The opinions expressed here are those of the authors. They do not necessarily reflect the views or positions of UK Finance or its members. 

This challenge lies at the heart of Root Cause Analysis (RCA) for Financial Services professionals.

When something goes wrong, there is understandable pressure to establish what happened, identify the cause and take corrective action. A process wasn't followed, a control failed, a colleague made an incorrect decision, information wasn't escalated or training was inadequate. All are plausible findings, but should we really be calling them root causes?

The challenge is that investigations create a strong incentive for resolution. We want to move from problem to explanation and from explanation to action(s). Once a credible cause has been found, we often stop looking. Yet this may be precisely the point at which the most useful part of the investigation should begin.

The first answer may only explain ‘what’ happened

Consider a customer who receives a detrimental outcome because a colleague failed to follow a procedure. It would be easy to conclude that the root cause was non-compliance and recommend refresher training.

But why wasn't the procedure followed? Perhaps it was difficult to interpret. Perhaps the system encouraged a workaround. Perhaps workload had increased significantly, or experienced colleagues routinely approached the task differently from the documented process. These are where the ‘how’ and the ‘why’ are hidden.

Each further level of explanation changes our understanding of the event and, crucially, what we might decide to do about it.

Effective RCA therefore distinguishes between different levels of causation. There may be an immediate or primary cause that explains the final event, contributory factors that made it more likely, and deeper organisational (root) causes that explain why those conditions existed in the first place.

What we call the cause determines what we fix

If an investigation concludes that an employee made an incorrect decision, the response is likely to focus on that employee: coaching, retraining, supervision or a reminder of the procedure.

If further analysis reveals that employees across the entire business are making similar decisions because guidance is ambiguous, systems are difficult to navigate or competing objectives are influencing behaviour, the appropriate response looks very different.

A useful test is to ask: if we removed or changed this factor, would we significantly reduce the long-term likelihood of the problem happening again? If the answer is no, there may be more work to do.

Cause levels really matter

Actions against primary causes might put the fire out in the short term. Addressing contributory causes may strengthen systems and processes in the medium term. Longer-term change only comes from understanding and acting on the deeper causes that allowed the problem to develop.

None of this means every problem requires an exhaustive investigation. Proportionality matters. A minor processing error and a significant regulatory or customer-impacting failure clearly warrant different levels of scrutiny.

But where a problem matters enough to investigate, stopping at the first credible explanation can create an illusion of understanding. The objective of RCA is not to find something that can be quickly labelled as the root cause. It is to understand the problem well enough to make the best decision about what to do next.