Fujitsu began a forecast improvement journey in 2000, moving from a spreadsheet-based process to a demand planning tool, says Chapman. Subsequently, the company launched a major initiative about every two years to improve service levels and inventory turns through demand planning.
By 2008, additional improvements from these efforts were "less than measurable" and the company concluded that the forecast was about as good as it could get, Chapman says. "We knew that if we wanted to continue to drive improvements through the demand planning process, we had to consider a different approach."
That is when the company started looking at segmenting demand planning strategies, he says. The idea behind segmentation "is to apply the appropriate inventory management scheme to a particular product based on its value and variant characteristics."
At Fujitsu, value is determined by a product's actual sales and its forecastability, with difficult to forecast products being the focus of the segmentation initiative. "The area where we needed to make changes was on parts that had high value and low forecastability," says Chapman. "We started a collaborative planning effort with our largest customers to jointly plan demand for those products." For low value/low forecastable products, the company implemented a min/max replenishment strategy based on historical demand.
With this segmentation strategy Fujitsu has reduced the average cost of inventory per product by about 30 percent, without negatively impacting service levels. "In some cases, we were even able to improve service levels," says Chapman. "Of course, these savings get passed on to our customers so they have seen improvements in inventory and service levels as well."
The company had a "couple of false starts" with the collaborative planning aspects, Chapman says. "We have highly configurable products, so a single product may have 50 to 60 parts. To sit down with customers and have meaningful discussions on that many parts overwhelmed the process." After the company pared the focus down to 10 or fewer parts, "everybody was focused and showed up prepared to talk to those parts. We had meaningful discussions and everyone was cognizant of what the plans were for those parts any changes were communicated rapidly. So reducing the menu of parts we were discussing really helped us get traction."
The biggest lesson learned in this process was "that we had to redefine what success looked like," Chapman says. "In some cases we were intentionally reducing inventory turns, so we were carrying more inventory than we had originally planned, with the expectation that overall inventory turns would actually increase. Those were difficult lessons, especially for the people who were responsible for managing the products whose turns were reduced."
The segmentation strategy is now in place with Fujitsu's largest customers and the company plans to continue rolling it out with some smaller customers and with new business.
