Getting more precise demand forecasts

Wistron gets a clear picture of long-term inventory flow

To avoid the cost of producing and storing excessive overstock, Wistron needs precise forecasts that provide an accurate picture of future demand and inventory flow.

Wistron, a leading original design manufacturer (ODM) for information and communication technology, uses SAS^® Service Parts Optimization to forecast parts demand with greater accuracy and to improve inventory transparency. As a result, the Taiwan-based company gains greater efficiency and, ultimately, more revenue.

In addition to its successful ODM business, Wistron has achieved eye-opening results in the service business. Service parts account for much of the business group’s revenue. It is customary to provide service parts for consumer electronics for at least 39 months after a product goes off production. For commercial electronics, the period extends up to seven years.

 

Long-term accuracy

Wistron has built a superb parts management system to deal with global business and demand forecasts. In order to implement SAS Service Parts Optimization, Wistron created SPOS, the Service Parts Optimization System – a refinement to its existing system that has helped the manufacturer keep pace with thriving business and deal with more complicated business procedures and data-processing requirements.

SPOS can manage and process data from ERP, customer service and maintenance systems, as well as front-end applications. As it imports data, SPOS runs statistical analyses for demand forecasts and replenishment plans. And it directs purchasing activity.

“We did have an SOP that would forecast parts demand for three to six months,” explains Jack Chen, Senior Director of Operation Division, “but we needed to create forecasts for three to six years. And the SAS solution offers a strong statistical analysis engine."

Though akin to SPOS in structure, the former system could only produce short-term forecasts. For mid-term and long-term forecasts, Wistron needed a system with stronger data-processing capacity and the ability to integrate detailed data from multiple sources.

In certain circumstances, when a product line phases out of the market, forecasting parts demand can be difficult. Overestimates could lead to overstock and waste; underestimates might trigger customer complaints. With SPOS, Wistron aimed high to achieve two goals: accurate forecast and inventory transparency.

Using SAS for cluster analysis, Wistron analyzes service parts with similar consumption levels with respect to different clients and product lines. The analysis is further processed into a trend model as a reference for forecasts.

“A precise forecast requires more than just experience,” Chen says. “Statistical analyses are imperative to build up long-term forecasts. With SPOS, we can successfully obtain accurate forecasts.”

Users can add non-default parameters to obtain higher-value results. For instance, if through statistical validation one non-default parameter was found to be significant in forecasts, this parameter could be added to the SPOS to perform forecasts and planning activity.

Global inventory view in one interface

While the previous system could only import data from limited service locations, the new SPOS allows users to monitor inventory across a global network of service points. SPOS manages inventory flow and sends purchasing or inventory allocation suggestions to planners around the world.

“Wistron has a wealth of experience in ODM and is familiar with the supply chain,” Chen says. “SAS has helped different industries implement the system. The combined effort between us triggered more ideas and inspiration.”

Reduced overstock

With support from management, Wistron enjoyed early success with the implementation. Next it plans to extend the application to service points globally.

As part of Wistron's goal to reduce inventory while improving customer service, the manufacturer understands the importance of influencing consumer behavior. When offered steeper discounts, for example, shoppers tend to purchase new instead of paying for repairs.

“These changes undoubtedly influence our business strategies,” Chen says. “It is also a new challenge for our planners to interpret the data derived from the analyses.

“Wistron constructed SPOS with technical support from SAS, and the system has successfully helped us achieve much more precise forecasts of parts demand and enhanced inventory transparency. With this, Wistron effortlessly deals with the complex management issues that arise due to our successful and growing business.”