Making a Difference Through Data: How Data Analysis Enhances Performance Audits and Evaluations

Author: Mario Luketić, Secretary General of the INTOSAI Working Group on Evaluation of Public Policies and Programs (WGEPPP), Swiss Federal Audit Office

Data has become a central part for the management of complex public programmes. It helps to highlight trends, better understand impacts and identify the need for policy adjustments. For Supreme Audit Institutions (SAIs), the question is therefore no longer whether data analysis should be used in performance audits and evaluations. Rather, the key issue is how data analysis can be planned, prepared and integrated into audit work in such a way that it is driven by the audit questions and contributes to better insights, more convincing recommendations and, ultimately, greater impact.

Within this context, the forum of the INTOSAI Working Group on the Evaluation of Public Policies and Programmes took place in Bern in October 2025 under the title “Making a Difference Through Data”. The aim of the forum was to discuss how data analysis can be embedded in the practice of SAIs and how it contributes to making performance audits and evaluations more effective.

Creating a Culture of Data Analysis in a SAI

A survey conducted among WGEPPP members and observers showed that data analysis is already part of audit work in many SAIs. It is used to analyse large amounts of data, highlight trends and identify differences between regions, institutions or population groups. Accessible applications such as data visualisations, digital surveys and descriptive analyses are particularly common. However, although data analytics is being used more widely, it is not yet systematically embedded in most SAIs. While some institutions have made it a strategic priority, many still apply it on a case-by-case basis. Only a minority of auditors are actively involved in data analysis, and specialised data units remain rare.

This underlines the need to build and sustain a data-driven culture within SAIs. Such a transformation is not merely a technical exercise. It requires lasting changes in behaviour and must become part of everyday audit work. The Swedish National Audit Office illustrated this point through an approach built around four mutually reinforcing elements: 1. skills, 2. processes and organisational design, 3. methods and use cases, and 4. infrastructure. Developed together, these elements provided the basis for long-term organisational change.

In the early stages, progress depended on clear leadership and decisive choices. Over time, the focus shifted towards experimentation, iterative learning and the creation of spaces where new approaches could be tested, such as an innovation lab. A clear purpose was also essential. Framing data analytics to translate technology into societal value helped reduce resistance and encouraged staff to take part in the transformation.

In practical terms, this requires SAIs to define clear objectives for the use of data analytics and to translate them into staff development and audit planning. This includes identifying the skills required, whether in a dedicated data unit or within existing audit teams. It also means assessing current capabilities and identifying auditors who can benefit from targeted training. Professional development plans can then help staff build relevant skills in a structured way. Finally, audit assignments should be organised so that staff can apply and strengthen these skills in practice. The Swiss Federal Audit Office has also developed a tool to assess the level of data analytics culture within an SAI, based on a comparative survey of auditors.

This leads to a key lesson from the forum. If SAIs wish to make sustainable use of data analysis, they must develop it as a core component of their working model. It is not enough to support individual projects with data analysis. Auditors themselves also need to build data analytics skills, so that they can identify relevant analytical opportunities, work effectively with data specialists and integrate data-based evidence into their audit judgement.

Integrating Data Analysis Throughout

Data analysis should be considered at two distinct stages of the audit or evaluation cycle. The first stage is audit programming, before the annual programme is finalised. At this point, SAIs should assess the potential for data analysis in relation to the topic under consideration and the data that may be accessible from the audited entity or from other sources. This early assessment helps to determine whether data analysis can add value, what type of skills are required and whether data analysts should be involved in the audit team from the outset.

The second stage is the preparation of the audit or evaluation itself. At this point, the audit questions involving data analysis need to be specified more precisely. The audit team should then assess whether the necessary data exist, whether they can be accessed and whether their quality is sufficient for the intended analysis. Where such data are not available, the team should consider whether they can be collected in a proportionate and reliable way within the audit. This helps ensure that data analysis is not added as an afterthought but is properly integrated into the audit design.

Added Value Stems from Clear Audit Questions

A key finding of the forum was that data analysis does not derive its value from the technical complexity of the method or the volume of available data. The added value is realised above all when it is closely linked to clear audit questions. Data analyses can directly answer important evaluation questions or back up qualitative findings. They can also help to present the context of an audit finding in a more comprehensible way. It follows that data analysis almost never stands alone. It delivers the greatest added value in combination with other methods, such as interviews, case studies or document analyses. For SAIs, this is a particularly important insight. Data does not replace the professional judgement of auditors. It broadens the basis for that judgement and can help to better justify conclusions.

The forum also demonstrated that data analysis is particularly valuable where traditional audit approaches reach their limits. It can reveal patterns that are difficult to identify in individual case audits. It can capture target groups more accurately and highlight discrepancies between policy objectives and actual outcomes. Data analysis becomes a tool that goes beyond description. It can help to identify problems more precisely and provide better justification for recommendations. The added value therefore lies not in more data, but in a strict analytical approach to the audit work.

Data Analytics for More Impactful Audit Reports

Another key topic was how data analysis contributes to the impact of audits. The answer is not always immediate. Data-driven audits rarely lead directly to visible improvements. Their impact is often indirect. They highlight problems and improve the evidence base for recommendations. This is precisely where their practical value lies. When SAIs can demonstrate where programmes are failing to reach their target groups or where resources are not being used effectively, they create a stronger foundation for improvements. The same applies when data gaps prevent appropriate management. Data analysis thus not only helps to make audit reports more precise. It can also help to ensure that recommendations become more concrete, verifiable and more useful for public administration and decision-makers.

Two additional factors make data analysis particularly relevant for SAIs. First, audited entities may not always have the resources, expertise or incentives to analyse their own data in a systematic way. By demonstrating the added value of data analysis, SAIs can encourage them to make better use of the information they already hold and to strengthen evidence-based policy management. Second, important data is often spread across different agencies. The horizontal perspective of SAIs enables them to connect, compare and cross-reference data from different sources, which can reveal patterns and insights that would otherwise remain hidden.

The WGEPPP Forum 2025 demonstrated that data analysis does not automatically improve performance audits and evaluations. It only realises the benefits when it is planned at an early stage, applied methodologically appropriately and combined with other sources of evidence. Equally important is that it is supported organisationally and remains linked to professional audit judgement. The central message of the forum is therefore that data analysis is a tool that can help SAIs to ask better questions, understand complex public programmes more accurately and formulate more effective recommendations. Innovation in performance audits therefore does not simply mean using new tools. It means bringing together data, methods, people and institutional learning in such a way that audits become more relevant, evidence-based and effective.