Mark’s Work Warehouse checks the weather to boost sales

With more than 300 stores across Canada, Mark’s Work Warehouse has increased sales by using a web-based weather-prediction service that lets it know as much as 18 months ahead of time whether to play up shorts or coats in its merchandising plans, the retailer says.

Mark’s, with stores ranging from 2,500 to 20,000 square feet, has long struggled to project weather patterns to better plan merchandise assortments. With 70% of its business tied to weather, “it’s difficult to plan and market assortments for all our stores,” Colin Laker, vice president of merchandising services, said this week at the Merchandising Innovation & Xcellence conference in Chicago, which shared space with the Retail Systems 2006 Conference & Exposition.

“We’ve tried using the Farmer’s Almanac and sticking our finger in the air to see if weather can help plan our business,” he quipped.

But using the web-based weather-prediction service from Planaytics Inc., which combines weather and sales data to produce merchandising recommendations, Calgary, Alberta-based Mark’s can now plan as much as 18 months ahead of time what its buyers and store managers for each region should emphasize in merchandise assortments, Laker says.

Mark’s executives, he adds, were initially surprised at the ability to improve sales through such long-range weather predictions, but were won over after the first time Planalytics advised that a warming trend called for a strong focus on warm-weather products. “It worked,” he says. “We actually did sell more summer than winter products.”

By having more shorts in its stores during a July-August period that Planalytics had predicted would be warmer than usual, Mark’s sold the items at average of $1 more per pair than it had expected, earning back more than the cost of the Planalytics system, Laker says.

Because Planalytics charges its fees based on the number of products for which it provides weather-based merchandising recommendations, Mark’s has figured out how to include a limited number of products that will produce usable recommendations for its entire product line, Laker says.

Mark’s has not revealed the overall increase in sales it gets from using the Planalytics system, but Laker says the weather-based merchandising recommendations could account for “15% to 25% of our sales increases.”

Mark’s sends historical SKU-level sales data to Planaytics, which combines it with weather predictions formed by its team of some 40 meteorologists to produce a report that recommends which product categories and items will sell best. Mark’s receives and views the data through a web interface over a series of time segments; the shorter the lead time, the more detailed the predictions of how weather will affect sales of products.

Data for 18 months ahead produces recommendations for general category merchandising; data for 9-month and 6-week lead times produce recommendations for specific items, such as winter boots or raincoats. Mark’s also receives data for 2-week lead times for making additional adjustments to its merchandise mix.

Other retailers using Planalytics, which is based in Wayne, PA, include J.C. Penney Co. Inc., The Home Depot and Gap Inc., a Planalytics spokesman says.