After all these years, on-shelf availability remains a problem for many retailers. There are two primary reasons for this state of affairs, says Tyagi. One is market-driven, based on an explosion of channels, the dominance of the internet, and growth of business-to-business commerce. Together they have resulted in a fragmentation of demand.
The second reason is a continuing demographic shift in buying power from Baby Boomers to Millennials. The latter are more fickle, and less loyal to a particular channel. “Capturing their behavior is not easy,” says Tyagi, adding that merchandisers must employ all possible sources of information. They need to achieve a “perfect marriage” of supply-chain planning and merchandising.
Retailers and suppliers still have a hard time with forecast accuracy. Many don’t know how to harness a “treasure trove” of data. In addition, traditional forecasting systems make it difficult to obtain a coherent answer, based on the wealth of information that is available to sellers today. Solutions require many parameters and significant scientific know-how.
Yet another challenge lies in the difficulty of managing more than a million item and location combinations a week. Retailers need to employ heuristics and come up with approximations. But in many companies, planning is an entry-level job. New hires simply don’t have the experience to manage the process, Tyagi says.
Science offers analytical techniques that can provide a deeper understanding of demand patterns, especially during promotions and seasonal peaks. Data must be deconstructed and incorporated into workable forecast models. “You have to look at a better way of doing this,” says Tyagi. Over the next five years, he expects new statistical models to become ingrained in forecasting tools.
