A central boiler sent steam through pipes to small steam engines at each point where power was required in the machine. Through a system of
62 | WHY GREAT COMPANIES CAN FAIL
pulleys, drums, and cables, these engines manipulated frontward-scooping buckets, as illustrated in Figure 3 . 1 . Originally, steam shovels were mounted on rails and used to excavate earth in railway and canal construction. American excavator manufacturers were tightly clustered in northern Ohio and near Milwaukee.
In the early 1 9 2 0 s , when there were more than thirty-two steam shovel manufacturers based in the United States, the industry faced a major technological upheaval, as gasoline-powered engines were substituted for steam power.2 This transition to gasoline power falls into the category that Henderson and Clark label radical technological transition. The fundamental technological concept in a key component (the engine) changed from steam to internal combustion, and the basic architecture of the product changed. Where steam shovels used steam pressure to power a set of steam engines to extend and retract the cables that actuated their
Figure 3.1 Cable-Actuated Mechanical Shovel Manufactured by Osgood General
Source: O s g o o d G e n e r a l p h o t o in H e r b e r t L. N i c h o l s , Jr., Moving the Earth: The Workbook of
Excavation ( G r e e n w i c h , CT: N o r t h C a s t l e , 1 9 5 5 ) .
Change in the Mechanical Excavator Industry | 63
buckets, gasoline shovels used a single engine and a very different system of gearing, clutches, drums, and brakes to wind and unwind the cable.
Despite the radical nature of the technological change, however, gasoline technology had a sustaining impact on the mechanical excavator industry.
Gasoline engines were powerful enough to enable contractors to move earth faster, more reliably, and at lower cost than any but the very largest steam shovels.
The leading innovators in gasoline engine technology were the industry's dominant firms, such as Bucyrus, Thew, and Marion. Twenty-three of the twenty-five largest makers of steam shovels successfully negotiated the transition to gasoline power.3 As Figure 3.2 shows, there were a few entrant firms among the gasoline technology leaders in the 1 9 2 0 s , but the established firms dominated this transition.
Beginning in about 1 9 2 8 , the established manufacturers of gasoline-powered shovels initiated the next major, but less radical, sustaining technological transition—to shovels powered by diesel engines and electric motors. A further transition, made after World War II, introduced the
Figure 3.2 Manufacturers of Gasoline-Powered Cable Shovels, 1920-1934
ource: D a t a are f r o m t h e Historical C o n s t r u c t i o n E q u i p m e n t A s s o c i a t i o n a n d f r o m The Thomas
Register, v a r i o u s y e a r s .
64 I WHY GREAT COMPANIES CAN FAIL
arched boom design, which allowed longer reach, bigger buckets, and better down-reaching flexibility. The established firms continued to embrace and succeed with each of these innovations.
Excavation contractors themselves actually pioneered a number of other important sustaining innovations, first modifying their own equipment in the field to make it perform better and then manufacturing excavators incorporating those features to sell to the broader market.4
T H E I M P A C T O F D I S R U P T I V E H Y D R A U L I C S T E C H N O L O G Y
The next major technological change precipitated widespread failure in the industry. Beginning shortly after World War II and continuing through the late 1 9 6 0 s , while the dominant source of power remained the diesel engine, a new mechanism emerged for extending and lifting the bucket: hydraulically actuated systems replaced the cable-actuated systems. Only four of the thirty or so established manufacturers of cable-actuated equipment in business in the 1 9 5 0 s (Insley, Koehring, Little Giant, and Link Belt) had successfully transformed themselves into sustainable hydraulic excavator manufacturers by the 1 9 7 0 s . A few others survived by withdrawing into making such equipment as huge, cable-actuated draglines for strip mining and dredging.5 M o s t of the others failed. The firms that overran the excavation equipment industry at this point were all entrants into the hydraulics generation: J. I. Case, J o h n Deere, Drott, Ford, J. C.
Bamford, Poclain, International Harvester, Caterpillar, O 8c K, Demag, Leibherr, Komatsu, and Hitachi.6 Why did this happen?
Performance Demanded in the Mechanical Excavator Market Excavators are one of many types of earthmoving equipment. Some equipment, such as bulldozers, loaders, graders, and scrapers, essentially push, smooth, and lift earth. Excavators7 have been used to dig holes and trenches, primarily in three markets: first and largest, the general excavation market, composed of contractors who dig holes for basements or civil engineering projects such as canal construction; second, sewer and piping contractors, who generally dig long trenches; and third, open pit or strip mining. In each of these markets, contractors have tended to measure the functionality of mechanical excavators by their reach or extension distance and by the cubic yards of earth lifted in a single scoop.8
In 1 9 4 5 , sewer and piping contractors used machines whose bucket
Change in the Mechanical Excavator Industry | 65
capacity averaged about 1 cubic yard (best for digging relatively narrow trenches), while the average general excavation contractor used excavators that hefted 2Vz cubic yards per scoop and mining contractors used shovels holding about 5 cubic yards. T h e average bucket capacity used in each of these markets increased at about 4 percent per year, a rate of increase constrained by other factors in the broader system-of-use. The logistical problems of transporting large machines into and out of typical construction sites, for example, helped limit the rate of increase demanded by contractors.
The Emergence and Trajectory of Improvement
of Hydraulic Excavation
The first hydraulic excavator was developed by a British company, J. C.
Bamford, in 1 9 4 7 . Similar products then emerged simultaneously in several American companies in the late 1 9 4 0 s , among them, the Henry Company, of Topeka,_ Kansas, and Sherman Products, Inc., of Royal Oak, Michigan. T h e approach was labeled "Hydraulically Operated Power Take-Off," yielding an acronym that became the name of the third entrant to hydraulic excavating in the late 1 9 4 0 s , H O P T O . 9
Their machines were called backhoes because they were mounted on the back of industrial or farm tractors. Backhoes excavated by extending the shovel out, pushing it down into the earth,10 curling or articulating the shovel under the slice of earth, and lifting it up out of the hole. Limited by the power and strength of available hydraulic pumps' seals, the capacity of these early machines was a mere Vi cubic yard, as graphed in Figure 3 . 3 . Their reach was also limited to about six feet. "Whereas the best cable excavators could rotate a full 3 6 0 degrees on their track base, the most flexible backhoes could rotate only 1 8 0 degrees.
Because their capacity was so small and their reach so short, hydraulic excavators were of no use to mining, general excavation, or sewer contractors, who were demanding machines with buckets that held 1 to 4 cubic yards. As a result, the entrant firms had to develop a new application for their products. They began to sell their excavators as attachments for the back of small industrial and farm tractors made by Ford, J. I. Case, J o h n Deere, International Harvester, and Massey Ferguson. Small residential contractors purchased these units to dig narrow ditches from water and sewer lines in the street to the foundations of houses under construction.
These very small jobs had never warranted the expense or time required
6 6 | W H Y G R E A T C O M P A N I E S C A N F A I L
Figure 3.3 Disruptive Impact of Hydraulics Technology in the Mechanical Excavator Market
Source: D a t a are f r o m t h e Historical C o n s t r u c t i o n E q u i p m e n t A s s o c i a t i o n .
to bring in a big, imprecise, cable-actuated, track-driven shovel, so the trenches had always been dug by hand. Hydraulic backhoes attached to highly mobile tractors could do these jobs in less than an hour per house, and they became extremely popular with contractors building large tract subdivisions during the housing booms that followed World War II and the Korean War. These early backhoes were sold through tractor and implement dealerships accustomed to dealing with small customers.
The early users of hydraulic excavators were, in a word, very different from the mainstream customers of the cable shovel manufacturers—in size, in needs, and in the distribution channels through which they bought.
They constituted a new value network for mechanical excavation. Interestingly, just as the performance of smaller-architecture disk drives was measured in different metrics than the performance of large drives (weight, ruggedness, and power consumption versus capacity and speed), the performance of the first backhoes was measured differently from the perfor-
Change in the Mechanical Excavator Industry | 67
mance of cable-actuated equipment. The metrics featured most prominently in early product literature of hydraulic backhoes were shovel width (contractors wanted to dig narrow, shallow trenches) and the speed and maneuverability of the tractor. Figure 3 . 4 , excerpted from an early product brochure from Sherman Products for its "Bobcat" hydraulic backhoe, illustrates this. Sherman called its Bobcat a "digger," showed it operating in tight quarters, and claimed it could travel over sod with minimum damage. The Bobcat was mounted on a Ford tractor. (Ford subsequently acquired the Sherman Bobcat line.) The featured attributes,
Figure 3.4 Hydraulic Backhoe Manufactured by Sherman Products
Source: B r o c h u r e f r o m S h e r m a n P r o d u c t s , Inc., R o y a l O a k , M i c h i g a n , early 1 9 5 0 s .
68 | WHY GREAT COMPANIES CAN FAIL
of course, were simply irrelevant to contractors whose bread was buttered by big earthmoving projects. These differences in the rank-ordering of performance attributes defined the boundaries of the industry's value networks.
The solid line in Figure 3.3 charts the rate of improvement in bucket size that hydraulics engineers were able to provide in the new excavator architecture. T h e maximum available bucket size had reached 3/8 cubic yard by 1 9 5 5 , 1/2 cubic yard by 1 9 6 0 , and 2 cubic yards by 1 9 6 5 . By 1 9 7 4 , the largest hydraulic excavators had the muscle to lift 10 cubic yards. This trajectory of improvement, which was far more rapid than the rate of improvement demanded in any of the excavator markets, carried this disruptive hydraulics technology upward from its original market through the large, mainstream excavation markets. The use of hydraulic excavators in general contracting markets was given a boost in 1 9 5 4 when another entrant firm in Germany, Demag, introduced a track-mounted model that could rotate on its base a full 3 6 0 degrees.
T H E R E S P O N S E T O H Y D R A U L I C S B Y T H E E S T A B L I S H E D
E X C A V A T O R M A N U F A C T U R E R S
Just as Seagate Technology was one of the first firms to develop prototype 3.5-inch drives, Bucyrus Erie, the leading cable shovel maker, was keenly aware of the emergence of hydraulic excavating technology. By 1 9 5 0
(about two years after the first backhoe appeared) Bucyrus purchased a fledgling hydraulic backhoe company, the Milwaukee Hydraulics Corporation. Bucyrus faced precisely the same problem in marketing its hydraulic backhoe as Seagate had faced with its 3.5-inch drives: Its most powerful mainstream customers had no use for it.
Bucyrus Erie's response was a new product, introduced in 1 9 5 1 , called the "Hydrohoe." Instead of using three hydraulic cylinders, it used only two, one to curl the shovel into the earth and one to "crowd" or draw the shovel toward the cab; it used a cable mechanism to lift the shovel.
T h e Hydrohoe was thus a hybrid of the two technologies, reminiscent of the early transoceanic steamships outfitted with sails.11 There is no evidence, however, that the Hydrohoe's hybrid design resulted from Bucyrus engineers' being "stuck" in some sort of cable-based engineering paradigm.
Rather, the cable lift mechanism was the only viable way at that time, based on the state of hydraulics technology, to give the Hydrohoe the
Change in the Mechanical Excavator Industry | 69
bucket capacity and reach that Bucyrus marketers thought they needed to appeal to their existing customers' needs.
Figure 3.5 presents an excerpt from an early Hydrohoe product brochure. Note the differences from Sherman's marketing approach: Bucyrus labeled the Hydrohoe a "dragshovel," showed it in an open field, and claimed it could "get a heaping load on every pass"—all intended to
Figure 3.5 Hydrohoe Manufactured by Bucyrus Erie
Source: B r o c h u r e f r o m B u c y r u s Erie C o m p a n y , S o u t h M i l w a u k e e , W i s c o n s i n , 1 9 5 1 .
70 ] WHY GREAT COMPANIES CAN FAIL
appeal to general excavation contractors. Rather than commercialize the disruptive technology in the value network in which the current attributes of hydraulics were prized, Bucyrus tried to adapt the technology to fit its own value network. Despite this attempt, the Hydrohoe was still too limited in capacity and reach and did not sell well to Bucyrus' customers.
Bucyrus kept its Hydrohoe on the market for over a decade, attempting periodically to upgrade its performance to make it acceptable to its customers, but the machine was never commercially successful. Ultimately, the company returned to the cable shovels that its customers needed.
Bucyrus Erie was the only maker of cable-actuated shovels known to have launched a hydraulic excavator between 1 9 4 8 and 1 9 6 1 : All of the other manufacturers continued serving their established customers, well and prosperously.12 In fact, the largest makers of cable-actuated excavators, Bucyrus Erie and Northwest Engineering, logged record profits until 1 9 6 6 — t h e point at which the disruptive hydraulics technology had squarely intersected with customers' needs in the sewer and piping segment. This is typical of industries facing a disruptive technology: The leading firms in the established technology remain financially strong until the disruptive technology is, in fact, in the midst of their mainstream market.
Between 1 9 4 7 and 1 9 6 5 , twenty-three companies entered the mechanical excavation market with hydraulic products. Figure 3 . 6 , which measures the total number of active entrants and established firms offering hydraulic excavators (net of the companies that had exited), shows how completely the entrants dominated the hydraulic excavator market.
