2015-05-29 08:31:51
Willy C. Shih
May 28, 2015
Technologies like 3-D printing, robotics, advanced motion controls, and new methods for continuous manufacturing hold great potential for improving how companies design and build products to better serve customers. But if the past is any indicator, many established firms will be slow to adjust because of a formidable obstacle: legacy assets and capabilities that they are reluctant to abandon. Why are older incumbent firms slow to adopt new technologies even when the economic or strategic benefits are clear?
The literature on this subject is enormous. Much of the early work focused on the adoption rate of new technologies following an S-curve, with some users going early, a lot in the middle, and some following late. These models assume that it takes a while for companies to find out about new technology and, once they do, for their employees to assimilate and use it. Robotics is a good example: It s obvious that it can increase productivity, but it takes some know-how to put robots to work.
Organizations develop processes through repeated problem solving. When a firm gets started, its founders solve problems for customers, and as they grow and establish routines, these ways of working come to embody their organizational capabilities. Managers constantly try to fit new market needs to existing processes and routines. Sometimes they are a fit, but often they are not. They can even require cultural change, which is difficult for established organizations.
U.S. automakers took decades to adopt lean production methods despite the obvious benefits from increased productivity and lower work-in-process inventory. Even General Motors, which had a bird s eye view of the Toyota Production System from its joint venture with Toyota at New United Motors Manufacturing Inc. (NUMMI) in California, did not deploy lean manufacturing across the company for decades despite the clear advantages. A former GM chairman told me that while his company learned much from NUMMI, implementing it across the GM manufacturing network was slow going.
Robert Abernathy and Kim Clark categorized technology shifts as competency-enhancing or competency-destroying. The latter are troublesome because the knowledge base and skills required to operate in the new realm are so fundamentally different. From the perspective of how the resources in an organization are deployed and how its processes are organized, this makes sense. The more fundamental the break from the previous technology, the greater the switching costs.
And then we have Clay Christensen s work on disruptive innovations: how a new technology that is initially not good enough for mainstream customers, initially gains a foothold with a set of low-end customers who are happy with a less expensive, good enough product. And then the innovators improve the technology steadily over time, allowing it to serve more and more of the overall market, which causes great trauma for incumbents who stuck to their legacy ways of doing things. One of Christensen s favorite examples is steel minimills, whose electric arc furnaces used scrap steel as their feedstock. While the quality of the steel that they initially produced was poor because of their inability to control the alloy contents due to contaminants in the scrap, technological advancements progressively eliminated this difficulty and increasingly made their lower-cost products viable substitutes for steel made from virgin materials.
But the legacy integrated steel makers in the United States were in trouble even before the low-end disruption from the minimills. They had fallen behind firms in Germany, Japan, and other countries in adopting new technologies for making steel from virgin ore and materials (e.g., the basic oxygen process and continuous casting). These weren t really that fundamental a technology change, but making the transition did require massive capital expenditures.
This is a great example of how, thanks to the financial tools that companies widely use, previous investments that have not been fully depreciated can get in the way of new investments. And once investments are fully depreciated, they have a habit of hanging on by virtue of marginal costing. One of my CFOs once told me that the return on new capital investment is long term and the success rate has a high variance, while the return on maintenance capital investments required to keep existing production facilities running is short term and has a low variance. Paralleling Jim March s classic analysis of exploration vs. exploitation, short term crowds out long term here as well.
Sometimes circumstances make it easy to walk away from old technology and established routines. World War II left the German and Japanese steel industries in ruins. So when they were rebuilding, it was easy for them to opt for basic oxygen furnaces, a technology developed in Europe in 1948 that reduced capital costs and process time, and improved labor productivity. Similarly, companies in countries like Taiwan, South Korea, and China that entered industries late have been able to grab the lead in many industries for a simple reason: They bought their capital equipment later and were able to leapfrog existing players technologically. For the same reasons, entrepreneurial start-ups have been quick to adopt cloud computing and new service-oriented architectures because they didn t have any installed base of legacy applications.
So what should established companies saddled with conventional equipment do when new technology appears?
The first thing they should do is distinguish between functional and economic obsolescence. Functional obsolescence occurs when an asset does not support competitive capability. Economic obsolescence means your depreciation schedule, which is set by local tax law, says the asset is at the end of its accounting life. The challenge is not to let residual book values of equipment get in the way of having the most competitive capabilities. Several U.S. commercial airlines and automakers failed to address this challenge along the way and ultimately had to resort to an extreme measure bankruptcy to deal with their obsolete assets. Clearly, this is not an exemplary approach.
In contrast, foreign airlines like Emirates and Singapore Air hold on to their fleets for much shorter times. Their younger average fleet age gives them a more competitive product.
Other alternatives are leasing or aggressively selling older equipment. Still another is moving the older equipment to less-demanding applications or to plants in emerging markets. One plastic products maker I visited did this with their conventional, hydraulic-drive injection-molding equipment when a new generation of electric-drive technology that allows for more precise control and better quality appeared. A major semiconductor firm I know cascades older equipment to less-demanding, trailing fabs.
But sometimes you just have to bite the bullet. When I was working in optical disk manufacturing, I took over management of a business unit that was just setting up a new factory. We had brand new equipment, much of it still in crates. What I didn t know at the time was that a European supplier had come out with a continuous-flow production system that had much lower capital costs. This reduced the entry barriers in the market, which was then flooded with product. I was faced with undertaking a massive restructuring and write-downs to get the unit back to a competitive position. Hard call? Our survival depended upon it.
Obviously, it is critical for established firms to track technological innovations in their industries and to understand their sources. How much of it is driven by your tool suppliers? How much of it comes from complementary materials processed by these tools or how you apply them? More importantly, managers need to monitor the tool innovation roadmap. Capital-equipment makers invariably have such a roadmap, and managers of companies that use their products need to make it their business to understand when new capabilities will arrive.
Managers also need to think differently about production systems. How can they be designed so they can be easily upgradable? That means building in a little flexibility on the plant floor and learning to adapt to variable production quantities.
Capturing the full benefit advanced tools requires making investments in complementary assets like the tooling (for example, molds and inserts for injection molding). In some areas like 3-D printing, innovation in the materials processed is just as important as the tools themselves. GE Aviation, both directly and through its acquisition of Morris Technologies, was an early explorer in laser sintering for 3-D printing of metals. It experimented extensively with sintering super alloys and is now moving into production on fuel injectors for its LEAP engine for the Airbus A320neo and Boeing 737MAX.
Finally, established companies should encourage experimentation with new tools and methods. This exploration is critical to the early formation of new capabilities. In his book From the American System to Mass Production, 1800-1932, business historian David Hounshell talks about how Henry Ford encouraged his production engineers to experiment extensively, even encouraging them to scrap processes and machine tools that no longer met their fancy. Ford s pioneering moving assembly line was a direct result. Managers need to strike the right balance between exploration and exploitation.
Keeping a firms asset base up to date and flexible is an important way to ensure that legacy doesn t turn into a liability in today s hypercompetitive environment.
Willy C. Shih is the Robert and Jane Cizik Professor of Management Practice in Business Administration at Harvard Business School.