To create this dataset we process structured and unstructured data sets, as well as traditional financial datasets, to ensure that all ESG-relevant data on around 17,000 public companies are gathered, measured for consistency and transparency, and scored in real-time across the E, S and G pillars. We track over 500 ESG factors and process around 200 features (based on statistical significance) to create our ESG scores. We have built a powerful proprietary algorithm to aggregate the data from raw values (such as carbon intensity metrics, employee satisfaction, diversity metrics etc.) to an overall score. The key differentiators of our scores are that they are: 1) Real-time 2) Explainable 3) Comprehensive across all 3 E, S, G pillars 1) Real-time Scores -------------------- Our platform uses the latest technologies which means that we are able to provide real-time ESG scores for our c.17,000 companies at scale. Some of the ESG factors that we use are real-time numbers such as people-related features like board diversity, management gender pay gap and employee satisfaction. As a result, as we get new data it will contribute to companies’ scores in real-time, which keeps investors up-to-date with the latest available numbers. 2) Explainable Scores -------------------- We have built a proprietary algorithm that aggregates ESG features to give an overall score. Our scores have 4 levels: overall score, ESG pillar level scores, theme level scores, and feature level scores so you can see exactly where these scores came from. 3) Comprehensive across all 3 E, S, G pillars ---------------------------------------------- There is a lot of focus on the Environmental pillar generally, but we still also process a huge volume of data for the Social and Governance pillars.