Appendix: Benchmark criteria

The benchmark criteria are divided into two parts: the content quality of the report and the quality of the iXBRL filing. Both components were given equal weight in the final score.

Part 1: Content quality criteria

1. Strategy and value creation

We looked at whether the report clearly explains the company’s strategic direction, business model and approach to long term value creation. This includes the choices and priorities behind the strategy, the external developments influencing those choices and the way financial and broader stakeholder value are connected.

2. Materiality and external environment

We assessed how well companies identify and explain the issues that matter most to their business and stakeholders. This includes material sustainability impacts, risks and opportunities, as well as broader market, regulatory and societal developments that could affect the company.

3. Financial integration

We looked at whether material sustainability issues are connected to their financial consequences. This includes effects on areas such as revenue, costs, investments, assets, financing and future performance, as well as how these considerations influence capital allocation and financial planning.

4. Targets and performance

We assessed whether strategic and sustainability ambitions are translated into clear targets and meaningful KPIs. We also looked at whether companies report progress over time, explain changes in performance and provide sufficient context to understand whether they are on track.

5. Transparency and credibility

We looked at how openly companies report both positive and negative developments. This includes explaining underperformance, uncertainties, limitations and changes in assumptions or methodology, rather than presenting only the positive side of the story.

6. Governance, data quality and assurance

We assessed whether responsibilities for reporting and performance are clearly defined and whether the information is supported by appropriate systems, processes and controls. We also considered the transparency of data quality, estimates and limitations, and the scope and level of external assurance.

7. Connectivity and integrated thinking

We looked at whether the different parts of the report form one coherent story. Material issues should connect with strategy, risks, targets, KPIs, investments, governance and financial performance. We also looked for evidence that these connections exist in management decision making, not only in the published report.

Part 2: iXBRL quality criteria

1. Currency consistency

It is important that the correct currency has been used for all the different facts within the report. Wrong currencies can lead to different valuations of certain facts which impacts machine analysis greatly.

2. Tag consistency

This means that same types of values have been tagged with the same concept throughout the years. This consistency is important for data comparability. When different concepts are used for the same types of values throughout reporting periods, it becomes very difficult to analyse trends over time.

3. Value consistency

Reports contain the values of multiple years, which means that throughout time there is an overlap of the same values between different reports. (e.g. the 2025 report contains values from both 2025 and 2024). These overlapping values should be exactly the same within the different reports, because if they differ, it becomes unclear which value is correct and which one is false.

5. Precision

Within XBRL validation it is possible to set a rounding margin. This is important to prevent false positive calculation errors. It is possible however to set this rounding margin so high, that false negatives can occur, which influences the quality of the tagged values. A high quality XBRL file contains the proper rounding margins per tagged value.

6. Sign consistency

Signs can be used to mark a tagged value as a negative value. In XBRL, however, most concepts already carry a debit/credit property, which defines if the value should be interpreted as positive or negative. Adding a negative sign in most cases means that an already negative value is marked as negative again, resulting in a double negative, which affects machine analysis.

6. Validation errors

XBRL Taxonomies contain validation rules. Any reporting package needs to be compliant against these validation rules. When a report triggers validation errors, there are likely either technical or content mistakes within the report, which affects machine analysis.

7. HTML quality

A report can contain a lot of different types of html styling to ensure maximum human readability. Machines however look at the actual html tags being used within the report. Most html tags contain semantic value that the machine can use as context to better understand the context. The more of these semantic tags are used, the higher the machine readable quality of the file.

8. Extension and anchoring quality

Extensions are necessary to ensure the proper data accuracy of values and anchoring ensures the possibility for data comparability. It is important that extensions are only used when absolutely necessary and the anchoring needs to be correct to prevent any misinterpretation of the values meaning.

9. Tagging quality

It is important that values are tagged to the right concept, so that the machine interprets a value exactly the same as a human would do. Tagging to a wrong concept greatly affects the difference in interpretation between a human reader and a machine. It also affects comparability between different reports.