The Gartner Group’s August 2013 Retail Forecasting and Replenishment Vendor Guide found that retailers’ ability to use demand forecasts still falls way short of the potential value, especially around forecasting demand for new items and promotions.
Why is this? Retailers have been forecasting demand and vendors have been selling demand forecasting solutions for years. Why aren’t we better at this by now?
According to Gartner part of the problem is that “With merchants using optimization to drive assortment, price, and space decisions, often with multiple forecasting engines that source data differently, the combined effects of these strategies are impossible to determine. “
I agree. Ask any category manager today how they forecast demand, and they will quickly rattle off multiple systems they use daily to make price, assortment, promotion or other decisions, each with it’s own embedded optimization model and forecast. The resulting "screen proliferation" not only makes it hard for category managers to efficiently do their jobs, but also prevents them from making merchandising decisions that are fundamentally interdependent in a holistic way.
So what’s the answer?
Retailers need to replace their siloed thinking (and associated siloed optimization/forecasting systems) with a more unified approach that lets individuals from different functional roles across the merchandizing organization – from category managers and pricing teams to space planners and procurement – use a common view of the business and shared platform of analytics and data to make smarter more holistic decisions.
That’s a lot to ask a retailer to do, especially during a period of weak economic recovery and limited IT investment. But with years of effort already expended and limited success to date in improving their effective use of demand forecasting, it’s high time for retailers to re-examine their approach and start thinking about analytics in a more holistic way.