As the concept of scale evolved, increasing attention was paid to the trade-offs — for example, scale versus distribution. Having scale entails large centralized plants, but companies still have to get the goods to the customers, who may be scattered far and wide. Trade-offs also arise in warehousing and inventory costs. Economic Order Quantity (EOQ) — a formula developed in the 1930s to guide managers in setting order sizes by estimating such variables as inventory holding cost, cycle time, and demand — was one attempt to quantify this problem. Today, elaborate linear and integer programming software is available to “optimize” all trade-offs.
Unfortunately, sophisticated tools rarely guarantee sound thinking. For example, many dot-com ventures requiring large capital investments were defended with the concept of scale. The online grocer Webvan Group Inc. signed a contract with Bechtel Group Inc. to develop 26 automated “picking and packing” warehouses for $30 million each. These warehouses were justified on the grounds that their huge scale and high throughput would give the firm a cost advantage. In fact, if the volume had ever materialized, the concept might have been solid. Webvan erred fatally, however, in assuming that economies of scale in one area (picking and packing) applied to its entire business. They did not. In the “last mile” — the final step in getting the goods to the customer — there are few economies of scale. Indeed, there may well be diseconomies of scale in the last mile. Venturing beyond certain very dense pockets of population (such as Manhattan’s Upper West Side), each incremental customer might increase the overall cost to serve; this sad fact, in turn, could prevent the provider from ever reaching the size necessary to capture the benefits of scale in its large centralized operations. (See “The Last Mile to Nowhere: Flaws & Fallacies in Internet Home-Delivery Schemes,” by Tim Laseter et al., s+b, Third Quarter 2000.) Webvan, Kozmo.com, and other online delivery companies failed in part by overapplying the concept of scale.
That failure was based on what we call “binary” thinking about scale: In industries where scale is believed to apply, it is generally thought to apply always and everywhere. We saw this phenomenon at work in the seafood industry, where the concept of scale was applied to the entire manufacturing process, with no distinctions made among loining, canning, and can making. On the other hand, where scale does not seem to apply — when there appears to be no advantage to size and concentration — its usefulness is too often summarily rejected. Witness the case of the battery maker that assumed because batteries are heavy (and thus expensive to distribute), scale could not apply to its business. Following that logic, the company built plants all over the country to be close to its customers, meanwhile losing out on virtually all the benefits of scale.
Recent work with a number of manufacturing clients has led us to question this binary perspective and develop a new way of looking at scale that is both intuitive and, in its way, a bit radical. We have come to believe, in manufacturing as in so many other affairs, that size may not matter. By breaking the value chain into its component parts, a company can adjust its operations to match the underlying scale economics of the specific process, like a seasoned angler custom-weighting the individual segments of his rod.
This reinvention of scale has clear implications for manufacturing industries. Manufacturing is a competitive, aging universe; decades of evolution have driven fat from many processes, while industry players seek continuous operational improvement. Each decade or so yields new techniques. Henry Ford found his panacea in vertical integration; in the 1980s, Ford Motor Company embraced W. Edwards Deming’s quality crusade. The 1990s saw a new version of Demingism (Six Sigma), as well as strategic sourcing and supply chain management.