In the 1 9 6 0 s , some of the strongest cable shovel makers introduced shovels with hydraulics. Almost all of these models were hybrids, however, like Bucyrus Erie's Hydrohoe, generally employing a hydraulic cylinder to articulate or curl the bucket and using cables to extend the bucket out and to lift the boom. When used in this way in the 1 9 6 0 s , hydraulics had a sustaining impact on the established manufacturers' products, improving their performance in the mainstream value networks. Some of the methods that engineers found to use hydraulics on the cable excavators were truly ingenious. All of this innovative energy, however, was targeted at existing customers.
The strategies employed by the excavator manufacturers during this period highlight an important choice that confronts companies encountering disruptive technological change. In general, the successful entrants accepted the capabilities of hydraulics technology in the 1 9 4 0 s and 1 9 5 0 s
Change in the Mechanical Excavator Industry | 71
Figure 3.6 Manufacturers of Hydraulic Excavators, 1948-1965
Source: D a t a are f r o m t h e Historical C o n s t r u c t i o n E q u i p m e n t A s s o c i a t i o n .
as a given and cultivated new market applications in which the technology, as it existed, could create value. And as a general rule, the established firms saw the situation the other way around: They took the market's needs as the given. They consequently sought to adapt or improve the technology in ways that would allow them to market the new technology to their existing customers as a sustaining improvement. The established firms steadfastly focused their innovative investments on their customers.
Subsequent chapters will show that this strategic choice is present in most instances of disruptive innovation. Consistently, established firms attempt to push the technology into their established markets, while the successful entrants find a new market that values the technology.
Hydraulics technology ultimately did progress to the point where it could address the needs of mainstream excavation contractors. That progress was achieved, however, by the entrant companies, who had first found a market for the initial capabilities of the technology, accumulated design and manufacturing experience in that market, and then used that commercial platform to attack the value networks above them. The established firms lost this contest. Only four cable excavator companies—Insley,
72 | WHY GREAT COMPANIES CAN FAIL
Koehring, Little Giant, and Link Belt—remained as viable suppliers to excavation contractors by successfully but belatedly introducing lines of hydraulic excavators to defend their markets.13
Aside from these, however, the other leading manufacturers of big cable machines in the mainstream excavation markets never introduced a commercially successful hydraulic excavator. Although some had employed hydraulics to a modest degree as a bucket-curling mechanism, they lacked the design expertise and volume-based manufacturing cost position to compete as hydraulics invaded the mainstream. By the early 1 9 7 0 s , all of these firms had been driven from the sewer, piping, and general excavation markets by the entrants, most of which had refined their technological capabilities initially in the small-contractor market.14
This contrast in strategies for profiting from change characterizes the approaches employed by entrant and established firms in many of the other industries affected by disruptive technologies—particularly disk drives, steel, computers, and electric cars.
T H E C H O I C E B E T W E E N C A B L E A N D H Y D R A U L I C S
In the trajectory map of Figure 3 . 3 , when hydraulics technology became capable of addressing the bucket-size needs of sewer and piping contractors (and a similar trajectory could be sketched for arm-reach), the competitive dynamics in the industry changed, and the mainstream excavation contractors changed the criteria by which they purchased their equipment.
Even today, the cable-actuated architecture can attain much longer reach and greater lift than can hydraulic excavators: They have roughly parallel technology trajectories. But once both cable- and hydraulics-actuated systems could satisfy mainstream market requirements, excavation contractors could no longer base their choice of equipment on which had longer reach and greater bucket capacity. Both were good enough, and the fact that cable was better ceased to have competitive relevance.
Contractors found, however, that hydraulic machines were much less prone to breakdowns than cable-actuated excavators. In particular, those who had experienced the life-threatening snap of a cable while hefting a heavy bucket embraced reliable hydraulics quickly, as soon as it was capable of doing the j o b . Once both technologies were good enough in the basic capabilities demanded, therefore, the basis of product choice in the market shifted to reliability. Sewer and piping contractors began
Change in the Mechanical Excavator Industry | 73
adopting hydraulic equipment rapidly beginning in the early 1 9 6 0 s , and general excavation contractors followed later in the decade.
C O N S E Q U E N C E S A N D I M P L I C A T I O N S O F T H E
H Y D R A U L I C S E R U P T I O N
W h a t went wrong within the companies that made cable-actuated excavators? Clearly, with the benefit of hindsight, they should have invested in hydraulics machines and embedded that piece of their organizations charged with making hydraulic products in the value network that needed them. But the dilemma in managing the disruptive technology in the heat of the battle is that nothing went wrong inside these companies. Hydraulics was a technology that their customers didn't need—indeed, couldn't use.
Each cable shovel manufacturer was one of at least twenty manufacturers doing everything they could to steal each other's customers: If they took their eyes off their customers' next-generation needs, existing business would have been put at risk. Moreover, developing bigger, better, and faster cable excavators to steal share from existing competitors constituted a much more obvious opportunity for profitable growth than did a venture into hydraulic backhoes, given how small the backhoe market was when it appeared in the 1 9 5 0 s . So, as we have seen before, these companies did not fail because the technology wasn't available. They did not fail because they lacked information about hydraulics or how to use it; indeed, the best of them used it as soon as it could help their customers. They did not fail because management was sleepy or arrogant. They failed because hydraulics didn't make sense—until it was too late.
The patterns of success and failure we see among firms faced with sustaining and disruptive technology change are a natural or systematic result of good managerial decisions. That is, in fact, why disruptive technologies confront innovators with such a dilemma. Working harder, being smarter, investing more aggressively, and listening more astutely to customers are all solutions to the problems posed by new sustaining technologies. But these paradigms of sound management are useless—even counterproductive, in many instances—when dealing with disruptive technology.
N O T E S
1. A summary of how this same mechanism might have affected a broader range of industries can be found in Richard S. Rosenbloom and Clayton
74 I WHY GREAT COMPANIES CAN FAIL
M. Christensen, "Technological Discontinuities, Organizational Capabilities, and Strategic Commitments," Industrial and Corporate Change (3), 1994, 655-686.
2. This information and the data used to calculate the graphs in this section were provided by Dimitrie Toth, Jr., and Keith Haddock, both National Directors of the Historical Construction Equipment Association. The associa-tion has a wealth of information about the earthmoving equipment industry in its archives, and Toth and Haddock were most gracious in sharing their knowledge and information with me. I am also indebted to them for their helpful comments on an earlier draft of this chapter. Other useful sources of information are Peter Grimshaw, Excavators (Poole, England: Blandford Press, 1985); The Olyslager Organisation, Inc., Earthmoving Vehicles (London: Frederick Warne & Co., Ltd., 1972); Harold F. Williamson and Kenneth H. Myers, Designed for Digging: The First 75 Years of Bucyrus Erie Company (Evanston, IL: Northwestern University Press, 1955); and J. L. Allhands, Tools of the Earthmover (Huntsville, TX: Sam Houston College Press, 1951).
3. Interestingly, the high success rate was only amongst the industry's twenty-five largest firms. Only one of the seven smallest steam shovel manufacturers survived this sustaining technology change to internal gasoline combustion.
Almost no information is available about these companies other than what is provided by their product brochures. I suspect, however, that the fact that the large and mid-sized firms cruised through this transition while the small ones were killed indicates that resources played a part in the story, a conclusion that complements the theoretical perspectives summarized in chapter 2
above. Some sustaining technologies clearly are so expensive to develop and implement or so dependent on proprietary or scarce expertise that some companies simply cannot successfully manage the transition. I am indebted to Professor Richard Rosenbloom for sharing his perspective on this issue.
4. An example of this is the development of the first dragline, by Page, a Chicago area contractor. Page dug Chicago's system of canals, and invented the dragline in 1903 to do that job more effectively. Page draglines were later used extensively in digging the Panama Canal, alongside steam shovels made by Bucyrus Erie and Marion. This finding that customers were significant sources of sustaining innovations is consistent with Professor Eric von Hippel's findings; see The Sources of Innovation (New York: Oxford University Press, 1988).
5. The companies that survived the invasion of hydraulics in this way found safe haven in a particular high-end market. Bucyrus Erie and Marion, for example, became the dominant makers of the huge stripping shovels used in strip mines. Marion's model 6360 stripping shovel was the largest frontward-scooping shovel ever built, able to heft 180 cubic yards in its bucket. (An
Change in the Mechanical Excavator Industry \ 75
advertisement showing Paul Bunyan standing aside the 6360 is one of the most stunning pieces of advertising art I have seen.) Harnischfeger is the world's largest maker of electric mining shovels, while Unit found a niche making the huge pedestal cranes used on offshore oil rigs. For a time, Northwest survived by making draglines for dredging ocean shipping lanes.
P & H and Lorain made huge cranes and draglines (all cable-actuated).
6. As the hydraulic excavator has matured, these companies have met with varying degrees of subsequent success. In 1996, the world's highest-volume excavator companies, Demag and O & K, were based in Germany.
7. Technically, excavators that scoop their buckets forward are power shovels.
This was the dominant design from 1837 through the early 1900s, and persisted as a major market segment through much of this century. Excavators that pull earth backward toward the cab are backhoes. As the hydraulic excavator became the dominant design during the 1970s, both types came to be called excavators. Until hydraulic actuation required the booms to be permanently attached to the unit, contractors could attach different booms or arms to their basic power units so that the same unit could work as a shovel, backhoe, or crane. Similarly, different buckets, sometimes called dip-pers, could be attached to move different types of material.
8. The true measure of performance in excavation was the number of cubic yards of earth that could be moved per minute. This measure was so dependent upon operator skill and upon the type of earth being dug, however, that contractors adopted bucket size as the more robust, verifiable metric.
9. These British and American pioneers were followed by several European manufacturers, each of which was also an entrant to the excavator industry, including France's Poclain and Italy's Bruneri Brothers.
10. The ability to push the shovel into the earth was a major advantage to the hydraulics approach. The cable-actuated excavators that pulled earth toward the operator all had to rely on gravity to drive the teeth of the heavy shovel into the earth.
11. Makers of early hybrid ocean transports, which were steam-powered but still outfitted with sails, used the same rationale for their design as did the Bucyrus Erie engineers: Steam power still was not reliable enough for the transoceanic market, so steam power plants had to be backed up by conventional technology. The advent of steam-powered ships and their substitution for wind-powered ships in the transoceanic business is itself a classic study of disruptive technology. When Robert Fulton sailed the first steamship up the Hudson River in 1819, it underperformed transoceanic sailing ships on nearly every dimension of performance: It cost more per mile to operate; it was slower; and it was prone to frequent breakdowns. Hence, it could not be used in the transoceanic value network and could only be applied in a
76 WHY GREAT COMPANIES CAN FAIL
different value network, inland waterways, in which product performance was measured very differently. In rivers and lakes, the ability to move against the wind or in the absence of a wind was the attribute most highly valued by ship captains, and along that dimension, steam outperformed sail. Some scholars (see, for example, Richard Foster, in Innovation: The Attacker's Advantage [New York: Summit Books, 1986]) have marveled at how myopic were the makers of sailing ships, who stayed with their aging technology until the bitter end, in the early 1900s, completely ignoring steam power.
Indeed, not a single maker of sailing ships survived the industry's transition to steam power. The value network framework offers a perspective on this problem that these scholars seem to have ignored, however. It was not a problem of knowing about steam power or of having access to technology.
The problem was that the customers of the sailing ship manufacturers, who were transoceanic shippers, could not use steam-powered ships until the turn of the century. To cultivate a position in steamship building, the makers of sailing ships would have had to engineer a major strategic reorientation into the inland waterway market, because that was the only value network where steam-powered vessels were valued throughout most of the 1800s. Hence, it was these firms' reluctance or inability to change strategy, rather than their inability to change technology, that lay at the root of their failure in the face of steam-powered vessels.
12. An exception to this is an unusual product introduced by Koehring in 1957: the Skooper combined cables and hydraulics to dig earth away from a facing wall; it did not dig down into the earth.
13. Bucyrus Erie does not fit easily into either of these groups. It introduced a large hydraulic excavator in the 1950s, but subsequently withdrew it from the market. In the late 1960s, it acquired the "Dynahoe" line of hydraulic loader-backhoes from Hy-Dynamic Corporation and sold them as utility machines to its general excavation customers, but, again, dropped this product line as well.
14. Caterpillar was a very late (but successful) entrant into the hydraulic excavation equipment industry, introducing its first model in 1972. Excavators were an extension of its line of dozers, scrapers, and graders. Caterpillar never participated in the excavation machine market when cable actuation was the dominant design.
C H A P T E R F O U R
What Goes Up, Can't Go Down
It is clear from the histories of the disk drive and excavator industries that the boundaries of value networks do not completely imprison the companies within them: There is considerable upward mobility into other networks. It is in restraining downward mobility into the markets enabled by disruptive technologies that the value networks exercise such unusual power. In this chapter we will explore these questions: Why could leading companies migrate so readily toward high-end markets, and why does moving downmarket appear to have been so difficult? Rational managers, as we shall see, can rarely build a cogent case for entering small, poorly defined low-end markets that offer only lower profitability. In fact, the prospects for growth and improved profitability in upmarket value networks often appear to be so much more attractive than the prospect of staying within the current value network, that it is not unusual to see well-managed companies leaving (or becoming uncompetitive with) their original customers as they search for customers at higher price points. In good companies, resources and energy coalesce most readily behind proposals to attack upmarket into higher-performance products that can earn higher margins.
