.svg)
After any data collection campaign, the question of how to leverage CSR reporting data arises. In short, how can you make your data speak to better contribute to managing the company's overall performance ?
Why analyze CSR reporting data?
As soon as they finish collecting their CSR reporting data, companies begin their analysis. An in-depth analysis offers numerous benefits.
Analyzing non-financial data allows companies to better manage their material issues. Companies evaluate and compare results across sites, subsidiaries, geographic areas, divisions, business lines, and more.
What are the limitations of "traditional" analysis?
What kind of analysis are we talking about here? The analysis of raw data and the qualitative information that illustrates it. This allows for trend analysis across various dimensions (site size, country, business line, region, continent, etc.).
To evaluate and understand these trends at a group level, one often starts by comparing indicator curves corresponding to material issues. For example, you might simultaneously study trends in water consumption, energy consumption, and the quantity of products manufactured. The idea is to observe whether or not they move in the same direction or, more generally, if they are correlated with one another.
That said, working with raw data does not allow for nuanced trend analysis.
Consequently, the second step is often to calculate "intensities." These are ratios such as "energy consumption per product manufactured," "water consumption per m2 of floor space," "number of training days per employee," etc. These intensities provide valuable information. They help put raw data like "energy consumption," "water consumption," or "number of training days" into perspective.
However, using these intensities is still difficult and sometimes even misleading:
- An intensity represents only one facet of a subject and does not account for the true complexity of the company. For example, the energy consumption of an industrial site certainly depends on the number of products manufactured, but it may also depend on the technology used, the size of the buildings, the number of staff on-site, the weather, etc. Therefore, one can better appreciate the site's actual energy performance. It is clear that intensity per product manufactured alone is not enough;
- At a group level, consolidated intensity metrics often lose their meaning. The reason: they can mask a wide heterogeneity of underlying situations.
- Analyzing the evolution of a consolidated group-level intensity metric over time is far more complex than it appears. For instance, calculating the "year-over-year variation in global group intensity" might suggest an improvement when, in reality, none has occurred. Take the intensity metric "energy consumption per unit produced" at the group level: it might decrease simply because the most energy-intensive activity has significantly declined (due, for example, to a drop in sales). In this case, concluding that the group’s energy performance has improved would be incorrect.
How can we go further?
It is clear that understanding a company's true performance requires a more granular analysis of CSR reporting data. This requires combining operational expertise with statistical modeling. Such an approach begins by identifying the factors (i.e., other indicators) that influence the raw data we are interested in, such as energy or water consumption, to use the examples mentioned above.
Next, analyzing the data and its correlations with these factors using statistical models allows for a better understanding of the company's performance. Which sites or activities are the most efficient? How is their performance evolving over time?
This type of analysis builds bridges between financial and non-financial data. It also opens up new possibilities for future integrated reporting. It encourages the sharing of useful information and best practices, as well as the (re)definition of priorities. It fosters innovation, which in turn feeds the company's "best-in-class" repository of practices.
Implementing statistical models also makes it easy to identify sites with high potential for improvement—a task that can prove complex when dealing with hundreds of sites, each with different characteristics.
This type of analysis will also lead to the definition of realistic improvement targets and the action plans required to achieve them.
It is worth noting that, in this spirit of continuous improvement, CSR reporting data analysis must also cover "coverage rates." Why? To ask the right questions about the evolution of your CSR reporting protocol, the indicator framework (for example, to remove unnecessary questions), and the validation process.
Conclusion
Making CSR reporting data speak for itself is an essential challenge for steering and improving a company's overall performance. This complex and exciting exercise requires combining operational insights with mathematical and statistical analysis.
This will only be possible for those who have reliable CSR reporting data and the necessary time. This implies being equipped with non-financial reporting software, which facilitates data collection, ensures data reliability, and enables multidimensional data analysis.
Photo credit: UX Indonesia