The adoption of distribution requirements planning systems in the late 1980s triggered the development of simple but robust models at P&G for setting inventory targets across a distribution network, one echelon at a time. Multiechelon technology from Optiant was selected to serve as its future inventory optimization platform. P&G executed a successful pilot in its most complex beauty care supply chain in 2005–2006.
The analytics that underlies the multiechelon inventory planning engine is based on the guaranteed service model of safety stock optimization. Although the basics of this research stream have been known to apply to strategic supply chain design for years, considerable advances in both operations research and computer science have been required to solve the tactical inventory problem at P&G’s scale. These analytics advances include modeling nonstationary demand to reflect the seasonal nature of consumer demand, modeling review periods to capture different replenishment frequencies across the supply chain, and most significantly, the ability to optimize inventory levels and locations in acyclic network topologies. It is this last advance that has allowed P&G to capture the multiechelon nature of their supply chains.