Indeed, the prospects for improving financial performance by moving toward upmarket value networks are so strong that one senses a huge
7 8
W H Y G R E A T C O M P A N I E S C A N F A I L
magnet in the northeast corner of the disk drive and excavator trajectory maps. This chapter examines the power of this "northeastern pull" by looking at evidence from the history of the disk drive industry. It then generalizes this framework by exploring the same phenomenon in the battle between minimill and integrated steel makers.
T H E G R E A T N O R T H E A S T M I G R A T I O N I N D I S K D R I V E S
Figure 4 . 1 plots in more detail the upmarket movement of Seagate Technology, whose strategy was typical of most disk drive manufacturers. Recall that Seagate had spawned, and then grew to dominate, the value network for desktop computing. Its product position relative to capacity demanded in its market is mapped by vertical lines which span from the lowest- to the highest-capacity drives in its product line, in each of the years shown.
The black rectangle on the line measuring each year's capacity span shows the median capacity of the drives Seagate introduced in each of those years.
Figure 4.1 Upmarket Migration of Seagate Products
Source: D a t a a r e f r o m v a r i o u s i s s u e s of Disk/Trend Report.
• What Goes Up, Can't Go Down | 79
Between 1 9 8 3 and 1 9 8 5 , the center of gravity of Seagate's product line was positioned squarely on the average capacity demanded in the desktop segment. It was between 1 9 8 7 and 1 9 8 9 that the disruptive 3.5-inch form invaded the desktop market from below. Seagate responded to that attack, not by fighting the disruptive technology head-on, but by retreating upmarket. It continued to offer models in the capacity ranges the desktop PC market demanded, but by 1 9 9 3 the focus of its energy had clearly shifted to the market for mid-range computers, such as file servers and engineering workstations.
Indeed, disruptive technologies have such a devastating impact because the firms that first commercialized each generation of disruptive disk drives chose not to remain contained within their initial value network.
Rather, they reached as far upmarket as they could in each new product generation, until their drives packed the capacity to appeal to the value networks above them. It is this upward mobility that makes disruptive technologies so dangerous to established firms—and so attractive to entrants.
V A L U E N E T W O R K S A N D C H A R A C T E R I S T I C C O S T
S T R U C T U R E S
W h a t lies behind this asymmetric mobility? As we have already seen, it is driven by resource allocation processes that direct resources toward new product proposals that promise higher margins and larger markets.
These are almost always better in the northeast portions of trajectory maps (such as Figures 1.7 and 3.3) than in the southeast. The disk drive manufacturers migrated to the northeast corner of the product-market map because the resource allocation processes they employed took them there.
As we saw in chapter 2, a characteristic of each value network is a particular cost structure that firms within it must create if they are to provide the products and services in the priority their customers demand.
Thus, as the disk drive makers became large and successful within their
"home" value network, they developed a very specific economic character: tuning their levels of effort and expenses in research, development, sales, marketing, and administration to the needs of their customers and the challenges of their competitors. Gross margins tended to evolve in each value network to levels that enabled the better disk drive makers to make money, given these costs of doing business.
80 WHY GREAT COMPANIES CAN FAIL
In turn, this gave these companies a very specific model for improving profitability. Generally, they found it difficult to improve profitability by hacking out cost while steadfastly standing in their mainstream market: The research, development, marketing, and administrative costs they were incurring were all critical to remaining competitive in their mainstream business. Moving upmarket toward higher-performance products that promised higher gross margins was usually a more straightforward path to profit improvement. Moving downmarket was anathema to that objective.
The obviousness of the path toward profit improvement is shown in Figure 4 . 2 . The three bars on the left depict the size of the desktop, minicomputer, and mainframe computer value networks in 1 9 8 1 and are labeled with the characteristic margins enjoyed by disk drive makers in each of those networks. Gross margins are clearly higher in higher-end markets, compensating manufacturers for the higher levels of overhead characteristic of those businesses.
The differences in the size of these markets and the characteristic cost structures across these value networks created serious asymmetries in the combat among these firms. Firms making 8-inch drives for the minicomputer market, for example, had cost structures requiring gross margins of 40 percent. Aggressively moving downmarket would have pitted them against foes who had honed their cost structures to make money at 25
percent gross margins. On the other hand, moving upmarket enabled them to take a relatively lower-cost structure into a market that was accustomed to giving its suppliers 60 percent gross margins. Which direction made sense? A similar asymmetry faced the makers of 5.25-inch drives in 1 9 8 6 , as they decided whether to spend their resources building a position in the emerging market for 3.5-inch drives in portable computers or to move up toward the minicomputer and mainframe companies.
Committing development resources to launch higher-performance products that could garner higher gross margins generally both offered greater returns and caused less pain. As their managers were making repeated decisions about which new product development proposals they should fund and which they should shelve, proposals to develop higher-performance products targeted at the larger, higher-margin markets immediately above them always got the resources. In other words, sensible resource allocation processes were at the root of companies' upward mobility and downmarket immobility across the boundaries of the value networks in the disk drive industry.
The hedonic regression analysis summarized in chapter 2 showed that
What Goes Up, Can't Go Down | 81
Figure 4.2 Views Upmarket and Downmarket for Established Disk Drive Manufacturers
Market view of 8-inch drive
Market view of 5.25-inch
makers in 1981, when the
drive makers in 1986,
5.25-inch desktop
when the 3.5-inch portable
computer value network
computer value network
was just emerging.
was just emerging.
Source: D a t a a r e f r o m v a r i o u s i s s u e s of Disk/Trend Report, c o r p o r a t e a n n u a l r e p o r t s , a n d d a t a p r o v i d e d in p e r s o n a l i n t e r v i e w s .
Note: P e r c e n t a g e s a b o v e e a c h bar indicate t y p i c a l g r o s s m a r g i n s in e a c h v a l u e n e t w o r k .
higher-end markets consistently paid significantly higher prices for incremental megabytes of capacity. Why would anyone opt to sell a megabyte for less when it could be sold for more? The disk drive companies' migration to the northeast was, as such, highly rational.
Other scholars have found evidence in other industries that as companies leave their disruptive roots in search of greater profitability in the market tiers above them, they gradually come to acquire the cost structures required to compete in those upper market tiers.1 This exacerbates their problem of downward immobility.
82 | WHY GREAT COMPANIES CAN FAIL
R E S O U R C E A L L O C A T I O N A N D U P W A R D M I G R A T I O N
Further insight into this asymmetric mobility across value networks comes from comparing two different descriptive models of how resources are allocated. The first model describes resource allocation as a rational, top-down decision-making process in which senior managers weigh alternative proposals for investment in innovation and put money into those projects that they find to be consistent with firm strategy and to offer the highest return on investment. Proposals that don't clear these hurdles are killed.
T h e second model of resource allocation, first articulated by Joseph Bower,2 characterizes resource allocation decisions much differently.
Bower notes that most proposals to innovate are generated from deep within the organization not from the top. As these ideas bubble up from the bottom, the organization's middle managers play a critical but invisible role in screening these projects. These managers can't package and throw their weight behind every idea that passes by; they need to decide which are the best, which are most likely to succeed, and which are most likely to be approved, given the corporate financial, competitive, and strategic climate.
In most organizations, managers' careers receive a big boost when they play a key sponsorship role in very successful projects—and their careers can be permanently derailed if they have the bad judgment or misfortune to back projects that fail. Middle managers aren't penalized for all failures, of course. Projects that fail because the technologists couldn't deliver, for example, often are not (necessarily) regarded as failures at all, because a lot is learned from the effort and because technology development is generally regarded as an unpredictable, probabilistic endeavor. But projects that fail because the market wasn't there have far more serious implications for managers' careers. These tend to be much more expensive and public failures. They generally occur after the company has made full investments in product design, manufacturing, engineering, marketing, and distribution. Hence, middle managers—acting in both their own and the company's interest—tend to back those projects for which market demand seems most assured. They then work to package the proposals for their chosen projects in ways geared to win senior management approval. As such, while senior managers may think they're making the resource allocation decisions, many of the really critical resource allocation decisions have actually been made long before senior management gets involved: Middle managers have made their decisions about which proj-
•What Goes Up, Can't Go Down | 83
ects they'll back and carry to senior management—and which they will allow to languish.
Consider the implications of this for a successful firm's downward and upward mobility from its initial value network in this hypothetical example. In the same week, two respected employees, one from marketing, the other from engineering, run two very different ideas for new products past their common manager two levels above them in the organization.
The marketer comes first, with an idea for a higher-capacity, higher-speed model. The two-levels-up manager starts her interrogation:
"Who's going to buy it?"
"Well, there's a whole segment in the workstation industry—they buy over $600 million in drives each year—that we've just never been able to reach because our capacity points just don't reach that high. I think this product just might get us there."
"Have you run this idea past any potential customers?"
"Yeah, I was in California last week. They all said they wanted prototypes as soon as they could get them. There's a design window opening up in nine months. They've been working with their current supplier [competitor X]
to get something ready, but someone we just hired from competitor X said they're having lots of trouble meeting the specs. I really think we can do it."
"But does engineering think we can do it?"
"They say it'll be a stretch, but you know them. They always say that."
"What kind of margins are we looking at up there?"
"That's what really excites me about this. If we can build it in our current factory, given the price per megabyte competitor X has been getting, I think we can get close to 35 percent."
Compare that conversation to the manager's interchange with the engineer whose idea is for a cheaper, smaller, slower, lower-capacity disruptive disk drive.
"Who's going to buy it?"
"Well, I'm not sure, but there's got to be a market out there somewhere for it. People are always wanting things smaller and less expensive. I could see them using it in fax machines, printers, maybe."
"Have you run this idea past any potential customers?"
"Yeah, when I was at the last trade show I sketched the idea out for one of our current customers. He said he was interested, but couldn't see how they could really use it. Today you really need 270 MB to run everything, and there's just no way we could get that kind of capacity on this thing—at least not for a while. His response doesn't surprise me, really."
84 | WHY GREAT COMPANIES CAN FAIL
"How about the guys who make fax machines? What do they think?"
"Well, they say they don't know. Again, it's an intriguing idea, but they already have their product plans pretty well set, and none of them use disk drives."
"You think we could make money on this project?"
"Well, I think so, but that depends on how we could price it, of course."
Which of the two projects will the two-levels-up manager back? In the tug-of-war for development resources, projects targeted at the explicit needs of current customers or at the needs of existing users that a supplier has not yet been able to reach will always win over proposals to develop products for markets that do not exist. This is because, in fact, the best resource allocation systems are designed precisely to weed out ideas that are unlikely to find large, profitable, receptive markets. Any company that doesn't have a systematic way of targeting its development resources toward customers' needs, in fact, will fail.3
The most vexing managerial aspect of this problem of asymmetry, where the easiest path to growth and profit is up, and the most deadly attacks come from below, is that "good" management—working harder and smarter and being more visionary—doesn't solve the problem. The resource allocation process involves thousands of decisions, some subtle and some explicit, made every day by hundreds of people, about how their time and the company's money ought to be spent. Even when a senior manager decides to pursue a disruptive technology, the people in the organization are likely to ignore it or, at best, cooperate reluctantly if it doesn't fit their model of what it takes to succeed as an organization and as individuals within an organization. Well-run companies are not populated by yes-people who have been taught to carry out mindlessly the directives of management. Rather, their employees have been trained to understand what is good for the company and what it takes to build a successful career within the company. Employees of great companies exercise initiative to serve customers and meet budgeted sales and profits.
It is very difficult for a manager to motivate competent people to energeti-cally and persistently pursue a course of action that they think makes no sense. An example from the history of the disk drive industry illustrates the impact of such employee behavior.
T H E C A S E O F T H E 1 . 8 - I N C H D I S K D R I V E
Managers in disk drive companies were very generous in helping me conduct the research reported in this book, and, as the results began
•What Goes Up, Can't Go Down | 85
emerging in 1 9 9 2 , 1 began feeding back the published papers that summarized what I was learning. I was particularly interested in whether the framework summarized in Figure 1.7 would have an impact on their decisions regarding the 1.8-inch drive, which was just then emerging as the industry's most recent disruptive technology. For industry outsiders, of course, the conclusion was obvious: "How many times does this have to happen before these guys learn?! Of course they've got to do it." T h e guys did, in fact, learn. By the end of 1 9 9 3 , each of the leading drive makers had developed 1.8-inch models and had them ready for introduction if and when the market developed.
In August 1 9 9 4 , 1 was visiting the C E O of one of the largest disk drive companies and asked him what his firm was doing about the 1.8-inch drive. This clearly touched a hot button. He pointed to a shelf in his office where a sample 1.8-inch drive was perched. "You see that?" he demanded.
"That's the fourth generation of 1.8-inch drives we've developed—each one with more capacity than the last. But we haven't sold any. We want to be ready when the market is there, but there just isn't a market for them yet."
I countered by reminding him that Disk/Trend Report, a highly regarded market research publication that was the source of much of the data used in my study, had measured the 1 9 9 3 market at $ 4 0 million, was projecting 1 9 9 4 sales to be $ 8 0 million, and forecast 1 9 9 5 volume at $ 1 4 0 million.
