How to learn from mistakes

Mistakes are part of how work gets done. There is common perception that failure naturally leads to learning, but in reality that’s not guaranteed.

Learning is the most valuable outcome that can come from failure, however we don't learn from experience itself but from reflecting on it. Learning does not just happen because something has gone wrong - it is entirely possible to go through failure and come away with no meaningful insight, to repeat the same mistake again, or to draw the wrong conclusion altogether. That tends to happen when the analysis is rushed, shaped by personal narratives, or influenced by what feels most comfortable rather than what is actually true.

There is a natural tendency to filter what happened in a way that protects ego or reinforces existing views, to overlook factors that are less visible or harder to accept, or to place too much weight on the most obvious or memorable part of the situation. It is just as easy to move in the opposite direction and overanalyse every detail until the learning becomes diluted and unclear. The value sits somewhere in between, and that is where a more structured approach becomes useful.

Root cause analysis offers a way of working through mistakes in a balanced and disciplined way, helping us move beyond surface-level explanations and towards learning that can actually improve how work is done.

1. Define the issue clearly

Start with what actually happened, including where and when it occurred, who was involved, and what the impact was. At this stage, the focus is on building a complete and accurate picture using facts only, without interpretation. If key details are missing or unclear, the rest of the analysis will be built on unstable ground.

2. Separate symptoms from causes

The most visible issue is often just the surface of a deeper problem. Missed deadlines, complaints, or breakdowns in delivery are usually symptoms rather than causes. If the analysis stops at what is most obvious, the response will only address what is immediately visible and the underlying learning will be missed.

3. Follow the cause chain

Working through the sequence of events that led to the outcome helps uncover what actually drove the situation. This can be done using simple methods such as the 5 Whys (like my 6 year-old often does), where you continue asking why something happened until you reach a point that is actionable, or by mapping out contributing factors and how they connect. The important point is to stop when you reach something that can be changed, as focusing on factors outside your control does not lead to useful learning.

4. Look at the system, not the individual

It is often tempting to attribute failure to an individual, but that usually brings the analysis to a premature conclusion. A more useful approach is to examine the system around the individual, including what in the process allowed the issue to occur, what conditions made it more likely, and what checks were missing or ineffective. Most meaningful learning sits within these systemic factors.

5. Use evidence, not assumptions

Opinions and recollections can be unreliable, particularly after something has gone wrong. A stronger analysis is grounded in evidence, such as data, specific examples, direct observations, and documented information. Where something cannot be supported by evidence, it should be treated with caution rather than accepted at face value.

6. Identify contributing factors, not just one cause

In most situations, there is not a single root cause but a combination of factors that together led to the outcome. These might include process gaps, communication issues, capability challenges, and contextual pressures. Capturing the full range of contributing factors allows for a more accurate understanding before deciding which ones matter most.

7. Listen and make a judgement call

Different perspectives are valuable because they help reduce individual bias and bring forward factors that might otherwise be missed. At the same time, not all contributing factors carry equal weight, and it is necessary to make a judgement on which ones are most significant. Without that judgement, the analysis can become overloaded with too many root causes, which dilutes the learning and makes it harder to act on.

8. Define corrective actions that address the cause

Actions need to link directly to the causes that have been identified. General statements are unlikely to change behaviour or outcomes, whereas specific actions that introduce clarity, structure, or accountability are more likely to have an impact. If an action does not reduce or remove the underlying cause, it will not contribute to meaningful improvement.

9. Make changes that prevent recurrence

The purpose of the analysis is not simply to resolve the immediate issue but to reduce the likelihood of it happening again. This often involves adjusting processes, introducing checkpoints, clarifying ownership, or strengthening capability in a targeted way, with a focus on improving how work is done going forward.

10. Validate and keep it proportionate

Once changes have been implemented, it is important to review whether they have had the intended effect and whether any new issues have emerged. At the same time, the level of analysis should be proportionate to the situation, as not every mistake requires a deep or time-intensive review, but every mistake should lead to some level of learning.

Used consistently, this approach becomes less of a formal process and more of a way of thinking about how work is reviewed and improved. It allows teams to move beyond surface-level explanations, make clear decisions about what matters, and take actions that strengthen how they operate.

The real value is not in the mistake itself but in what is taken from it and what changes as a result. Without that, failure does not carry any benefit and simply becomes something that is repeated rather than something that improves future outcomes.