Reportedly, for some time the demand planning solution from the incumbent Demand Solutions was doing an adequate job at Wayfair, including in terms of attribute-based forecasting for retail items that come in multiple variations. But as the number of stock-keeping units (SKUs) increased and lifecycles shrank, the challenge became to forecast a wide variety of heterogeneous demand behaviors (black, white, and shades of grey inbetween), such as at different stages of the product lifecycle (i.e., new product introduction [NPI], mature product, end of life [EOL]). If the process is manually intensive, imagine how unmanageable it becomes for an Internet retailer like Wayfair, who carries more than a million items.
“We had outgrown our existing infrastructure,” said James Saravese, Wayfair’s senior vice president of operations. “ToolsGroup offered us the opportunity to introduce an entirely new level of automation and intelligence into the planning process.”
ToolsGroup has self-adaptive functionality that automates this process for items with lots of attributes/options and also brings an attribute-based forecasting capability it calls “Macro-Item,” which originated with its apparel and fashion industry customers. The frequency-based forecasting sifts through the frequency and quantity history of individual SKU line orders to produce reliable statistical forecast at the item-location level and identify inflection points in the SKU lifecycle.
TEC article (Feb 2012): ToolsGroup—Going Back to Its SCP Roots