Managing Complexity in the Supply Chain – Part I

I recently sat on a panel of supply chain executives in Las Vegas, at the Better Management Live! Meeting sponsored by SAS. The theme of this panel was to have several people reflect and brainstorm about what supply chain complexity is, how it relates to product design, how it can be measured, and whether new supply chain strategies such as eAuctions and global sourcing is increasing complexity. Finally, we were also challenged with the problem of how to manage supply chain complexity, in an increasingly complex supply chain environment. Seated on the panel were Jonathan Whitaker from Accenture, Lowell Hoffman (a retired Senior VP of Purchasing at Kraft foods and Colgate Palmolive), Ken Newton (a retired Senior VP of Purchasing at Texas Instruments and Hewlett Packard), and Jeanne Staudacher, Senior VP of Supply Chain Management at Motorola’s cellphone division. (Needless to say, I felt a little intimidated by the level of expertise on this panel!)

We began by discussing how product design impacts complexity – the feeling on this issue was that there definitely was an impact. Anytime that marketing can reduce the number of options or SKU’s available to customers, the resulting impact on the supply chain is immediately felt. One goal was to be able to provide the “appearance” of endless choices, such as the Dell online ordering model, which in fact directs users to a limited number of SKU’s and options. Even though many automotive manufacturers such as Toyota and BMW are pushing the “build to order” model a la Dell, this simply isn’t congruent with their existing manufacturing systems. I brought up a discussion I had with a senior executive from Toyota who was presenting at the Southern Automotive Manufacturing Conference last year in Greenville, SC. When I asked him how he was reconciling the increased complexity demanded by the BTO model with the inherent limitations on product options required in the lean manufacturing model of the Toyota Production System, his only response was “We’re still working on that!”

Measuring Product Complexity
Next, we discussed the problem of measuring product complexity in terms of supply chain impacts. Jeanne Staudacher at Motorola discussed a measure of complexity that Motorola had introduced in response to this problem, that was used as a focus point for discussion. The measure was initially introduced by Theresa Metty, Jeanne’s predecessor. (Details regarding how this measure was integrated with their other initiatives at Motorola are discussed next time). Jeanne noted that “We’ve actually identified 10 complexity factors which include the following:

Approved by purchasing
Volume per part number in inventory
% of industry standard parts

Volume per model (at what point do you customize in the system)

Design for X (DFX)
Design for assembly
Design for part count
Design for reuse
Design for postponement

How easy is it to pull apart

Cycle time

Staudacher notes that “The key to using this measure is to assign a metric for every one of those 10 factors on our products and our competitors’ products. If a product has a complexity index of 1.0, it means it’s at parity with best in class. If it’s anything over 1.0 — if it’s 1.5 — that means it’s significantly more complex than our competitors. The only way we could benchmark it against our competitors was through basic engineering “teardowns” and secondary research on their websites. We would estimate their volumes, tear down the product and assign metrics to all of their components at the product level and the portfolio level, and benchmark it against our product, which then led to a question of how to decide on the different tradeoffs. This was lots of work and required some really good engineers – so we do it every 6 months every time there is a product refresh.”

The panel also decided that a good proxy measure of supply chain complexity is cycle time – how quickly can you respond to the DRIVERS of complexity. Some of the primary drivers in addition to design are the number of SKU’s, promotions, regulation, seasonality, and others. These drivers have to be weighed off against the value to customers.

Jonathan Whittaker commented that one could envision this as a two by two matrix, with complexity as a cost (lo versus high) and value to the customer on the other axis (low versus high). Obviously, one would want to eliminate products that are high cost and low value to customer – and focus on those elements that are of high value to the customer. However, those that are in the “high cost, high value” quadrant should be priced higher according to customers’ requirements. Using this matrix as a filter may be one way of thinking about how to better manage supply chain complexity.

Ken Newton had another perspective. “I never went to work and thought to myself ‘I’m going to work on reducing complexity today!’ Instead, you have to decide where do you let it in? This involves first recognizing complexity when you see it, and deciding whether or not you are willing to live with it or not. You need to attack each example when you encounter it – and there really is no “one size fits all” strategy. You also have to be aware that people work in silos – reducing complexity in one area (procurement) could create additional complexities in another silo (marketing or operations). That means you need to look at it holistically across the organization.”

Lowell Hoffman also discussed how complexity can have a major impact on cost. He cited the example of IBM. In the past, each IBM division – PCs, servers, storage devices and others – operated its own supply chain, running distinct manufacturing lines, negotiating its own deals with components suppliers and using its own billing systems and ways of naming parts and products. By standardizing across drivisions, IBM was able to reduce supply chain costs by over $3 billion, which did not count an additional $2.6 billion in price reductions occurring in component markets, for a total of $5.6 billion in savings. That required division leaders to give up control of sourcing, manufacturing and even design of some products. This also brought up a discussion of how organizations may need to restructure in order to deploy complexity reduction initiatives.

The increasing globalization of supply chains was also discussed as a driver of complexity. The impact of SARS made managing Asian suppliers very complex. In addition, the events surrounding 9/11 caused problems with shutdowns at Motorola, TI, and others. Finally, the increasing problems associated with counterfeiting pharmaceuticals, terrorist activities, and heists on semiconductors creates additional security problems associated with managing complexity.

The thoughts shared by these people will be shown in a webcast later this month on