In the past few years, we have seen increasing temperatures, rising sea levels and extreme weather across the globe. According to NASA, 2016 was the hottest year on record and 2017 was the second warmest year on record. In 2017 alone, the world witnessed massive heat waves in the Arctic and Australia, dangerous droughts in Somalia and horrific hurricanes in North America and the Caribbean, just to name a few. As erratic weather promises to continue, businesses, including retailers, are taking note. In fact, the Center for Climate and Energy Solutions found that 90% of the multi-national, blue chip companies on the Standard and Poor’s Global 100 Index recognize climate change as a major risk to business.

Demand for products has always been driven by fairly predictable seasonal weather, both in brick-and-mortar stores and online. It seems obvious that colder weather boosts the sales of things like boots and snow removal products while warmer weather increases sales for sun care products and outdoor lighting. Planning tends to become more difficult when weather becomes less predictable. For example, when people are stuck at home due to winter storms, e-commerce sales do not increase like one may think; instead, when people are home from work, they are busy dealing with things like power outages and household chores, and are not online shopping like they may have been at their desks in the office. What becomes even less obvious is how to plan for extreme weather conditions.

Extreme weather, like storms or wildfires, has the potential to disrupt supply and distribution chains, cause inadequate staffing, render products too scarce or abundant and distort prices. Loss in revenue due to store closures or decreased foot traffic is business that is rarely made up. In fact, atypical weather disrupts the operational and financial performanceof 70% of businesses throughout the world and weather variability is estimated to cost the United States about $630 billion each year. Small businesses and new businesses are especially disadvantaged by extreme weather.

Retailers historically plan for the year ahead based on the seasons. As businesses grapple with extreme weather, though, companies are understanding the value in weather prediction services. After all, the prior year’s weather is only an adequate indicator of the next year’s weather about 15% of the time. By utilizing weather data and analytics, retailers hope to better understand how weather affects customer traffic, sales, staffing, production and pricing.

Lucky for retailers, weather forecasting programs have also become more accurate, thanks in large part to artificial intelligence (AI) which enables researches to analyze massive amounts of weather-related data faster and more efficiently than ever before. IBM’s Deep Thunder, for example, utilizes weather, location and traffic data to predict the weather and therefore assist businesses in making smarter and more informed decisions ahead of time. Interestingly, IBM bought The Weather Company in 2015 to access its data in conjunction with Watson, IBM’s AI platform.

Similarly, there are weather-focused consulting firms that encourage companies to quantify the impact of weather on their businesses by measuring the impact across time and location. Weather-based sales distortions can then be removed from the company’s sales history to create a baseline for planning purposes. In a report by the National Retail Federation (NRF) in partnership with Planalytics, NRF found that businesses that remove the historical impacts of weather from their sales history can drive a 20-80 basis point annual improvement in profitability in inventory management alone.

As extreme weather continues to disrupt the globe, businesses will undoubtedly need to consider ways to develop more weather forecasting technologies and weather-focused planning services.