Using data to prepare for extreme weather


Weather can be a primary driver of consumer demand and behavior. Now there’s a service spearheaded by IBM that integrates geographically precise weather-related data and analytics with retailers’ point-of-sale and marketing/operating data to offer the potential of delivering operational savings and revenue growth.

That potential was demonstrated in a survey of 1,000 C-level executives representing 13 industries in 15 countries, “How Weather Insights Can Grow Your Bottom Line,” conducted by the IBM Institute for Business Value and Oxford Economics.

Several of the organizations surveyed are early users of IBM’s Watson system and IBM’s Weather Company division, a 35-year-old business acquired three years ago and expanded to integrate weather data with Watson analytics.

The Weather Company’s services can be customized to provide weather-related insights on ways to enhance marketing, merchandising, operations, finance, administration and the supply chain, says Paul Walsh, director of weather strategy and business meteorologist for IBM’s Global Business Services.

Nearly all surveyed executives believe improved weather insights can reduce annual operating costs, with nearly a quarter saying the savings can be between 2 and 5 percent. More than half expect the same cost reduction potential for legal, insurance and risk mitigation costs. And one in five put the annual revenue growth opportunity at the same percentage.

Those opportunities translate to additional revenue growth of up to $50 million per $1 billion of annual revenues, the report said; scaling to the Fortune Global 500 could amount to annual revenue growth of more than $500 billion.


“By understanding what a weather forecast is and layering on top of that analytics derived from POS data, you can get to a very precise understanding of how that weather forecast is going to shape demand, whether consumers are likely to go to a store or shop online from home,” Walsh says.

“Even more importantly, you’re able to determine what people would be going to a store or online to buy.”

According to the report, a retailer using weather-related data could improve its in-store upsell/cross-sell rates from the bottom quarter to the median, which could potentially increase retail revenue by $45 million for every $1 billion of sales.

Marks and Spencer uses weather data to help predict the level of demand it can expect for items across its U.K. stores, says Ricky Mompalao, head of forecasting for the British retailer. “We use this information to plan our stock and, at the same time, minimize waste. To help us get this right, we’re looking at weather forecasts weeks in advance with a level of confidence that means we can better prepare our stores.”

Walsh says a national drug store chain used advance weather data from IBM to “reach and engage consumers when and where symptoms were likely to flare up or when prescriptions were most likely to be filled.”

By identifying the right consumers and reaching them in the right moments, the drug store chain’s stores improved store traffic and key advertising statistics, he said.

Minneapolis-based Red Wing Shoes generated 21,000 incremental visits, a 41 percent lift, during a recent promotion done in partnership with IBM’s Watson Advertising. Walsh says Red Wing Shoes worked with Watson Advertising across its more than 500 stores to match individual transactions to weather conditions. It then used that analysis to create customized WeatherFX triggers sent to customers when weather conditions were likely to cause a spike in work boot sales.

Through data optimization, Watson Advertising reallocated 19 million of 74 million total impressions to better performing weather conditions, saving nearly one-third of Red Wing’s plan from under-achieving placements, Walsh says.

For one high-volume client in the United States, the Weather Company currently provides integrated weather-related data to help its lawn and garden store managers determine whether they should water their trees and plants.

“By doing this, they were significantly able to reduce waste,” Walsh says, “and the savings for them is significantly more than the fee they are paying for the service.”

He says IBM is currently working with several retailers and consumer goods companies “as both industries rethink how they can leverage weather-related data from being reactive to proactive.”


The volatility of this year’s extreme spring weather fluctuations, from unusually cold to unusually warm, caused shoppers to postpone buying seasonal items in April. And when they did begin buying in May, many retailers had already taken markdowns; while sales did rebound, profits went south.

For most marketing applications, Walsh says, forecasts are typically applied days out from a planned event. From a supply chain perspective, forecasts are typically a month or six weeks out.

“From a distribution perspective, you can anticipate weather-related demand about five weeks out and then make sure there is enough product in the distribution centers to meet that demand,” he says. “We can also do long range regional forecasts that go out about four months.”

Store managers can use the service to determine staffing requirements, allocating fewer hours to staff on days when the weather is projected to be bad. Loss prevention also benefits, gaining insights into when the weather is projected to be nice, which would drive more traffic into stores, potentially creating a greater likelihood of higher theft.

Before a retailer fully commits to a rollout of IBM’s program, it first conducts a “proof of concept or value” to determine how much of a return it can get from their investment. “The price they will pay is relative to the value they will be getting,” Walsh says.

Some retailers who have committed to the program manage the integration and analysis of the data for themselves; others leave the analytical work mostly to the Weather Company.

Overall, “retailers are participating in a massive transformation, realizing and reacting to the fact that weather insights can drive their bottom line,” Walsh says. “Ultimately, we can’t control the weather, but weather insights are helping retailers and others do a better job controlling their reactions to the weather.”

Liz Parks is a Union City, N.J.-based writer with extensive experience reporting on retail, pharmacy and technology issues.


Comments are closed.