"I know that's what they think," he responded. "But they're wrong.
There isn't a market. We've had that drive in our catalog for 18 months.
Everyone knows we've got it, but nobody wants it. The market just isn't there. We just got way ahead of the market." I had no other basis for pressing my point with this manager, who is one of the most astute managers I've ever met. Our conversation moved to other issues.
About a month later I was leading a case discussion in the Harvard M B A program's technology and operations management course about the development of a new engine at Honda. One of the students in the class had previously worked in Honda's research and development organization, so I asked him to take a few minutes to tell the class what it was like working there. It turned out that he had been working on dashboard mapping and navigation systems. I couldn't resist interrupting his talk by asking, "How do you store all that data for the maps?"
Said the student: "We found a little 1.8-inch disk drive and put it in there. It's really neat—almost a solid-state device, with very few moving parts. Really rugged."
"Who do you buy them from?" I pressed.
86 | WHY GREAT COMPANIES CAN FAIL
"It's kind of funny," he replied. "You can't buy them from any of the big disk drive companies. We get them from a little startup company somewhere in Colorado—I can't remember the name."
I have since reflected on why the head of this company would insist so stubbornly that there was no market for 1.8-inch drives, even while there was, and why my student would say the big drive makers didn't sell these drives, even though they were trying. The answer lies in the northeast-southeast problem, and in the role that the hundreds of well-trained decision makers in a good company play in funneling resources and energy into those projects they perceive will bring the company the greatest growth and profit. The C E O had decided that the company was going to catch this next disruptive wave early and had shepherded the project through to a successful, economical design. But among the employees, there was nothing about an $ 8 0 million, low-end market that solved the growth and profit problems of a multibillion dollar company—especially when capable competitors were doing all they could to steal away the customers providing those billions. (The revenue figure is disguised.) And way at the other end of the company, there was nothing about supplying prototype quantities of 1.8-inch drives to an automaker that solved the problem of meeting the 1 9 9 4 quotas of salespeople whose contacts and expertise were based so solidly in the computer industry.
For an organization to accomplish a task as complex as launching a new product, logic, energy, and impetus must all coalesce behind the effort. Hence, it is not just the customers of an established firm that hold it captive to their needs. Established firms are also captive to the financial structure and organizational culture inherent in the value network in which they compete—a captivity that can block any rationale for timely investment in the next wave of disruptive technology.
V A L U E N E T W O R K S A N D M A R K E T V I S I B I L I T Y
The impetus to drift upmarket can be particularly powerful when a firm's customers themselves are migrating upmarket. In such circumstances, suppliers of an intermediate component such as a disk drive may not sense their northeasterly migration because they are embedded among competitors and customers experiencing a similar drift.
In this light, we can see how easy it would have been for the leading 8-inch disk drive makers—Priam, Quantum, and Shugart—to miss the 5.25-inch generation of drives. Not a single one of their core customers,
. What Goes Up, Can't Go Down | 87
for example, Digital Equipment, Prime Computer, Data General, Wang Laboratories, and Nixdorf, successfully introduced a desktop computer.
Instead, each was moving upmarket itself toward ever higher performance segments of their markets, trying to win the business of customers who historically had used mainframes. Similarly, not a single one of the customers of the 14-inch drive makers—mainframe makers such as Univac, Burroughs, N C R , I C L , Siemens, and Amdahl—ever made a bold enough move downmarket into minicomputers to became a significant player there.
Three factors—the promise of upmarket margins, the simultaneous upmarket movement of many of a company's customers, and the difficulty of cutting costs to move downmarket profitably—together create powerful barriers to downward mobility. In the internal debates about resource allocation for new product development, therefore, proposals to pursue disruptive technologies generally lose out to proposals to move upmarket.
In fact, cultivating a systematic approach to weeding out new product development initiatives that would likely lower profits is one of the most important achievements of any well-managed company.
An important strategic implication of this rational pattern of upmarket movement is that it can create vacuum in low-end value networks that draws in entrants with technologies and cost structures better suited to competition. One of these powerful downmarket voids occurred in the steel industry, for example, when entrant companies employing disruptive minimill process technology entered through low-end beachheads; they have attacked relentlessly upmarket ever since.
T H E N O R T H E A S T E R L Y M I G R A T I O N
O F I N T E G R A T E D S T E E L
Minimill steel making first became commercially viable in the mid-1960s.
Employing widely available and familiar technology and equipment, minimills melt scrap steel in electric arc furnaces, continuously cast it into intermediate shapes called billets, and then roll those into products such as bars, rods, beams, or sheets. They are called minitnills because the scale at which they produce cost-competitive molten steel from scrap is less than one-tenth of the scale required for an integrated mill to produce cost-competitive molten steel from iron ore in blast and basic oxygen furnaces. (Integrated mills take their name from the integrated process of transforming iron ore, coal, and limestone into final steel shapes.) Inte-
88 | WHY GREAT COMPANIES CAN FAIL
grated mills and minimills look much the same in their processes of continuous casting and rolling operations. Scale is the only difference: The output of efficiently sized blast furnaces requires that integrated mills'
casting and rolling operations must be much greater than those of the minimills.
North America's steel minimills are the most efficient, lowest-cost steel makers in the world. In 1 9 9 5 , the most efficient minimill required 0.6
labor-hours per ton of steel produced; the best integrated mill required 2 . 3 labor-hours. In the product categories in which they compete, the average minimill can make product of equivalent quality, on a fully costed basis, at about a 15 percent lower cost than the average integrated mill.
In 1 9 9 5 , it cost about $ 4 0 0 million to build a cost-competitive steel minimill and about $6 billion to build a cost-competitive integrated mill.4
In terms of capital cost per ton of steel making capacity, integrated mills are more than four times as costly to build.5 As a result, minimills' share of the North American market has grown from nothing in 1 9 6 5 to 19
percent in 1 9 7 5 , 32 percent in 1 9 8 5 , and 40 percent in 1 9 9 5 . Experts predict they will account for half of all steel production by the turn of the century.6 Minimills virtually dominate the North American markets for rods, bars, and structural beams.
Yet not a single one of the world's major integrated steel companies to date has built a mill employing minimill technology. Why would none of them do something that makes so much sense? The explanation forwarded most frequently by the business press, especially in the United States, is that the managers of the integrated companies are conservative, backward-looking, risk-averse, and incompetent. Consider these indictments.
Last year, U.S. Steel Corp. closed fifteen of its facilities, claiming they had become "noncompetitive." Three years ago, Bethlehem Steel Corp. shuttered major portions of its plants in Johnstown, PA, and Lackawanna, NY. . . .
The closing of these major steel complexes is the final dramatic concession from today's chief executives that management has not been doing its job.
It represents decades of maximizing profits to look good for the short term.7
If the U.S. steel industry were as productive in tons per man-hour as it is in rhetoric per problem, it would be a top-notch performer.8
Surely there is some credibility to such accusations. But managerial incompetence cannot be a complete answer for the failure of North American integrated mills to counter the conquest by minimills of vast portions
. What Goes^Up, Can't Go Down | 89
of the steel industry. None of what most experts regard as the best-managed and most successful of the world's integrated steel makers—
including Nippon, Kawasaki, and N K K in Japan; British Steel and Hoogovens in Europe; and Pohang Steel in Korea—has invested in minimill technology even though it is demonstrably the lowest-cost technology in the world.
At the same time, in the last decade the management teams at integrated mills have taken aggressive steps to increase mill efficiency. U S X , for example, improved the efficiency of its steel making operations from more than nine labor-hours per ton of steel produced in 1 9 8 0 to just under three hours per ton in 1 9 9 1 . It accomplished this by ferociously attacking the size of its workforce, paring it from more than 9 3 , 0 0 0 in 1 9 8 0 to fewer than 2 3 , 0 0 0 in 1 9 9 1 , and by investing more than $2 billion in modernizing its plant and equipment. Yet all of this managerial aggressive-ness was targeted at conventional ways of making steel. H o w can this be?
Minimill steelmaking is a disruptive technology. When it emerged in the 1 9 6 0 s , because it used scrap steel, it produced steel of marginal quality.
The properties of its products varied according to the metallurgical compo-sition and impurities of the scrap. Hence, about the only market that minimill producers could address was that for steel reinforcing bars (rebars)—right at the bottom of the market in terms of quality, cost, and margins. This market was the least attractive of those served by established steel makers. And not only were margins low, but customers were the least loyal: They would switch suppliers at will, dealing with whoever offered the lowest price. The integrated steel makers were almost relieved to be rid of the rebar business.
The minimills, however, saw the rebar market quite differently. They had very different cost structures than those of the integrated mills: little depreciation and no research and development costs, low sales expenses (mostly telephone bills), and minimal general managerial overhead. They could sell by telephone virtually all the steel they could make—and sell it profitably.
Once they had established themselves in the rebar market, the most aggressive minimills, especially Nucor and Chaparral, developed a very different view of the overall steel market than the view that the integrated mills held. Whereas the downmarket rebar territory they seized had looked singularly unattractive to their integrated competitors, the minimills' view upmarket showed that opportunities for greater profits and expanded
9 0 I W H Y G R E A T C O M P A N I E S C A N F A I L
sales were all above them. With such incentive, they worked to improve the metallurgical quality and consistency of their products and invested in equipment to make larger shapes.
As the trajectory map in Figure 4 . 3 indicates, the minimills next attacked the markets for larger bars, rods, and angle irons immediately above them.
By 1 9 8 0 , they had captured 90 percent of the rebar market and held about 30 percent of the markets for bars, rods, and angle irons. At the time of the minimills' attack, the bar, rod, and angle iron shapes brought the lowest margins in the integrated mills' product lines. As a consequence, the integrated steel makers were, again, almost relieved to be rid of the business, and by the m i d - 1 9 8 0 s this market belonged to the minimills.
Once their position in the market for bars, rods, and angle irons seemed secure, the minimills continued their march upmarket, this time toward structural beams. Nucor did so from a new minimill plant in Arkansas, and Chaparral launched its attack from a new mill adjacent to its first
Figure 4.3 The Progress of Disruptive Minimill Steel Technology
• What Goes Up, Can't Go Down | 91
one in Texas. The integrated mills were driven from this market by the minimills as well. In 1 9 9 2 , U S X closed its South Chicago structural steel mill, leaving Bethlehem as the only integrated North American structural steel maker. Bethlehem closed its last structural beam plant in 1 9 9 5 , leaving the field to the minimills.
An important part of this story is that, throughout the 1 9 8 0 s , as they were ceding the bar and beam business to the minimills, the integrated steel makers experienced dramatically improving profit. Not only were these firms attacking cost, they were forsaking their lowest-margin products and focusing increasingly on high-quality rolled sheet steel, where quality-sensitive manufacturers of cans, cars, and appliances paid premium prices for metallurgically consistent steel with defect-free surfaces.
Indeed, the lion's share of integrated mills' investments in the 1 9 8 0 s had been targeted at improving their ability to provide the most demanding customers in these three markets with the highest-quality product and to do so profitably. Sheet steel markets were an attractive haven for the integrated producers in part because they were protected from minimill competition. It cost about $2 billion to build a state-of-the-art, cost-competitive sheet steel rolling mill, and this capital outlay simply had been too much for even the largest of the minimills.
Targeting the premium end of the market pleased the integrated mills'
investors: For example, Bethlehem Steel's market value had leapt from $ 1 7 5 million in 1 9 8 6 to $ 2 . 4 billion in 1 9 8 9 . This represented a very attractive return on the $ 1 . 3 billion the company invested in R&cD and plant and equipment during this period. The business press generously acknowledged these aggressive, well-placed investments.
Walter Williams (Bethlehem's CEO) has worked wonders. Over the past three years he mounted a highly personal campaign to improve the quality and productivity of Bethlehem's basic steel business. Bethlehem's metamor-phosis has outclassed even its major U.S. competitors—which as a whole are now producing at lower costs than their Japanese rivals and are fast closing the quality gap. Customers notice the difference. "It's nothing short of miraculous," says a top purchaser of sheet steel at Campbell Soup. [Italics
. added.]9
Another analyst made similar observations.
While almost no one was looking, a near miracle occurred: Big Steel is making a quiet comeback. Gary Works (US Steel) is back in the black . . .
92 I WHY GREAT COMPANIES CAN FAIL
pouring out a glowing river of molten iron at the rate of 3 million tons per year—a North American record. Union-management problem-solving teams are everywhere. Instead of making steel in all shapes and sizes, Gary has focused almost entirely on higher-value flat-rolled steel. [Italics added.]10
Almost all of us would agree that these remarkable recoveries were the fruits of good management. But where will good management in this genre lead these firms?
M I N I M I L L T H I N - S L A B C A S T I N G F O R S H E E T S T E E L
While integrated steel makers were busy engineering their recoveries, more disruptive clouds began gathering on the horizon. In 1 9 8 7 , a German supplier of equipment for the steel industry, Schloemann-Siemag A G , announced that it had developed what it called "continuous thin-slab casting" technology—a way for steel to be continuously cast from its molten state into long, thin slabs that could be transported directly, without cooling, into a rolling mill. Rolling the white-hot, already thin slab of steel to the final thickness of coiled sheet steel was much simpler than the traditional task mastered by the integrated mills of reheating and rolling sheet from thick ingots or slabs. M o s t important, a cost-competitive continuous thin-slab casting and rolling mill could be built for less than $ 2 5 0 million—one-tenth the capital cost of a traditional sheet mill and a relatively manageable investment for a minimill steel maker. At this scale, an electric arc furnace could easily supply the required quantity of molten steel. Moreover, thin-slab casting promised at least a 20 percent reduction in the total cost of making sheet steel.
