to a remote access portal and moves laterally to a file server. The proximate cause is the attachment. The systemic causes are the ones that turned a single click into a breach:
Fixing the proximate cause means more training. Fixing the systemic causes means the next click does not matter nearly as much. Both belong in the output, but only the second changes the risk.
Five Whys is quick and effective for narrow problems, provided the chain is followed past the first organisational answer. The common failure is stopping at “the patch was not applied” rather than asking why that system was outside the patching process and why nobody noticed for fourteen months.
Fishbone diagrams group contributing factors into categories such as people, process, technology, and supplier, which is useful when an incident has several independent contributors rather than one chain.
Causal factor charting builds the timeline of events and conditions and marks each point where a different decision or control would have changed the outcome. It fits security work well because it reuses the timeline the forensic team has already built.
Whichever method is used, the input must be evidence. A root cause analysis built on recollection a fortnight later will converge on whatever the most senior person in the room believes.
The British Library’s published account of its October 2023 ransomware attack identifies the most likely entry as the compromise of privileged account credentials used to reach a terminal services server, installed in 2020 to give trusted partners remote access, on which multi-factor authentication had not been enabled. The report then goes further, naming the conditions that amplified the damage: historic underinvestment, a complex legacy estate, and reliance on systems that could not be restored in their original form.
That structure is the model. One trigger, several enabling conditions, and a set of actions that address the conditions rather than only the trigger.
A review that assigns individual fault produces incomplete information in every subsequent incident, because people conceal their part. The discipline from safety engineering applies directly: ask what made the mistake likely rather than who made it. An administrator who kept a credential in a text file did so because the vault was unusable from the jump host at three in the morning. That is a design finding, not a disciplinary one.
The output should be a short list of actions, each with a named owner, a date, and a defined verification step. Three implemented actions beat twenty-five recorded ones. Feed them into the risk register, report progress to the board, and check six months later that the control is working rather than merely installed.
One practical habit: write the root cause statement so that it holds true for the class of incident, not only this one. “No phishing-resistant authentication on externally reachable services” is actionable across the estate. “The attacker used Jane’s password” is not.
Our analysts produce this analysis as a deliverable at the end of Investigation and Incident Response Services, because an incident that is closed without it tends to recur in a slightly different form.
The investigation establishes what happened: the entry point, the attacker’s actions, the affected systems, and the data involved. Root cause analysis asks why the environment allowed it, identifying the control gaps and organisational conditions behind the events. The investigation produces facts and a timeline; the root cause analysis produces the changes that reduce the chance of a repeat.
Almost always. Serious incidents result from several weaknesses aligning rather than one failure, which is why models that only follow a single chain of causation can mislead. A good analysis distinguishes the proximate trigger from the enabling conditions and lists contributing factors separately, so that the resulting actions address the whole set rather than the most visible element.
State the principle at the start, exclude line management from the technical session where possible, and phrase every question about conditions rather than individuals: what information was available at the time, what the system made easy, and what pressure the person was under. Share the findings without names. People who expect blame withhold detail, and the detail is what makes the analysis useful.
After recovery is under control but while evidence and memory are fresh, typically within two to four weeks of the incident. Waiting until the pressure has gone entirely means logs have rolled over and participants reconstruct events rather than recall them. Running it too early competes with the response itself and produces conclusions based on an incomplete picture.
You May Be Interested In These
We use a few cookies that are needed to make this website work, and they are always on. We would also like to set analytics cookies, to see which pages are useful; functional cookies, so that our live chat works; and advertising cookies, to measure our campaigns. We will not set any of those until you say yes, and you can change your mind at any time from the Cookie Settings link in the footer.
Choose the cookies you are happy for us to use. Strictly necessary cookies are always on; everything else is off until you turn it on.
These cookies keep the website working: they hold your security checks, your form submissions and the choice you make here. The site cannot run without them, so they are always on.
These let our live chat open and remember your conversation, and let the newsletter sign-up work. Switch them off and the chat window will not appear.
These tell us which pages are read and how people find us, so that we can improve the site. The reports are about groups of visitors, never about you personally.
These let us measure our advertising and show you our adverts on other websites, such as LinkedIn, Google and Meta. Switch them off and you will still see adverts, but they will be less relevant.