Professor David Simchi-Levi Provides Insight on Push-Pull Strategy

By Monica Nakamine
September 6, 2002

Uncertainty is inherent in every supply chain since customer demand can never be forecasted accurately. Simultaneously, delivery lead times, manufacturing yields, and component availability may vary and can have a major impact on supply chain performance. On July 23, 2002, as part of an ongoing series of webcasts for partner companies and alumni hosted by LFM-SDM, David Simchi-Levi, professor of engineering systems at MIT, provided insight on a new supply chain paradigm that could help manufacturers maintain or increase levels of service and at the same time reduce supply chain cost – the Push-Pull strategy.

Traditional Supply Chain Management

In supply chain management, there are many pitfalls – long lead time, uncertain demand, increased product variety, component availability, and system variation – that can oftentimes mean the downfall of a company, says Simchi-Levi. For instance, in an automotive supply chain, it could take 60 to 70 days to complete custom orders, the variety of products can create a high level of demand uncertainty, and it is sometimes difficult to match supply and demand.

These pitfalls are usually associated with traditional supply chain strategies that are referred to as Push Strategies. In a Push-based supply chain strategy, all production and distribution are based on demand forecasts. Simchi-Levi says that there are three principles of all forecasting techniques:

  • Forecasts are always wrong.
  • The longer the forecast horizon — the worse the forecast is.
  • Aggregate forecasts are more accurate.

Simchi-Levi suggests that a number of innovative companies have recently taken advantage of the third principle by implementing a Push-Pull system. In such a system, one identifies the portion of the supply chain time line in which uncertainty is relatively small and manages this part using a push strategy, which is based on forecast. On the other hand, the portion of the supply chain time line in which uncertainty is relatively high is managed based on realized demand, or a pull strategy.

Thus, in manufacturing, push-pull is a make-to-order system in which the supply chain manages component inventory based on forecast, while assembly is driven by accurate customer demand. In his talk, Simchi-Levi provided examples from other industries, including books, automotive, grocery and the furniture industry that illustrated how various companies implemented this concept.

This session also was notable for being shared with the University of Michigan’s Tauber Manufacturing Institute (TMI) students. LFM and TMI actively cooperate in ways to improve both programs.

“Sharing between the programs has been of mutual benefit for a number of years through the National Coalition of Manufacturing Leaders,” said Paul Gallagher, research associate for LFM-SDM who coordinates the webcasts. ” Leveraging technology as we did in this session is an example of how we continue improving how we work together.”