Because of its promise, thin-slab casting was carefully evaluated by every major player in the steel industry. Some integrated mills, such as U S X , worked very hard to justify installation of a thin-slab facility.11 In the end, however, it was minimill Nucor Steel, rather than the integrated mills, that made the bold move into thin-slab casting. Why?
At the outset, thin-slab casting technology could not offer the smooth, defect-free surface finish required by the integrated mills' mainstream customers (makers of cans, cars, and appliances). The only markets were those such as construction decking and corrugated steel for culverts, pipes, and Quonset huts, in which users were more sensitive to price than to surface blemishes. Thin-slab casting was a disruptive technology. Furthermore, large, capable, and hungry integrated competitors were busy trying
• What Goes Up, Can't Go Down \ 93
to rob each other's most profitable business with the large auto, appliance, and can companies. It made no sense for them to target capital investment at thin-slab casting, positioned as it was in the least-profitable, most price-competitive and commodity-like end of their business. Indeed, after seriously considering between 1 9 8 7 and 1 9 8 8 whether to invest in thin-slab casting at an amount then projected to be about $ 1 5 0 million, both Bethlehem and U S X elected instead to invest in conventional thick-slab continuous casters at a cost of $ 2 5 0 million to protect and enhance the profitability of the business with their mainstream customers.
Not surprisingly, Nucor saw the situation another way. Unencumbered by the demands of profitable customers in the sheet steel business and benefiting from a cost structure forged at the bottom of the industry, Nucor fired up the world's first continuous thin-slab casting facility in Crawfordsville, Indiana, in 1 9 8 9 , and constructed a second mill in Hick-man, Arkansas, in 1 9 9 2 . It increased its capacity at both sites by 80
percent in 1 9 9 5 . Analysts estimate that Nucor had captured 7 percent of the massive North American sheet market by 1996—hardly enough to concern the integrated mills, because Nucor's success has been limited to the commoditized, least-profitable end of their product line. Of course, in its effort to win higher-margin business with higher-quality products from these mills, Nucor has already improved the surface quality of its sheet steel substantially.
Thus, the integrated steel companies' march to the profitable northeast corner of the steel industry is a story of aggressive investment, rational decision making, close attention to the needs of mainstream customers, and record profits. It is the same innovator's dilemma that confounded the leading providers of disk drives and mechanical excavators: Sound managerial decisions are at the very root of their impending fall from industry leadership.
N O T E S
1. This process of moving to higher tiers of the market and then adding the costs to support business at that level was described by Professor Malcom P. McNair, of the Harvard Business School, in a way that strikingly parallels the disk drive story. Writing in a history of retailing, McNair describes how successive waves of retailers entered the field with disruptive technologies (though he does not use the term):
The wheel always revolves, sometimes slowly, sometimes more rapidly,
94 | WHY GREAT COMPANIES CAN FAIL
but it does not stand still. The cycle frequently begins with the bold new concept, the innovation. Somebody gets a bright new idea. There is a John Wanamaker, a George Hartford (A&P), a Frank Woolworth, a W. T.
Grant, a General Wood (Sears), a Michael Cullen (supermarkets), a Eugene Ferkauf. Such an innovator has an idea for a new kind of distributive enterprise. At the outset he is in bad odor, ridiculed, scorned, condemned as "illegitimate." Bankers and investors are leery of him. But he attracts the public on the basis of the price appeal made possible by the low operating costs inherent in his innovation. As he goes along he trades up, improves the quality of his merchandise, improves the appearance and standing of his store, attains greater respectability. . . .
During this process of growth the institution rapidly becomes respectable in the eyes of both consumers and investors, but at the same time its capital investment increases and its operating costs tend to rise. Then the institution enters the stage of maturity. . . . The maturity phase soon tends to be followed by topheaviness . . . and eventual vulnerability. Vulnerability to what? Vulnerability to the next fellow who has a bright idea and who starts his business on a low-cost basis, slipping in under the umbrella that the old-line institutions have hoisted.
See Malcom P. McNair, "Significant Trends and Developments in the Post-war Period," in Albert B. Smith, ed., Competitive Distribution in a Free High-Level Economy and Its Implications for the University (Pittsburgh: University of Pittsburgh Press, 1958) 17-18. In other words, the very costs required to become competitive in higher-end markets restrict downward mobility and create further incentive to move upmarket.
2. Joseph Bower, Managing the Resource Allocation Process (Homewood, IL: Richard D. Irwin, 1970).
3. The use of the term systematic in this sentence is important, because most resource allocation systems work in a systematic way—whether the system is formal or informal. It will be shown later in this book that a key to managers' ability to confront disruptive technology successfully is their ability to intervene and make resource allocation decisions personally and persistently. Allocation systems are designed to weed out just such proposals as disruptive technologies. An excellent description of this dilemma can be found in Roger Martin, "Changing the Mind of the Corporation," Harvard Business Review, November-December 1993, 81-94.
4. Because of slow growth in steel demand in many of the world's markets, fewer large integrated steel mills are being built in the 1990s. Those integrated mills that are being built these days are in high-growth, rapidly developing countries such as Korea, Mexico, and Brazil.
5. Professor Thomas Eagar of the Department of Materials Science at the Massachusetts Institute of Technology provided these estimates.
' What Goes Up, Can't Go Down \ 95
6. "The U.S. Steel Industry: An Historical Overview," Goldman Sachs U.S.
Research Report, 1995.
7. "What Caused the Decline," Business Week, June 30, 1980, 74.
8. Donald B. Thompson, "Are Steel's Woes Just Short-term," Industry Week, February 22, 1982, 31.
9. Gregory L. Miles, "Forging the New Bethlehem," Business Week, June 5, 1989, 108-110.
10. Seth Lubove and James R. Norman, "New Lease on Life," Forbes, May 9, 1994, 87.
11. The experience of the team at U.S. Steel charged with evaluating continuous thin-slab casting technology is chronicled in the Harvard Business School teaching case "Continuous Casting Investments at USX Corporation," No.
697-020.
MANAGING DISRUPTIVE
TECHNOLOGICAL CHANGE
In the search for reasons why so many strong companies in three very different industries stumbled or failed, the research summarized in the preceding chapters casts doubt on several conventional explanations other researchers have offered. It wasn't the case that the leading companies'
engineers tended to get stuck in a particular technological paradigm or ignored innovations that were "not invented here." T h e cause of failure could not be solely attributed to established firms' inadequate competence in new technological fields or their inability to stay atop their industry's
"technological mudslide." Of course, these problems do afflict some companies. But as a general rule, the evidence is very strong that as long as the new technology was required to address the needs of their customers, established firms were able to muster the expertise, capital, suppliers, energy, and rationale to develop and implement the requisite technology both competitively and effectively. This has been true for incremental as well as radical advances; for projects that consumed months as well as those lasting more than a decade; in fast-paced disk drives, in the slower-paced mechanical excavator industry, and in the process-intensive steel industry.
Probably the most important outcome of this attempt to define the problem is that it ruled out poor management as a root cause. Again,
98 | MANAGING DISRUPTIVE TECHNOLOGICAL CHANGE
this is not to say that good and bad management aren't key factors affecting the fortunes of firms. But as a general explanation, the managers of the companies studied here had a great track record in understanding customers' future needs, identifying which technologies could best address those needs, and in investing to develop and implement them. It was only when confronted with disruptive technology that they failed. There had, therefore, to be a reason why good managers consistently made wrong decisions when faced with disruptive technological change.
The reason is that good management itself was the root cause. Managers played the game the way it was supposed to be played. The very decision-making and resource-allocation processes that are key to the success of established companies are the very processes that reject disruptive technologies: listening carefully to customers; tracking competitors' actions carefully; and investing resources to design and build higher-performance, higher-quality products that will yield greater profit. These are the reasons why great firms stumbled or failed when confronted with disruptive technological change.
Successful companies want their resources to be focused on activities that address customers' needs, that promise higher profits, that are technologically feasible, and that help them play in substantial markets. Yet, to expect the processes that accomplish these things also to do something like nurturing disruptive technologies—to focus resources on proposals that customers reject, that offer lower profit, that underperform existing technologies and can only be sold in insignificant markets—is akin to flapping one's arms with wings strapped to them in an attempt to fly.
Such expectations involve fighting some fundamental tendencies about the way successful organizations work and about how their performance is evaluated.
Part Two of this book is built upon detailed case studies of a few companies that succeeded, and many more that failed, when faced with disruptive technological change. Just as in our analogy to man's finally learning to fly when aviators ultimately came to understand and either harness or accommodate some fundamental laws of nature, these case studies show that those executives who succeeded tended to manage by a very different set of rules than those that failed. There were, in fact, four fundamental principles of organizational nature that managers in the successful firms consistently recognized and harnessed. The firms that lost their battles with disruptive technologies chose to ignore or fight them. These principles are:
Managing-Disruptive Technological Change | 99
1. Resource dependence: Customers effectively control the patterns of resource allocation in well-run companies.
2. Small markets don't solve the growth needs of large companies.
3. The ultimate uses or applications for disruptive technologies are unknowable in advance. Failure is an intrinsic step toward success.
4. Technology supply may not equal market demand. The attributes that make disruptive technologies unattractive in established markets often are the very ones that constitute their greatest value in emerging markets.
H o w did the successful managers harness these principles to their advantage?
1. They embedded projects to develop and commercialize disruptive technologies within an organization whose customers needed them. When managers aligned a disruptive innovation with the
"right" customers, customer demand increased the probability that the innovation would get the resources it needed.
2. They placed projects to develop disruptive technologies in organizations small enough to get excited about small opportunities and small wins.
3. They planned to fail early and inexpensively in the search for the market for a disruptive technology. They found that their markets generally coalesced through an iterative process of trial, learning, and trial again.
4. When commercializing disruptive technologies, they found or developed new markets that valued the attributes of the disruptive products, rather than search for a technological breakthrough so that the disruptive product could compete as a sustaining technology in mainstream markets.
Chapters 5 through 8 in Part Two describe in more detail how managers can address and harness these four principles. Each chapter starts by examining how harnessing or ignoring these principles affected the fortunes of disk drive companies when disruptive technologies were emerging.1 Each chapter then branches into an industry with very different characteristics, to show how the same principles drove the success and failure of firms confronted with disruptive technologies there.
1 0 0 I MANAGING DISRUPTIVE TECHNOLOGICAL CHANGE
The sum of these studies is that while disruptive technology can change the dynamics of industries with widely varying characteristics, the drivers of success or failure when confronted by such technology are consistent across industries.
Chapter 9 shows how these principles can be used by illustrating how managers might apply them in a case study of a particularly vexing technology—the electric vehicle. Chapter 10 then reviews the principal findings of the book.
N O T E S
1. The notion that we exercise power most effectively when we understand the physical and psychological laws that define the way the world works and then position or align ourselves in harmony with those laws, is of course not new to this book. At a light-hearted level, Stanford Professor Robert Burgelman, whose work is extensively cited in ths book, once dropped his pen onto the floor in a lecture. He muttered as he stooped to pick it up, "I hate gravity."
Then, as he walked to the blackboard to continue his line of thought, he added, "But do you know what? Gravity doesn't care!"
At a more serious level, the desirability of aligning our actions with the more powerful laws of nature, society, and psychology, in order to lead a productive life, is a central theme in many works, particularly the ancient Chinese classic, Tao De Ching.
C H A P T E R F I V E
Give Responsibility for
Disruptive Technologies to
Organizations Whose
Customers Need Them
M o s t executives would like to believe that they're in charge of their organizations, that they make the crucial decisions and that when they decide that something should be done everyone snaps to and executes. This chapter expands on the view already introduced: that in practice, it is a company's customers who effectively control what it can and cannot do. As we have seen in the disk drive industry, companies were willing to bet enormous amounts on technologically risky projects when it was clear that their customers needed the resulting products. But they were unable to muster the wherewithal to execute much simpler disruptive projects if existing, profitable customers didn't need the products.
This observation supports a somewhat controversial theory called resource dependence, propounded by a minority of management scholars,1
which posits that companies' freedom of action is limited to satisfying the needs of those entities outside the firm (customers and investors, primarily) that give it the resources it needs to survive. Drawing heavily upon concepts from biological evolution, resource dependence theorists assert that organizations will survive and prosper only if their staffs and systems serve the needs of customers and investors by providing them with the products, services, and profit they require. Organizations that do not will ultimately die off, starved of the revenues they need to survive.2
1 0 2 I MANAGING DISRUPTIVE TECHNOLOGICAL CHANGE
Hence, through this survival-of-the-fittest mechanism, those firms that rise to prominence in their industries generally will be those whose people and processes are most keenly tuned to giving their customers what they want. The controversy with this theory arises when its proponents conclude that managers are powerless to change the courses of their firms against the dictates of their customers. Even if a manager has a bold vision to take her or his company in a very different direction, the power of the customer-focused people and processes in any company well-adapted to survival in its competitive environment will reject the manager's attempts to change direction. Therefore, because they provide the resources upon which the firm is dependent, it is the customers, rather than the managers, who really determine what a firm will do. It is forces outside the organization, rather than the managers within it, that dictate the company's course.
