Data & Methodology
GaiaLens optimises explainability to enable users to drill into and fully understand the detail behind the numbers. Our proprietary algorithm enables the system to aggregate a massive volume of data in order to calculate overall ESG scores.
We are flexible when it comes to data delivery and can deliver via our API, data dumps or sharing via a bucket.
Scientifically Sourced ESG Scores
We calculate ESG scores for each of our c.20,000 companies in real-time using our proprietary ESG scoring algorithm. Our scoring system has four levels of hierarchy starting with transforming raw ESG data into Factor Level Scores, then calculating Theme Level Scores, Pillar Scores and finally Overall Scores. We have historical scores going back 20 years.
Proprietary ESG Factors
We engineer proprietary datapoints which are inputted into our scoring algorithm. These include Beneish M Score, Board Gender Pay Gap, Ethnic Diversity on the Board, Percentage of Independent Directors on Board, Percentage of Women on the Board, Employee Satisfaction Rating, Employee Turnover, Corporate Misconduct Penalties.
Real-time ESG News
We have millions of articles that we generate signals from using the latest Large Language Model (LLM) technologies. We heavily process each article and extract as much value as possible including calculating a company relevance score (which measures how relevant the article is to the company); we also calculate sentiment scores for over 50 different ESG topics including SASB themes and UN SDGs. We have news going back over 10 years.
