By Nicholas P. Brown

(Reuters) -Big-name retailers such as Walmart are increasingly using analytics to blunt the impact of one of the most unpredictable performance variables of shopping: weather. Weather data, once used strictly for inventory planning, is now helping retailers localize advertising and decide when to discount seasonal items such as sweaters. Walmart, whose inventory planning with artificial-intelligence software incorporates weather analysis, reduced sunscreen prices a couple weeks earlier than usual this year in parts of the U.S. Weather data forecasting a wetter-than-usual autumn in some U.S. regions was a factor in its decision, whereas several years ago, it likely would not have been, said Kirby Doyle, a skin-care category replenishment adviser to the world's biggest retailer. "In the beginning, (weather data) was just a forecast model for high-level planning," said Doyle, who works for Beiersdorf, which makes personal-care products. "Now we’re infusing it into pre-season planning and throughout the season to diagnose the impact of weather, and for things like scheduling promotions.”

Walmart declined to comment. A niche group of weather consultants — from Germany's Meteonomiqs to U.S. firms Planalytics and Weather Trends International — is using breakthroughs such as cloud computing to process once-unimaginable amounts of data. Demand for such data is growing amid heightened weather volatility due to climate change. The National Retail Federation, which is chaired by a Walmart executive, issued a report with Planalytics in July, recommending retailers pay more attention to weather analysis. New weather-data tools, centered on pricing, may soon be hitting the market. Planalytics and BearingPoint, a management consultancy, are partnering to build software retailers can integrate into their analytical models for setting prices.

“Weather is something you can’t control,” BearingPoint managing consultant Ryan Orabone said at an industry workshop last month to unveil the new initiative. “But you can control the analytics. And pricing, you absolutely control.” It is natural for a warm October, like this year's in the U.S., to cause retailers to sweat ahead of the holidays. "It needs to get cold for our business to really perform well in Q4,” Tractor Supply CEO Hal Lawton said last month on a quarterly call.

The company, which uses weather analytics, sells cold-weather products like heating pellets and outerwear. Weather analytics can help companies like Tractor Supply decide whether to discount winter items, said Planalytics CEO Fred Fox, whose clients include Dick's Sporting Goods and Ross Stores. If November temperatures in the U.S. drop below 2023 levels - which forecasts suggest is likely - a discount now could mean a missed opportunity later, Fox said. As intuitive as that may seem to a retailer, they do not always get it right. In August, Lowe's Chief Financial Officer Brandon Sink cited cold, wet weather in May as the reason for weaker sales in the prior quarter. But that description is inaccurate, said Bill Kirk, founder of Weather Trends, whose clients include Target, Gap, and Tractor Supply. May was indeed wet, Kirk's data shows, but not cold. It was the hottest May in six years for the U.S., he said, and third-hottest in four decades. "Welcome to the world of retail excuses not based on facts," he said.