Resource dependence theorists conclude that the real role of managers in companies whose people and systems are well-adapted to survival is, therefore, only a symbolic one.
For those of us who have managed companies, consulted for management, or taught future managers, this is a most disquieting thought. We are there to manage, to make a difference, to formulate and implement strategy, to accelerate growth and improve profits. Resource dependence violates our very reason for being. Nonetheless, the findings reported in this book provide rather stunning support for the theory of resource dependence—especially for the notion that the customer-focused resource allocation and decision-making processes of successful companies are far more powerful in directing investments than are executives' decisions.
Clearly, customers wield enormous power in directing a firm's investments. What, then, should managers do when faced with a disruptive technology that the company's customers explicitly do not want? One option is to convince everyone in the firm that the company should pursue it anyway, that it has long-term strategic importance despite rejection by the customers who pay the bills and despite lower profitability than the upmarket alternatives. The other option would be to create an independent organization and embed it among emerging customers that do need the technology. Which works best?
Managers who choose the first option essentially are picking a fight with a powerful tendency of organizational nature—that customers, not managers, essentially control the investment patterns of a company. By contrast, managers who choose the second option align themselves with this tendency, harnessing rather than fighting its power. The cases pre-
Responsibility for Disruptive Technologies | 1 0 3
sented in this chapter provide strong evidence that the second option offers far higher probabilities of success than the first.
I N N O V A T I O N A N D R E S O U R C E A L L O C A T I O N
T h e mechanism through which customers control the investments of a firm is the resource allocation process—the process that determines which initiatives get staff and money and which don't. Resource allocation and innovation are two sides of the same coin: Only those new product development projects that do get adequate funding, staffing, and management attention have a chance to succeed; those that are starved of resources will languish. Hence, the patterns of innovation in a company will mirror quite closely the patterns in which resources are allocated.
Good resource allocation processes are designed to weed out proposals that customers don't want. When these decision-making processes work well, if customers don't want a product, it won't get funded; if they do want it, it will. This is how things must work in great companies. They must invest in things customers want—and the better they become at doing this, the more successful they will be.
As we saw in chapter 4, resource allocation is not simply a matter of top-down decision making followed by implementation. Typically, senior managers are asked to decide whether to fund a project only after many others at lower levels in the organization have already decided which types of project proposals they want to package and send on to senior management for approval and which they don't think are worth the effort.
Senior managers typically see only a well-screened subset of the innovative ideas generated.3
And even after senior management has endorsed funding for a particular project, it is rarely a "done deal." Many crucial resource allocation decisions are made after project approval—indeed, after product launch—by mid-level managers who set priorities when multiple projects and products compete for the time of the same people, equipment, and vendors. As management scholar Chester Barnard has noted:
From the point of view of the relative importance of specific decisions, those of executives properly call for first attention. But from the point of view of aggregate importance, it is not decisions of executives, but of non-executive participants in organizations which should enlist major interest. [Italics added.]4
1 0 4 | MANAGING DISRUPTIVE TECHNOLOGICAL CHANGE
So how do non-executive participants make their resource allocation decisions? They decide which projects they will propose to senior management and which they will give priority to, based upon their understanding of what types of customers and products are most profitable to the company. Tightly coupled with this is their view of how their sponsorship of different proposals will affect their own career trajectories within the company, a view that is formed heavily by their understanding of what customers want and what types of products the company needs to sell more of in order to be more profitable. Individuals' career trajectories can soar when they sponsor highly profitable innovation programs. It is through these mechanisms of seeking corporate profit and personal success, therefore, that customers exert a profound influence on the process of resource allocation, and hence on the patterns of innovation, in most companies.
S U C C E S S I N D I S R U P T I V E D I S K D R I V E T E C H N O L O G Y
It is possible to break out of this system of customer control, however.
Three cases in the history of the disk drive industry demonstrate how managers can develop strong market positions in a disruptive technology.
In two cases, managers harnessed, rather than fought, the forces of resource dependence: They spun out independent companies to commercialize the disruptive technology. In the third, the manager chose to fight these forces, and survived the project, exhausted.
Quantum and Plus Development
As we have seen, Quantum Corporation, a leading maker of 8-inch drives sold in the minicomputer market in the early 1 9 8 0 s , completely missed the advent of 5.25-inch drives: It introduced its first versions nearly four years after those drives first appeared in the market. As the 5.25-inch pioneers began to invade the minicomputer market from below, for all the reasons already described, Quantum's sales began to sag.
In 1 9 8 4 several Quantum employees saw a potential market for a thin 3.5-inch drive plugged into an expansion slot in I B M X T - and AT-class desktop computers—drives that would be sold to personal computer users rather than the O E M minicomputer manufacturers that had accounted for all of Quantum's revenue. They determined to leave Quantum and start a new firm to commercialize their idea.
Responsibility for Disruptive Technologies | 1 0 5
Rather than let them leave unencumbered, however, Quantum's executives financed and retained 80 percent ownership of this spinoff venture, called Plus Development Corporation, and set the company up in different facilities. It was a completely self-sufficient organization, with its own executive staff and all of the functional capabilities required in an independent company. Plus was extremely successful. It designed and marketed its drives but had them manufactured under contract by Matsushita Kotobuki Electronics ( M K E ) in Japan.
As sales of Quantum's line of 8-inch drives began to evaporate in the mid-1980s, they were offset by Plus's growing "Hardcard" revenues. By 1 9 8 7 , sales of Quantum's 8- and 5.25-inch products had largely disappeared. Quantum then purchased the remaining 20 percent of Plus, essentially closed down the old corporation, and installed Plus's executives in Quantum's most senior positions. They then reconfigured Plus's 3.5-inch products to appeal to O E M desktop computer makers, such as Apple, just as the capacity vector for 3.5-inch drives was invading the desktop market, as shown in the disk drive trajectory map in Figure 1.7. Quantum, thus reconstituted as a 3.5-inch drive maker, has aggressively adopted sustaining component technology innovations, moving upmarket toward engineering workstations, and has also successfully negotiated the sustaining architectural innovation into 2.5-inch drives. By 1 9 9 4 the new Quantum had become the largest unit-volume producer of disk drives in the world.5
Control Data in Oklahoma
Control Data Corporation (CDC) effected the same self-reconstitution—
once. C D C was the dominant manufacturer of 14-inch drives sold into the O E M market between 1 9 6 5 and 1 9 8 2 ; its market share fluctuated between 55 and 62 percent. When the 8-inch architecture emerged in the late 1 9 7 0 s , however, C D C missed it—by three years. The company never captured more than a fraction of the 8-inch market, and those 8-inch drives that it did sell were sold almost exclusively to defend its established customer base of mainframe computer manufacturers. The reason was resources and managerial emphasis: Engineers and marketers at the company's principal Minneapolis facility kept getting pulled off the 8-inch program to resolve problems in the launch of next-generation 14-inch products for CDC's mainstream customers.
C D C launched its first 5.25-inch model two years after Seagate's pio-
1 0 6 I MANAGING DISRUPTIVE TECHNOLOGICAL CHANGE
neering product appeared in 1 9 8 0 . This time, however, C D C located its 5.25-inch effort in Oklahoma City. This was done, according to one manager, "not to escape CDC's Minneapolis engineering culture, but to isolate the [5.25-inch product] group from the company's mainstream customers." Although it was late in the market and never regained its former dominant position, CDC's foray into 5.25-inch drives was profitable, and at times the firm commanded a 20 percent share of higher-capacity 5.25-inch drives.
Micropolis: Transition by Managerial Force
Micropolis Corporation, an early disk drive leader founded in 1 9 7 8 to make 8-inch drives, was the only other industry player to successfully make the transition to a disruptive platform. It did not use the spin-out strategy that had worked for Quantum and Control Data, however, choosing instead to manage the change from within the mainstream company. But even this exception supports the rule that customers exert exceptionally powerful influence over the investments that firms can un-dertake successfully.
Micropolis began to change in 1 9 8 2 , when founder and C E O Stuart M a b o n intuitively perceived the trajectories of market demand and technology supply mapped in Figure 1.7 and decided that the firm should become primarily a maker of 5.25-inch drives. While initially hoping to keep adequate resources focused on developing its next generation of 8-inch drives so that Micropolis could straddle both markets,6 he assigned the company's premier engineers to the 5.25-inch program. M a b o n recalls that it took " 1 0 0 percent of my time and energy for eighteen months" to keep adequate resources focused on the 5.25-inch program, because the organization's own mechanisms allocated resources to where the customers were—8-inch drives.
By 1 9 8 4 , Micropolis had failed to keep pace with competition in the minicomputer market for disk drives and withdrew its remaining 8-inch models. With Herculean effort, however, it did succeed in its 5.25-inch programs. Figure 5.1 shows why this struggle occurred: In making the transition, Micropolis assumed a position on a very different technological trajectory. It had to walk away from every one of its major customers and replace the lost revenues with sales of the new product line to an entirely different group of desktop computer makers. M a b o n remembers the experience as the most exhausting of his life.
Responsibility for Disruptive Technologies | 1 0 7
Figure 5.1 Technology Transition and Market Position at Micropolis Corporation
Source: D a t a a r e f r o m v a r i o u s i s s u e s of Disk/Trend Report.
Micropolis finally introduced a 3.5-inch product in 1 9 9 3 . That was the point at which the product had progressed to pack more than 1
gigabyte in the 3.5-inch platform. At that level, Micropolis could sell the 3.5-inch drive to its existing customers.
D I S R U P T I V E T E C H N O L O G I E S A N D T H E T H E O R Y O F
R E S O U R C E D E P E N D E N C E
The struggles recounted earlier of Seagate Technology's attempts to sell 3.5-inch drives and of Bucyrus Erie's failed attempt to sell its early Hydrohoe only to its mainstream customers illustrate how the theory of resource dependence can be applied to cases of disruptive technologies.
In both instances, Seagate and Bucyrus were among the first in their
1 0 8 | MANAGING DISRUPTIVE TECHNOLOGICAL CHANGE
industries to develop these disruptive products. But despite senior managers' decisions to introduce them, the impetus or organizational energy required to launch the products aggressively into the appropriate value networks simply did not coalesce—until customers needed them.
Should we then accept the corollary stipulated by resource-dependence theorists that managers are merely powerless individuals? Hardly. In the Introduction, exploring the image of how people learned to fly, I noted that all attempts had ended in failure as long as they consisted of fighting fundamental laws of nature. But once laws such as gravity, Bernoulli's principle, and the notions of lift, drag and resistance began to be understood, and flying machines were designed that accounted for or harnessed those laws, people flew quite successfully. By analogy, this is what Quantum and Control Data did. By embedding independent organizations within an entirely different value network, where they were dependent upon the appropriate set of customers for survival, those managers harnessed the powerful forces of resource dependence. The C E O of M i c r o -
polis fought them, but he won a rare and costly victory.
Disruptive technologies have had deadly impact in many industries besides disk drives, mechanical excavators, and steel.7 The following pages summarize the effect of disruptive technologies in three other industries—
computers, retailing, and printers—to highlight how the only companies in those industries that established strong market positions in the disruptive technologies were those which, like Quantum and Control Data, harnessed rather than fought the forces of resource dependence.
D E C , I B M , A N D T H E P E R S O N A L C O M P U T E R
Quite naturally, the computer industry and the disk drive industry have parallel histories, because value networks of the latter are embedded in those of the former. In fact, if the axes and intersecting trajectories depicted on the disk drive trajectory map in Figure 1.7 were relabeled with computer-relevant terms, it would summarize equally well the failure of leading computer industry firms. I B M , the industry's first leader, sold its mainframe computers to the central accounting and data processing departments of large organizations. The emergence of the minicomputer represented a disruptive technology to I B M and its competitors. Their customers had no use for it; it promised lower, not higher, margins; and the market initially was significantly smaller. As a result, the makers of mainframes ignored the minicomputer for years, allowing a set of
Responsibility for Disruptive Technologies | 1 0 9
entrants—Digital Equipment, Data General, Prime, Wang, and N i x -
dorf—to create and dominate that market. I B M ultimately introduced its own line of minicomputers, but it did so primarily as a defensive measure, when the capabilities of minicomputers had advanced to the point that they were performance-competitive with the computing needs of some of IBM's customers.
Similarly, none of the makers of minicomputers became a significant factor in the desktop personal computer market, because to them the desktop computer was a disruptive technology. The PC market was created by another set of entrants, including Apple, Commodore, Tandy, and I B M . The minicomputer makers were exceptionally prosperous and highly regarded by investors, the business press, and students of good management—until the late 1 9 8 0 s , when the technological trajectory of the desktop computer intersected with the performance demanded by those who had previously bought minicomputers. The missile-like attack of the desktop computer from below severely wounded every minicomputer maker.
Several of them failed. None established a viable position in the desktop personal computer value network.
