What is root cause analysis?

Root cause analysis is a systematic method for identifying the underlying reasons for variances, errors, or undesirable events, for example in business processes. In controlling and business intelligence, it is used not only to identify symptoms but also to uncover the actual drivers behind KPI variances – providing the basis for targeted measures and informed decisions.

Root cause analysis – often referred to as RCA – is a structured method that goes beyond identifying symptoms. Its objective is to determine the actual underlying cause of a problem so that corrective measures have a lasting effect rather than addressing only the short-term symptoms.

In a business context, the need for action typically arises when KPIs deviate from plan or prior-year values. Simply stating, “Revenue has fallen by 8%,” is not sufficient for effective management. Root cause analysis systematically asks: Why has revenue declined? In which segment, region, or product did the decline occur? And what triggered this development?

These questions can usually only be answered by viewing, breaking down, and relating data from different perspectives. This is where root cause analysis connects with modern BI systems.

Various methodological approaches are available, depending on the context and complexity of the problem:

The 5 Whys method is a simple but effective tool. Starting with the observed problem, the question “Why?” is asked five times to progressively identify the underlying root cause. The method is particularly suitable for operational problems with a linear chain of causes.

Example:

1. Why has contribution margin declined? – Because raw material costs have increased.
2. Why have raw material costs increased? – Because a supplier failed.
3. Why did the supplier fail? – Because no alternative supplier had been contractually secured.
4. Why was no alternative supplier secured? – Because the supplier portfolio was not regularly evaluated.
5. Why was this evaluation missing? – Because no process had been defined for it.

The Ishikawa diagram visualizes cause-and-effect relationships. It categorizes potential causes into major groups – typically people, machines, materials, methods, environment, and management – enabling a structured team discussion. It is particularly suitable for complex problems involving multiple parallel causal chains.

Based on the Pareto principle (80/20 rule), Pareto analysis identifies the few causes that account for the largest share of a problem. In controlling, it helps focus resources on the areas with the greatest potential impact.

In a BI context, drill-down is a key tool for root cause analysis. Starting from an aggregated KPI, the analysis progressively moves to lower hierarchy levels – from the overall level through regions, product groups, and customer groups down to individual transactions. Drill-down analysis makes hidden patterns visible and leads the analyst directly to the relevant underlying data.

In controlling, root cause analysis is closely linked to variance analysis. As soon as plan-versus-actual or year-over-year comparisons reveal significant differences, root cause analysis begins. It does not merely answer how large a variance is, but where it comes from.

Modern BI systems support this process through:

• interactive dashboards with drill-down and drill-through functionality
• automatic variance highlighting, such as traffic-light systems, sparklines, and variance bars
• multidimensional data analysis, which examines causes across multiple dimensions simultaneously
• commenting functions, which link qualitative explanations directly to KPIs
• AI functions that can explain causes rather than merely display figures

A central principle in controlling is: A variance whose cause is unknown cannot be addressed effectively. Root cause analysis is therefore not an optional add-on, but an integral part of an effective management and control process.

Initial situation: A retail company determines in its monthly reporting that gross profit for the current quarter is 6.3% below plan. The variance is visible at the overall level, but its cause is unclear.

Approach with DeltaMaster:

The controller opens the plan-versus-actual report in DeltaMaster and uses the integrated drill-down functionality to progressively break down the variance:

1. Overall company level: Gross profit –6.3% versus plan
2. Business unit level: The “Electronics” business unit shows a significantly larger variance of –14.2% than the other units
3. Product group level: Within “Electronics,” the variance is concentrated in the “Large household appliances” product group
4. Customer/region level: Two major customers in the southern region have reduced their order volumes by approximately 40%
5. Pricing and terms level: An analysis of sales prices shows that unplanned discounts were granted in this segment
6. Semantic context: A stored conversation note indicates that one of the two major customers had been presented with a competitor’s offer and that the discounts were granted as a short-term retention measure.

Result: The root cause analysis does not stop at the figure but extends into the semantic context. Variances are not only quantitatively localized but also explained in business terms – making them relevant for management action: Sales management recognizes that the issue is not a demand problem but a pricing and terms process issue and can take targeted action.

DeltaMaster makes this complete causal chain – from the aggregated variance to the qualitative explanation – visible through automatic variance detection, sparklines, and the integrated commenting function in the report layout.

Root cause analysis is an essential tool in controlling and business intelligence. It goes beyond merely identifying variances and uncovers their underlying drivers. Modern BI systems such as DeltaMaster support this process through interactive drill-down functionality, automatic variance highlighting, and integrated AI functions. Knowing the causes of variances enables effective management action – and helps prevent the same problems from recurring.