80–20: Empowering businesses to make better climate risk-related decisions
Originally published in March 2022 through StartU
As extreme weather events have become more frequent and destructive, it is hard not to be apprehensive of the escalating impact of climate change in our local communities and broader world. In the 2022 Global Risk Report, the World Economic Forum reported that extreme weather and climate action failure are among the top five short-term and long-term risks to the global economy. Understanding the impact of climate change on businesses and humanity, 80–20 has developed a solution to increase climate projection accuracy, improve insurance market efficiencies, and help companies diversify their portfolios in the context of climate risk.
About 80–20
80–20’s SaaS platform empowers businesses to make better climate risk-related decisions by creating more accurate climate forecasts and translating these forecasts into actionable insights for the insurance industry. The platform applies machine learning algorithms to identify cross-climate risk correlations, improving projection accuracies by up to 80% compared to current publicly available models.
I had the pleasure of interviewing Nicole Engels, co-founder and CEO of 80–20, to learn more about the company’s platform and mission. Nicole began by highlighting the report from Aon that global economic losses from natural disasters were $343 billion in 2021, and the global catastrophe protection gap reached 62% in 2021 with $213 billion in uninsured losses related to natural catastrophe and severe weather events. To protect against increasing claims, insurance companies are increasing premiums and pricing more people out of insurance, further widening the gap.
Nicole views knowledge and data as key to tackling climate change. She says, “To be able to mitigate the effects of climate change, you have to first know what is happening. The focus of our business is taking the first step to increase information gathering and sharing. With a data-driven approach, we can get people and businesses to agree on what is happening and then how we can address it.”
80–20’s Climate Forecasts
80–20 is fully transparent about its data sources. They leverage publicly available historical data and forward-looking models to build climate forecasts beyond a 2-week horizon. By applying machine learning algorithms to identify correlations of climate factors, they can understand how the climate behaves holistically. As more correlations are identified, projection accuracy increases, which is currently up to 80% more accurate than publicly available models. Their forecasts highlight areas impacted by climate disasters but also areas of opportunity that may be less threatened due to the movement of where these extreme weather events occur. This allows insurance companies to diversify their portfolios by geography and by category of climate risk.
As an example of the type of correlation embedded in 80–20’s models, Nicole described data that shows that a wildfire in one area can lead to a higher chance of flooding the next year, since the soil is no longer able to absorb high levels of precipitation. This information can be used to understand areas with increased need for flood insurance in addition to wildfire risks.
As another example, the images below show FEMA’s policy counts, which are standard reference points for the insurance industry in making decisions about where to sell flood coverage. The middle image shows FEMA’s actual claims count for 2020. 80–20’s model predicted FEMA’s actual claim count with 83% accuracy.
Market Intelligence Tool
80–20’s climate forecasts are used to inform their market intelligence tool, which aims to make the climate insights relevant for the property and casualty insurance industry. The tool, which users can use at their own discretion, aggregates industry data and gives the user seamless access to data on rate fillings, rate changes, pricing opportunities, builders’ data, supply and disaster chain relationships, amongst other data. 80–20’s climate forecasts inform the market intelligence tool to project future changes.
The Founders
The 80–20 journey started only 6 months ago at Columbia University’s Hacking for Humanity hackathon, where the two co-founders first met. Nicole, with a background in marketing at Anheuser Busch, was looking for the right opportunity to build a business in the climate space. Perry Beaumont, Ph.D., a data scientist currently with AWS Cloud Computing and a Lecturer at Columbia University, reached out to Nicole about his idea, and the two quickly found that they worked well together and shared a passion for mitigating the risks of climate change.
Perry, co-founder and CTO, had already begun developing the technology behind 80–20, having uncovered the need to improve the accuracy of climate projections, particularly for insurance, through his decades-long career in insurance and finance. Over the last 30 years, he has led several high-profile insurance and banking initiatives, including serving as the first Head of Data Science and Analytics at Distinguished Programs (an insurance firm), successfully launching a fintech firm that was acquired, and managing a multi-national division for AXA Insurance with their asset management group. Perry brought on Nicole to help develop the business model and ultimately bring the product to market.
Roadmap
After finishing Columbia’s Hacking for Humanity competition as a top 10 finalist in September, the founders completed Columbia Business School’s Launch Your Startup course and were recently accepted into Columbia’s Entrepreneurial Greenhouse Accelerator Program. The highly-selective program offers funding for startup-related expenses, access to key experts, and individualized support, and it has previously helped launch successful ventures such as Flexport, Otherland, and Workrise (formerly RigUp).
This month, 80–20 intends to conduct a pilot project with a large re-insurance company to continue to validate the technology, and the company will seek fundraising later this spring to build out the user interface. The platform will initially target the 2,900 insurance companies in the U.S. through a tiered subscription model. Long-term, they hope to expand outside of the US and go beyond insurance into broader financial services and agriculture. As 80–20 continues to collect more data, the models will achieve even greater accuracy and provide patterns and insights throughout the world.
It is clear that 80–20 is providing critical insights for the insurance industry and is at the forefront of addressing climate risks, and I look forward to following their progress.