A similar sequence of events characterized the emergence of the portable computer, where the market was created and dominated by a set of entrants like Toshiba, Sharp, and Zenith. Apple and I B M , the leading desktop makers, did not introduce portable models until the portables'
performance trajectory intersected with the computing needs of their customers.
Probably none of these firms has been so deeply wounded by disruptive technology as Digital Equipment. D E C fell from fortune to folly in just a few years, as stand-alone workstations and networked desktop computers obviated most customers' needs for minicomputers almost overnight.
D E C didn't stumble for lack of trying, of course. Four times between 1 9 8 3 and 1 9 9 5 it introduced lines of personal computers targeted at consumers, products that were technologically much simpler than DEC's minicomputers. But four times it failed to build businesses in this value network that were perceived within the company as profitable. Four times it withdrew from the personal computer market. Why? D E C launched all four forays from within the mainstream company.8 For all of the reasons so far recounted, even though executive-level decisions lay behind the move into the PC business, those who made the day-to-day resource allocation decisions in the company never saw the sense in investing the necessary money, time, and energy in low-margin products that their
1 1 0 I MANAGING DISRUPTIVE TECHNOLOGICAL CHANGE
customers didn't want. Higher-performance initiatives that promised upscale margins, such as D E C ' s super-fast Alpha microprocessor and its adventure into mainframe computers, captured the resources instead.
In trying to enter the desktop personal computing business from within its mainstream organization, D E C was forced to straddle the two different cost structures intrinsic to two different value networks. It simply couldn't hack away enough overhead cost to be competitive in low-end personal computers because it needed those costs to remain competitive in its higher-performance products.
Yet I B M ' s success in the first five years of the personal computing industry stands in stark contrast to the failure of the other leading mainframe and minicomputer makers to catch the disruptive desktop computing wave. H o w did I B M do it? It created an autonomous organization in Florida, far away from its New York state headquarters, that was free to procure components from any source, to sell through its own channels, and to forge a cost structure appropriate to the technological and competitive requirements of the personal computing market. The organization was free to succeed along metrics of success that were relevant to the personal computing market. In fact, some have argued that IBM's subsequent decision to link its personal computer division much more closely to its mainstream organization was an important factor in I B M ' s difficulties in maintaining its profitability and market share in the personal computer industry. It seems to be very difficult to manage the peaceful, unambiguous coexistence of two cost structures, and two models for how to make money, within a single company.
T h e conclusion that a single organization might simply be incapable of competently pursuing disruptive technology, while remaining competitive in mainstream markets, bothers some "can-do" managers—and, in fact, most managers try to do exactly what Micropolis and D E C did: maintain their competitive intensity in the mainstream, while simultaneously trying to pursue disruptive technology. The evidence is strong that such efforts rarely succeed; position in one market will suffer unless two separate organizations, embedded within the appropriate value networks, pursue their separate customers.
K R E S G E , W O O L W O R T H , A N D D I S C O U N T R E T A I L I N G
In few industries has the impact of disruptive technology been felt so pervasively as in retailing, where discounters seized dominance from tradi-
Responsibility for Disruptive Technologies | 1 1 1
tional department and variety stores. The technology of discount retailing was disruptive to traditional operations because the quality of service and selection offered by discounters played havoc with the accustomed metrics of quality retailing. Moreover, the cost structure required to compete profitably in discount retailing was fundamentally different than that which department stores had developed to compete within their value networks.
The first discount store was Korvette's, which began operating a number of outlets in New York in the mid-1950s. Korvette's and its imitators operated at the very low end of retailing's product line, selling nationally known brands of standard hard goods at 20 to 40 percent below department store prices. They focused on products that "sold themselves" because customers already knew how to use them. Relying on national brand image to establish the value and quality of their products, these discounters eliminated the need for knowledgeable salespeople; they also focused on the group of customers least attractive to mainstream retailers:
"young wives of blue collar workers with young children."9 This was counter to the upscale formulas department stores historically had used to define quality retailing and to improve profits.
Discounters didn't accept lower profits than those of traditional retailers, however; they just earned their profits through a different formula.
In the simplest terms, retailers cover their costs through the gross margin, or markup, they charge over the cost of the merchandise they sell. Traditional department stores historically marked merchandise up by 40 percent and turned their inventory over four times in a year—that is, they earned 40 percent on the amount they invested in inventory, four times during the year, for a total return on inventory investment of 1 6 0 percent. Variety stores earned somewhat lower profits through a formula similar to that used by the department stores. Discount retailers earned a return on inventory investment similar to that of department stores, but through a different model: low gross margins and high inventory turns. Table 5.1
summarizes the three positions.
The history of discount retailing vividly recalls the history of minimill steel making. Just like the minimills, discounters took advantage of their cost structure to move upmarket and seize share from competing traditional retailers at a stunning rate: first at the low end, in brand-name hard goods such as hardware, small appliances, and luggage, and later in territory further to the northeast such as home furnishings and clothing.
Figure 5.2 illustrates how stunning the discounters' invasion was: Their
1 1 2 M A N A G I N G D I S R U P T I V E T E C H N O L O G I C A L C H A N G E
Table 5.1 Different Pathways to Profits Typical
Typical
Return on
Company
Gross
Inventory
Inventory
Retailer Type
Example
Margins
Turns
Investment*
Department stores
R. H. Macy
40%
4x
160%
Variety stores
F. W. Woolworth
36%
4x
144%
Discount retailers
Kmart
20%
8x
160%
' C a l c u l a t e d a s M a r g i n s x T u r n s , i n o t h e r w o r d s , t h e total o f t h e m a r g i n s e a r n e d t h r o u g h s u c c e s s i v e t u r n o v e r s e a c h year.
Source: A n n u a l c o r p o r a t e r e p o r t s o f m a n y c o m p a n i e s i n e a c h c a t e g o r y f o r v a r i o u s y e a r s .
Figure 5.2 Gains in Discount Retailers' Market Share, 1960-1966
Source: D a t a are f r o m v a r i o u s i s s u e s of Discount Merchandiser.
share of retailing revenues in the categories of goods they sold rose from 10 percent in 1 9 6 0 to nearly 40 percent a scant six years later.
Just as in disk drives and excavators, a few of the leading traditional retailers—notably S. S. Kresge, F. W. Woolworth, and Dayton Hudson—
saw the disruptive approach coming and invested early. None of the other
Responsibility for Disruptive Technologies | 1 1 3
major retail chains, including Sears, Montgomery Ward, J. C. Penney, and R. H. Macy, made a significant attempt to create a business in discount retailing. Kresge (with its Kmart chain) and Dayton Hudson (with the Target chain) succeeded.10 They both created focused discount retailing organizations that were independent from their traditional business. They recognized and harnessed the forces of resource dependence. By contrast, Woolworth failed in its venture (Woolco), trying to launch it from within the F. W. Woolworth variety store company. A detailed comparison of the approaches of Kresge and Woolworth, which started from very similar positions, lends additional insight into why establishing independent organizations to pursue disruptive technology seems to be a necessary condition for success.
S. S. Kresge, then the world's second largest variety store chain, began studying discount retailing in 1 9 5 7 , while discounting was still in its infancy. By 1 9 6 1 , both Kresge and its rival F. W. Woolworth (the world's largest variety store operator) had announced initiatives to enter discount retailing. Both firms opened stores in 1 9 6 2 , within three months of each other. The performance of the Woolco and Kmart ventures they launched, however, subsequently differed dramatically. A decade later, Kmart's sales approached $ 3 . 5 billion while Woolco's sales were languishing unprofitably at $ 0 . 9 billion.11
In making its commitment to discount retailing, Kresge decided to exit the variety store business entirely: In 1 9 5 9 it hired a new C E O , Harry Cunningham, whose sole mission was to convert Kresge into a discounting powerhouse. Cunningham, in turn, brought in an entirely new management team, so that by 1 9 6 1 there "was not a single operating vice president, regional manager, assistant regional manager, or regional merchandise manager who was not new on the j o b . " 1 2 In 1 9 6 1 Cunningham stopped opening any new variety stores, embarking instead on a program of closing about 10 percent of Kresge's existing variety operations each year. This represented a wholesale refocusing of the company on discount retailing.
Woolworth, on the other hand, attempted to support a program of sustaining improvements in technology, capacity, and facilities in its core variety store businesses while simultaneously investing in disruptive discounting. The managers charged with improving the performance of Woolworth's variety stores were also charged with building "the largest chain of discount houses in America." C E O Robert Kirkwood asserted that Woolco "would not conflict with the company's plans for growth
1 1 4 I MANAGING DISRUPTIVE TECHNOLOGICAL CHANGE
and expansion in the regular variety store operations," and that no existing stores would be converted to a discount format.13 Indeed, as discount retailing hit its most frenzied expansion phase in the 1 9 6 0 s , Woolworth was opening new variety stores at the pace it had set in the 1 9 5 0 s .
Unfortunately (but predictably), Woolworth proved unable to sustain within a single organization the two different cultures, and two different models of how to make a profit, that were required to be successful in variety and discount retailing. By 1 9 6 7 it had dropped the term "discount"
from all Woolco advertising, adopting the term "promotional department store" instead. Although initially Woolworth had set up a separate administrative staff for its Woolco operation, by 1 9 7 1 more rational, cost-conscious heads had prevailed.
In a move designed to increase sales per square foot in both Woolco and Woolworth divisions, the two subsidiaries have been consolidated operationally on a regional basis. Company officials say the consolidation—which involves buying offices, distribution facilities and management personnel at the regional level—will help both to develop better merchandise and more efficient stores. Woolco will gain the benefits of Woolworth's buying resources, distribution facilities and additional expertise in developing specialty departments. In return, Woolworth will gain Woolco's knowhow in locating, designing, promoting and operating large stores over 100,000 sq. ft.14
W h a t was the impact of this cost-saving consolidation? It provided more evidence that two models for how to make money cannot peacefully coexist within a single organization. Within a year of this consolidation, Woolco had increased its markups such that its gross margins were the highest in the discount industry—about 33 percent. In the process, its inventory turns fell from the 7x it originally had achieved to 4 x . The formula for profit that had long sustained F. W. Woolworth (35 percent margins for four inventory turns or 1 4 0 percent return on inventory investment) was ultimately demanded of Woolco as well. (See Figure 5.3.) Woolco was no longer a discounter—in name or in fact. Not surprisingly, Woolworth's venture into discount retailing failed: It closed its last Woolco store in 1 9 8 2 .
Woolworth's organizational strategy for succeeding in disruptive discount retailing was the same as Digital Equipment's strategy for launching its personal computer business. Both founded new ventures within the mainstream organization that had to earn money by mainstream rules,
Responsibility for Disruptive Technologies | 1 1 5
Figure 5.3 Impact of the Integration of Woolco, and F. W. Woolworth on the Way Woolco Attempted to Make Money
Source: D a t a a r e f r o m v a r i o u s a n n u a l r e p o r t s o f F . W . W o o l w o r t h C o m p a n y a n d f r o m v a r i o u s i s s u e s of Discount Merchandiser.
and neither could achieve the cost structure and profit model required to succeed in the mainstream value network.
S U R V I V A L B Y S U I C I D E : H E W L E T T - P A C K A R D ' S L A S E R J E T
A N D I N K - J E T P R I N T E R S
Hewlett-Packard's experience in the personal computer printer business illustrates how a company's pursuit of a disruptive technology by spinning out an independent organization might entail, in the end, killing another of its business units.
Hewlett-Packard's storied success in manufacturing printers for personal computers becomes even more remarkable when one considers its management of the emergence of bubble-jet or ink-jet technology. Beginning in the mid-1980s, HP began building a huge and successful business around laser jet printing technology. The laser jet was a discontinuous
1 1 6 | MANAGING DISRUPTIVE TECHNOLOGICAL CHANGE
improvement over dot-matrix printing, the previously dominant personal computer printing technology, and HP built a commanding market lead.
When an alternative way of translating digital signals into images on paper (ink-jet technology) first appeared, there were vigorous debates about whether laser jet or ink jet would emerge as the dominant design in personal printing. Experts lined up on both sides of the question, offering HP extensive advice on which technology would ultimately become the printer of choice on the world's desktops.15
Although it was never framed as such in the debates of the time, ink-jet printing was a disruptive technology. It was slower than the laser jet, its resolution was worse, and its cost per printed page was higher. But the printer itself was smaller and potentially much less expensive than the laser jet. At these lower prices, it promised lower gross margin dollars per unit than the laser jet. Thus, the ink-jet printer was a classic disruptive product, relative to the laser jet business.
Rather than place its bet exclusively with one or the other, and rather than attempt to commercialize the disruptive ink-jet from within the existing printer division in Boise, Idaho, HP created a completely autonomous organizational unit, located in Vancouver, Washington, with responsibility for making the ink-jet printer a success. It then let the two businesses compete against each other. Each has behaved classically. As shown in Figure 5 . 4 , the laser jet division has moved sharply upmarket, in a strategy reminiscent of 14-inch drives, mainframe computers, and integrated steel mills. HP's laser jet printers can print at high speeds with exceptional resolution; handle hundreds of fonts and complicated graphics; print on two sides of the page; and serve multiple users on a network. They have also gotten larger physically.
