Planning for the Afterlife

Regardless of your market sector, at some point you'll need to make the decision: Keep a product on life support... or pull the plug?

Most semiconductor manufacturers hope to keep their product designs alive either until the competition out-designs and under-prices them or the demand evaporates. When either will happen, whatever the reason, is something few OEMs can control.

The decision whether to keep a legacy product active or retire it can become a real balancing act. Product demand, technical support, component availability, competitive landscape, required resources, new product introduction, and product roadmaps all play their parts. And, unless and until the decision is made to retire a product, the evaluation and analysis are ongoing. Return on investment (ROI) and market matrices are always on the move. They must be monitored constantly.

Lifecycle expectations have a large impact on forecasting demand for ongoing product support. In the world of finished goods, consumer electronics have an average life expectancy of nine months, commercial products about two years, industrial three, medical devices five to seven, and military up to twenty years.

While a component is still being actively manufactured, the demand volume and lifecycle are often somewhat known in advance. Ramp-ups and ramp-downs for a product manufacturing line can usually be planned beforehand, at least at some level. However, once a product has reached end-of-life (EOL), the demands change. Forecasting EOL demand is different from forecasting demand during the life of a product. Basing spare parts demand at EOL based on demand in the manufacturing phase is difficult at best.

Typical problems related to the EOL phase in the product lifecycle are very uneven demand patterns, difficulties in ordering -- or even locating -- spare parts, and a high probability of stock-outs or shortages. For most parts, the need to support warranty and maintenance requirements lasts longer than the period during which the necessary components are available from the manufacturer.

When it comes to EOL planning, the ultimate goal is to ensure that there will be enough (but not too much) of the right components in stock, throughout the life of the product. However, determining when to replenish a given SKU and the quantity to inventory and store can be problematic. Since EOL doesn't last for the same number of months or years for all products, over time the complexity increases as more EOL products are added to the mix.

There are multiple approaches to managing EOL parts procurement and management. Whatever the approach, with the potential need for thousands of items to be inventoried and managed, a systematic approach is an absolute requirement.

Setting up an internal process is one way to handle EOL component management, especially when the number of products is small and life expectancy is short. However, given the ongoing nature of the process, as additional products reach EOL, OEMs can expect to have to increase the resources and budget allocations for EOL planning, management, and inventory control.

Another approach, especially when the number and volume of components is large, is to partner with an independent distributor with the capacity and capability to provide ongoing EOL component management and support for any number of products throughout their entire active lives. This relieves the OEM of the need to divert resources from its core business to address the myriad issues associated with EOL planning and management and makes it possible to budget more accurately for them.

Have a comment or a horror story about EOL? I'd love to hear from you.