By Sridhar Tayur
Inventory optimization is one of the hottest topics in supply chain circles today — and rightfully so. As a growing number of organizations have proved, astute planning and management can wring 20-30 percent out of current inventories, saving "multiple millions" in direct costs and achieving huge gains in operational performance, all while maintaining — or even improving — product availability and customer service.
Given the enormous benefits at stake, it's no surprise that a recent survey by Aberdeen Group places inventory management software at the top of the list for supply chain technology investments. Within that category, the highest priority is clearly multi-stage inventory optimization, which generates optimal inventory levels for each item across each of the stages or tiers within an organization's supply chain network. More than 80 percent of respondents cited multi-stage inventory optimization as a top priority, nearly twice the number who named any other type of inventory management technology.
The Case for "Multi-stage" Inventory Optimization
This heightened interest in optimizing every step in the supply chain is a function of the increasing size and scope of today's networks. As supply chains grow, often stretching across the globe, so does the complexity of homing in on appropriate, time-varying inventory targets at each point along the way. In fact, because the steps are interdependent, the complexity multiplies. At the same time, the inherent risks and uncertainties at each point virtually ensure that more problems will occur and that when they do, they will send shockwaves up and down the supply chain.
The challenges of managing every type of inventory that an organization maintains — safety, cycle, pre-build, pipeline and merchandising stock — at multiple locations and/or stages in its supply chain are indeed formidable. The complexity of the task has clearly outgrown many of the planning processes and tools that organizations have adopted even in just the past few years. Quite frankly, it outgrew the best efforts of human beings long before that.
Multi-stage or "multi-echelon" solutions provide powerful capabilities for modeling the most complex supply chains and analyzing a staggering number of variables, constraints and "what-ifs." Sophisticated algorithms, combined with a stochastic (probabilistic) approach, enable multi-stage inventory solutions to assess vast amounts of historic and real-time information, considering multiple variabilities and interdependencies.
A case in point is Deere & Company's Commercial & Consumer Equipment Division, which implemented a solution to optimize inventory levels for more than 300 commercial and consumer equipment products held at 2,500 North American dealer locations, plants and warehouses. To do so, the software considers 52 million variables and 26 million constraints. In four hours each week, the system generates optimal targets that have enabled Deere to reduce inventory by more than $1 billion, while significantly improving on-time shipments from factories and maintaining customer service levels at 90 percent or better.
Another example is packaged food giant ConAgra, whose supply chain includes 65 manufacturing facilities and a network of co-located buffer warehouses that feed into 14 mixing centers. Early results from ConAgra's multi-stage inventory optimization project include a reduction in finished goods inventory and significant improvements in case-fill and store in-stock percentages, plus reductions in forecast error and bias. ConAgra recently reported that, after less than one year of implementing its multi-stage inventory optimization solution to set policies governing downstream stocking locations, supply chain savings and productivity gains are ahead of target.
Moving on: Enterprise Inventory Optimization