How AI integrates into our data design process
In 2024, we turned our attention to ESG (Environment, Social, Governance). The three pillars encompass a range of global issues that are measured in a quantitative and regulated way.
ESG aims to provide stakeholders — such as investors, consumers, and regulators — with transparent insights into a company’s sustainability practices. In practice though, it is not without controversy. ESG faces challenges such as inaccurate data collection, the risk of misleading claims, and pressure that may lead companies to overstate their sustainability efforts.
In response, new EU legislation will soon require nearly 50,000 companies to disclose their environmental and social impact in an effort to “end greenwashing, strengthen the EU’s social market economy, and lay the groundwork for new global sustainability reporting standards.”
— EU Parliament press release, November 10th 2022
Whilst this legislation marks a positive step toward increased accountability and transparency, we believe such critical data needs better data design practices to reach it’s full potential.
At CLEVER°FRANKE, we regularly engage with new domains, data types, and specialized areas. In exploring the ESG domain, we sought to understand how sustainability is measured. We were struck by a complex web of various actors, regulations, and proprietary information, little to none of which caters to the needs of the general public.
We believe that data should help people understand the world around them. In our view, the ESG domain ought to be:
As data designers, our expertise lies in translating complex data into meaningful, insightful experiences. We set out to analyze how ESG is currently presented, and investigate how design can make ESG more accessible, transparent and comparable.
We analyzed three components of ESG: Metrics represent the most detailed view of this data, and are consolidated into extensive reports by the companies themselves. ESG reports, along with other forms of disclosure, are evaluated by external bodies to assign an ESG score. We explored the design choices across each to better understand the existing landscape.
Sources suggest there may be as many as 600+ different ESG scoring systems. We focused on a number of scoring systems that were regularly cited and referenced, and crucially, that had information available that outlined their methodology. We compared their design choices in terms of:
We analyzed the content and hierarchy to understand which information scoring companies prioritize. They were generally consistent, prioritizing the following data:
Despite presenting the same content, each scoring system chose different visual-communication techniques.
Each scoring company uses different terminology to communicate the score itself, ranging from numbers to letters to grades. Some scoring systems evaluate performance, others evaluate risk.
These inconsistencies in terminology make it challenging to compare their findings and as a result, any limited understanding of one system cannot be applied to another.
We found that each ESG scoring system employed a unique color scale, including a wide range of hues and combining both diverging, sequential, classed and unclassed palettes.
Color is an important part of interpreting data. In data visualization, data is typically classified as sequential, divergent, or categorical, each of which has a corresponding color palette.
Unclassed data represents values along a continuous gradient or spectrum.
Classed data is organized into distinct steps, categories or buckets.
Sequential data is best visualized using a color palette that varies uniformly in lightness, saturation or hue, ordered to effectively represent progression or magnitude.
Divergent data is best visualized with a split color palette centered on a neutral midpoint, highlighting variations above and below a median value.
Differences in color through hue, palette, and classification not only makes the results harder to interpret and understand but also hinders meaningful comparisons of their findings.
ESG scoring companies present their grading systems and score breakdowns using various visual forms, from bullet charts to sunbursts. Each format has its merits and limitations, revealing different focuses on the data.
Our work at CLEVER°FRANKE often involves determining which type of chart is most appropriate for the data, and creating custom visualizations to support it.
Below are abstracted versions of some of the visual forms we encountered, with examples of the data types they’re most appropriate for, and some of their limitations.
The choice of chart can vary significantly based on the audience, intent, and data involved. When assessing the ESG impact of a single company, however, the variability in visualization techniques, terminology, and color schemes can lead to differing interpretations of the scores. This inconsistency highlights the importance of careful chart selection and presentation, as it can significantly influence understanding and perception of a company’s ESG performance.
When comparing all these elements of ESG scores, the most significant takeaway was the total lack of consistency. Whilst we can assume that market competition and slight variation in audience drives them to differentiate themselves from each other, the true failure is the inability for non-experts to interpret them, or make comparisons between them.
An example of consistent grading can be found within the energy sector. Whilst there are some differences in how scores are visualized between countries, strong similarities enable a degree of comparison.
This example of consistency in visual communication has persisted around the world for almost thirty years, giving consumers time to become familiar with the system and build understanding of what it represents.
For more information on our work with ESG, visit cleverfranke.com/esg.
How AI integrates into our data design process
Particle wind font
A font based on wind animation
Geographical scrollytelling