By Ron Menich, EVP and Chief Scientist, Predictix
Are retail sales in a slump? Don’t just blame the weather—that’s an easy cop-out, but one not likely to fly well with your CEO. Instead, your organization needs to understand the impacts of weather and understand how best to use that knowledge to properly position inventory ahead of demand.
Advertisement
According to a recent CNBC article, retailers often explain away weaker-than-expected sales results by saying, “unfavorable weather conditions negatively impacted sales.” The article points out, “a number of retailers and analysts blamed tepid sales and traffic over the early summer weeks on an unusually cooler and wetter late spring and early summer. For some, the weather truly could have been a headwind.”
There is absolutely no doubt that weather impacts retail sales. But what separates high-performing organizations from the rest of the pack is what actions they take to position inventory to coincide with demand, given knowledge of weather impacts.
A hot spring day can mean tables full of sweaters with no one buying them and, at the same time, empty shelves where the fans are typically stocked. As retailers know too well, weather variances can wreak havoc with seasonal merchandise—leading to overstocks and markdowns, or else out-of-stocks and missed sales. More than the absolute value of temperature or humidity, it is the variability from the norm at a given location and time of year that is key to understanding the demand impacts of weather (e.g., a 39 degree day in February might be warm in Minneapolis but cold in Houston). Many in the scientific community have concluded that global climate change will increase climate variability, and if so, retailers will have to grapple with weather impacts on demand for a long time to come. As Planalytics, the business weather intelligence provider, sums up: “on a day-to-day basis, weather is the most volatile external factor influencing consumer and market behavior.”
That’s only a small part of the weather-related challenges that retailers face, however. The problems multiply if retailers don’t consider the effect of abnormal weather patterns when planning next year’s demand forecast. If retailers simply pick up last year’s sales data as the basis for their new 12-month demand forecast, they can seriously underforecast or overforecast, and end up with out-of-stocks or overstocks. That’s because weather repeats itself year-to-year only 25% of the time, which means if retailers plan or forecast using last year as the template, they will be wrong a staggering 75 percent of the time.
To avoid this, retailers must “deweatherize” the sales data—in other words, create a weather-neutral forecast that is based on historical norms, which is increasingly important as unpredictable extreme weather events occur with growing frequency. The deweatherized plan accounts for the year-over-year demand volatility generated by changing weather patterns and provides a better baseline for future plans, improving accuracy by as much as 3X.
What can retailers do to properly account for the effects of weather in their demand forecasts? Here are three suggestions that can lead to better forecasts, rain or shine:
· Start with a more accurate forecast. A surprising number of retailers rely on outdated demand forecasting systems that cannot effectively analyze the huge amounts of data that can feed into a retail forecast; these older systems get around the limitations by taking shortcuts and making analytical compromises. Moreover, retailers can’t afford the IT infrastructure that’s needed to process all this data. A new generation of cloud-based forecasting systems can help retailers overcome these limitations by providing access to the unlimited computing resources of the cloud, which is ideal for highly sophisticated, configurable and up-to-date forecasting science that utilizes all kinds of information—including weather data.
· Utilize all the data. Big Data is all the rage in retail these days, and complex weather information is simply another form of Big Data. The best retail forecasts are able to consider all of the available data: sales data and multiple promotional attributes, social media, door counts, product reviews, customer loyalty information, weather patterns and any other variable that can affect demand.
· Deweatherize your demand forecast. Once you’ve considered every possible variable, including weather, the final step is to weatherproof the forecast. Deweatherizing your demand forecast will help ensure that the final result isn’t based upon last year’s unique weather conditions —greatly improving your ability to strike the perfect balance between inventory and demand.
Ron Menich is one of the foremost innovators in predictive analytics for Fortune 500 enterprises. He previously served as Chief Scientist of Pricing and Revenue Management at JDA, where he was the conceptual designer of many of the core modules in JDA’s Pricing and Revenue Management suite. Menich is a proven professional with 18 years of experience in demand forecasting, optimization and large-scale recommendation processing systems. He holds a bachelor’s degree in mathematics from the University of Illinois at Urbana-Champaign, a master’s degree in operations research, and a doctoral degree in industrial and systems engineering, the latter two from the Georgia Institute of Technology.
