Saturday, May 14, 2005

Applying Science to Supply Chain Management: Pendulum Swinging Back?

The last 5 years, we have seen companies move away from sophisticated planning tools, towards very simplistic use of supply chain decision support tools and a focus on rapid change of plans based on demand and supply changes. This is changing rapidly now with best practice companies starting to apply advanced science (optimization, statistics, business intelligence) to decision support. Venture capitalists are starting to fund a lot of companies that are focused on providing tools that help enterprises apply this advanced science to planning. Is the pendulum swinging back to applying advanced techniques to planning / decision support?

A few years back, most enterprises did not trust their ability to run sophisticated algorithms for planning and did not see the need for it given significantly diminished demand, because of the downturn, and significantly increased supply, through low cost production centers in Asia. Large scale planning problems took a long time to solve -12 hours, is not atypical counting the time to collect data, to solve the problem and update relevant systems. When optimization was applied to short term planning problems, the decision support tools gave results which were not valid given the changes that have happened since the start of the solving run. There were isolated instances like airlines, utilities and oil and gas companies using advanced math but the rest of the enterprise world was content with simple planning logic.

Today enterprises are starting to understand the competitive advantages of applying advanced science to decision support given the current environment. Bottlenecks are starting to creep up in transportation (ports for ocean shipment, nodes for rail network, availability of containers for road transportation). The charge-backs by retailers for misses in perfect order fulfillment are starting to reach in the millions of dollars per year for large vendors. AMR Benchmark Analytix data indicates that out–of–stocks average 3% on consumer packaged goods and up to 18% for consumer electronics companies. Retailers are storing less SKUs in their distribution centers and forcing more store deliveries with less than 72 hour order turnaround requirement - which causes LTL (Lessthan-Truck-Load) shipments to stores. Shorter product lifecycle means less ability to buffer supply chain misses with inventory. Enterprises have to understand the upside risk and the downside risk that they have with supplier contracts as market demand changes. On the positive side, data (point of sale data from retailers & CPFR based information sharing, inventory tracking data from logistics network accelerated by RFID adoption, manufacturing and procurement plan data from suppliers) is lot more available now and can be taken advantage of for sophisticated decision support. According to AMR Research, in a recent interview with Linda Dillman, CIO of Wal–Mart, she commented that "refining forecasting processes is the single most important area of focus for a consumer products manufacturer in improving service levels to Wal–Mart."

But instead of looking at the opportunity and concluding anything from that, let us follow the money - where venture capitalists are investing their money and where customers are showing interest. Besides established and recognized companies which have focused on applying optimization like i2 Technologies, Aspen Technologies, SAS / MarketMax etc. here are start-up or early stage companies focused on applying science and business intelligence to Supply Chain Management problems - 4RSystems, Evant, Innovative Scheduling, Logictools, Optiant, Prescient, Rapt, Silvon, Smartops, Terratechnology, Valdero, Vivecon.

I predict even more companies will show up in this space.

The pendulum is definitely swinging back. What do you think?

Karthik Mani

No comments:

Post a Comment