What real-time postponement looks like
Gillette's successful postponement strategy for a finished goods network relies on optimization software.
By Roberto Michel, Editor at Large -- Modern Materials Handling, 9/1/2006
 |
| Optimization software helps Gillette stage the correct inventory levels for each SKU flowing into the pack centers.
|
To optimize transportation costs, Gillette began to co-locate product from multiple divisions, such as Right Guard deodorant, Oral-B toothbrushes and Duracell batteries, in the same regional pack centers. While strategy and facility design were important, the company added to the mix inventory optimization software (Optiant, 781-238-8855). It determines the optimal distribution network for a particular product line, and the required safety stock for each stock keeping unit (SKU) to achieve desired service levels.
Simply put, the software takes inputs like customer service, demand variability, lead time and cost through the supply chain and calculates the optimal supply chain, according Maggie Walenty, a senior inventory modeling analyst with Gillette.
"We use the software to assist us in determining the optimal number and location of distribution centers in a finished goods network," says Walenty. "We use the output of our models to help determine the number of pallet spaces required in a distribution center."
In addition to the regional pack centers, Gillette's postponement strategy also relies on a central supply warehouse. This facility, says Walenty, reduces the amount of excess inventory that might exist at regional DCs if there is a forecast inaccuracy. It also reduces the costs associated with redeploying or reworking inventory from multiple DCs.
While a number of systems support execution of the postponement strategy, Gillette found that inventory optimization was important to designing the network and the inventory levels.
"There is a common assumption that postponement always improves the supply chain," Walenty says. "This is not true, however, without optimization. Until we implemented the software, we did not have a robust method to examine the 15,000-20,000 SKUs that we manage globally and determine the optimal supply chain for each of those SKUs."
View All Blogs