T h e ink-jet printer isn't as good as the laser jet and may never be. But the critical question is whether the ink jet could ever be as good a printer as the personal desktop computing market demands. T h e answer appears to be yes. The resolution and speed of ink-jet printers, while still inferior to those of laser jets, are now clearly good enough for many students, professionals, and other un-networked users of desktop computers.
HP's ink-jet printer business is now capturing many of those who would formerly have been laser jet users. Ultimately, the number of users at the highest-performance end of the market, toward which the laser jet division is headed, will probably become small. One of HP's businesses may, in the end, have killed another. But had HP not set up its ink-jet business as a separate organization, the ink-jet technology would probably have languished within the mainstream laser jet business, leaving one of the
Responsibility for Disruptive Technologies | 1 1 7
Figure 5.4 Speed Improvement in InkJet and LaserJet Printers
Source: H e w l e t t - P a c k a r d p r o d u c t b r o c h u r e s , v a r i o u s y e a r s .
other companies now actively competing in the ink-jet printer business, such as Canon, as a serious threat to HP's printer business. And by staying in the laser business, as well, HP has joined IBM's mainframe business and the integrated steel companies in making a lot of money while executing an upmarket retreat.16
A D D I T I O N A L O R G A N I Z A T I O N A L I M P L I C A T I O N S O F
D I S R U P T I V E T E C H N O L O G I E S
In addition to the advisability of creating substantial organizational distance between the mainstream enterprise and the organization charged
1 1 8 | MANAGING DISRUPTIVE TECHNOLOGICAL CHANGE
with commercializing disruptive technology, there is another dimension to the framework that should guide decisions about how to organize for innovation: The team should be structured to facilitate the cross-functional interaction that is characteristic of different types of projects.
Scholars have shown that in the initial phases of an industry's history, most technical energy is expended on architectural innovation, using materials and technologies that are generally available in the marketplace.17
Product designs tend to be integral, meaning that the design of most individual components is integral to, or impacts the design of, many other components. Hence, projects at this stage of an industry's development are best managed by tightly integrated teams. Eventually, however, a dominant architectural design for the product category emerges. After that point, the application of technological energies tends to shift toward improving the performance and cost-effectiveness of individual components and subsystems within the framework of the dominant architectural design. Standard interfaces or specifications emerge, which define how the components interact with each other. This makes it possible to plug and play, or mix and match, components from a variety of sources.
As the shift from architectural- to component-level innovation occurs, technologically specialized groups tend to be organized in established firms to create deep expertise and focus on improvements to each component. M a j o r functional groups such as marketing and manufacturing are established, each to focus on its own component of the company's value-added. This specialization occurs within specific functions such as engineering as well. For example, the division within an automobile manufacturer responsible for designing steering gear typically will be organized in groups that mirror the components that comprise the steering gear: a steering column group, a rack-and-pinion gear group, a tie rod group, a power steering pump group, and so on.
T h e way the components work together to constitute a steering system is defined by the design of the system's architecture. Scholars have found that, over time, the patterns of interaction and communication between the individuals and groups within the organization will come to mirror the manner in which the components themselves interact within the product architecture. In the steering gear example, members of the groups responsible for designing hoses and power steering pumps would interact regularly, because the components for which they have responsibility interact inti-mately. When they make a change in pump design, the pump engineers will know who in the hose group needs to be informed of the change,
Responsibility for Disruptive Technologies | 1 1 9
what they need to know, and when they need to know it to get their own work done effectively. By the same token, members of the power steering pump group would rarely interact with those in the group responsible for tie rods. Because the two components have little to do with each other in the architecture of the steering gear, their respective engineering groups will never learn how to work together. This symmetry between the product's architecture and the organization's design is depicted symbolically in Figure 5 . 5 .
This organizational structure and the patterns of working together that it implies serve a company well as long as the innovations are modular, that is, as long as the technological change is largely self-contained within each component and does not require that other components in the system be redesigned to accommodate the change. In particular, as long as the interface does not change, the new component will fit both into the product architecture, and into the organizational structure. The organization will know how to work to get the j o b done. In projects of this character, loosely structured or lightweight project teams, in which the project leader is essentially a coordinator and an expediter, can be quite successful.18
However, when a project involves a new architectural design, then the organization's structure and its well-defined and lubricated patterns of
Figure 5.5 The Organizational Structure Mirrors the Established Product's Architecture
1 2 0 | MANAGING DISRUPTIVE TECHNOLOGICAL CHANGE
communication and interaction can become barriers to, rather than facili-tators of, effective innovation. If redesign of a component impacts the performance of other component in a different way, engineers may not know with whom they need to work, what they need to know, when they need to know it, or how to solve new mutual problems when an innovation involves a significant change in product architecture and changes the intermodule interfaces significantly.
The same is true at a broader level. When a product redesign requires that manufacturing, purchasing, and marketing coordinate their activities around different questions at different points in the project than had been the historical norm—when the way they interface must change—then a strong or heavyweight team structure is important. When heavyweight team members can work together in a dedicated way, they are free from organizational rhythms, habits, and reporting requirements and are free to forge new patterns of interaction and problem solving.
Hence, there appear to be two factors to determine what type of organizational structure will best facilitate a project's success: The extent to which the innovation will require people and groups to interact with different people, upon different subjects, and with differing timing, than has habitually been the case; and the extent of disruptiveness inherent in the technology. The matrix in Figure 5.6 summarizes this framework: Its left axis measures the degree of modularity inherent in the innovation, which drives the appropriate organization of the development team. The bottom axis measures the degree of the innovation's disruptiveness, which determines the necessary organizational distance from the mainstream organization required to successfully commercialize products of each type.
The implication here is that no single development or commercial structure is appropriate for all types of products and technologies. M a j o r development projects that involve significant architectural innovation, requiring different patterns of working together, but which are sustaining in character, can be managed through heavyweight teams within the mainstream organization. Disruptive projects, however, even technologically straightforward ones, can thrive only within organizationally distinct units.
H A R N E S S I N G O R I G N O R I N G T H E F O R C E S O F R E S O U R C E
D E P E N D E N C E : A S U M M A R Y
In the cases of Quantum, Control Data, I B M , and Hewlett-Packard cited above, the innovating managers who were faced with disruptive technol-
Responsibility for Disruptive Technologies j 1 2 1
Figure 5.6 The Organizational Structures for New Products and Services
Appropriate Position of Responsible Commercial Structure
ogy created organizations whose cost structures enabled them to make money in the value network where the disruptive technology was taking root, and where customers' power and the managers' intentions were aligned. By harnessing their efforts to the forces of resource dependence, they succeeded in the face of disruptive change. Kresge's Cunningham achieved this alignment differently, by cutting off the customer base that historically had provided the company's resources, thereby intensifying Kresge's dependence on the new source of resources in discount retailing.19
In other known instances in the industries described in this chapter, the managers of companies that faced disruptive technology change also proved the resource dependence theorists right: They simply lacked the power to change the course of their companies when their own capabilities were pitted against the power of customer-focused, profit-driven, rational systems of resource allocation in their companies.20
And finally, lest readers take these observations as a blanket recommendation to create insulated skunk works as the key to the success of every project, this chapter proposed a framework that describes the circumstances that should dictate both the structure of project teams and the
122 MANAGING DISRUPTIVE TECHNOLOGICAL CHANGE
distance or insulation that should be created or need not be created, between the team and the power and processes of the mainstream organization.
N O T E S
1. The theory of resource dependence has been most thoroughly argued by Jeffrey Pfeffer and Gerald R. Salancik in The External Control of Organizations: A Resource Dependence Perspective (New York: Harper & Row, 1978).
2. This implies that, in managing business under both normal conditions and conditions of assault by a disruptive technology, the choice of which customers the firm will serve has enormous strategic consequences.
3. Joseph L. Bower, in Managing the Resource Allocation Process (Homewood, IL: Richard D. Irwin, 1972), presents an elegant and compelling picture of the resource allocation process.
4. Chester Barnard, The Functions of the Executive (Cambridge, MA: Harvard University Press, 1938), 190-191.
5. Quantum's spin-out of the Hardcard effort and its subsequent strategic reorientation is an example of the processes of strategy change described by Robert Burgelman, in "Intraorganizational Ecology of Strategy-Making and Organizational Adaptation: Theory and Field Research," Organization Science (2), 1991, 239-262, as essentially a process of natural selection through which suboptimal strategic initiatives lose out to optimal ones in the internal competition for corporate resources.
6. The failure of Micropolis to maintain simultaneous competitive commitments to both its established technology and the new 5.25-inch technology is consistent with the technological histories recounted by James Utterback, in Mastering the Dynamics of Innovation (Boston: Harvard Business School Press, 1994). Utterback found that firms that attempted to develop radically new technology almost always tried to maintain simultaneous commitment to the old and that they almost always failed.
7. A set of industries in which disruptive technologies are believed to have played a role in toppling leading firms is presented by Richard S. Rosenbloom and Clayton M. Christensen in "Technological Discontinuities, Organizational Capabilities, and Strategic Commitments," Industrial and Corporate Change (3), 1994, 655-685.
8. In the 1990s, DEC finally set up a Personal Computer Division in its attempt to build a significant personal computer business. It was not as autonomous from DEC's mainstream business; however, the Quantum and Control Data spin-outs were. Although DEC set up specific performance metrics for the PC division, it was still held, de facto, to corporate standards for gross margins and revenue growth.
Responsibility for Disruptive Technologies \ 1 2 3
9. "Harvard Study on Discount Shoppers," Discount Merchandiser, September, 1963, 71.
10. When this book was being written, Kmart was a crippled company, having been beaten in a game of strategy and operational excellence by WalMart.
Nonetheless, during the preceding two decades, Kmart had been a highly successful retailer, creating extraordinary value for Kresge shareholders.
Kmart's present competitive struggles are unrelated to Kresge's strategy in meeting the original disruptive threat of discounting.
11. A detailed contrast between the Woolworth and Kresge approaches to discount retailing can be found in the Harvard Business School teaching case.
"The Discount Retailing Revolution in America," No. 695-081.
12. See Robert Drew-Bear, "S. S. Kresge's Kmarts," Mass Merchandising: Revolution and Evolution (New York: Fairchild Publications, 1970), 218.
13. F. W. Woolworth Company Annual Report, 1981, p. 8.
14. "Woolco Gets Lion's Share of New Space," Chain Store Age, November, 1972, E27. This was an extraordinarily elegant, rational argument for the consolidation, clearly crafted by a corporate spin-doctor extraordinaire.
Never mind that no Woolworth stores approached 100,000 square feet in size!
15. See, for example, "The Desktop Printer Industry in 1990," Harvard Business School, Case No. 9-390-173.
16. Business historian Richard Tedlow noted that the same dilemma had confronted A&P's executives as they deliberated whether to adopt the disruptive supermarket retailing format:
The supermarket entrepreneurs competed against A&P not by doing better what A&P was the best company in the world at doing, but by doing something that A&P did not want to do at all. The greatest entrepreneurial failure in this story is Kroger. This company was second in the market, and one of its own employees (who left to found the world's first supermarket) knew how to make it first. Kroger executives did not listen. Perhaps it was lack of imagination or perhaps, like the executives at A&P, those at Kroger also had too much invested in the standard way of doing business. If the executives at A&P endorsed the supermarket revolution, they were ruining their own distribution system.
That is why they sat by paralyzed, unable to act until it was almost too late. In the end, A&P had little choice. The company could ruin its own system, or see others do it.
See Richard Tedlow, New and Improved: The Story of Mass Marketing in America (Boston: Harvard Business School Press, 1996).
17. See, for example, James M. Utterback and William J. Abernathy, "A Dynamic
124 | MANAGING DISRUPTIVE TECHNOLOGICAL CHANGE
Model of Process and Product Innovation," Omega (33:6), 1975, 639-656; and Clayton M. Christensen, Fernando F. Suarez, and James M. Utterback,
"Strategies for Survival in Fast-Changing Industries," Harvard Business School working paper, 1996.
18. The notion of lightweight and heavyweight project teams was originally introduced by Robert H. Hayes, Steven C. Wheelwright, and Kim B. Clark,
"Managing Product and Process Development Projects," chapter 11 of Dynamic Manufacturing: Creating the Learning Organization (New York: The Free Press, 1988). The concept has been amplified by Steven C. Wheelwright and Kim B. Clark, in chapter 8 of Revolutionizing Product Development (New York: The Free Press, 1992), and in Kim B. Clark and Steven C.
Wheelwright, "Organizing and Leading 'Heavyweight' Development Teams,"
California Management Review (34), Spring, 1992, 9-28. These authors note that in general, breakthrough and platform-type projects call for heavyweight project management, whereas derivative projects can be handled by lightweight teams. However, "Hospital Equipment Corporation," Harvard Business School, Case No. 697-086, illustrates a more fundamental principle: It demonstrates that lightweight managers can effectively run projects, even technologically complex platform projects, as long as the individuals and groups within the organization know how to work together because new modules can be plugged and played in the established architectural system.
It also demonstrates that even technologically straightforward projects may require heavyweight teams when individuals and groups within the organization are required to work together in different ways, on different problems, and at different times in the project than had been customary.
19. In advocating the setting up or spinning out of an organizationally independent business unit to focus on commercializing a disruptive technology, I do not mean in any way to present this as a managerially simple panacea to a difficult problem. The challenge of "corporate venturing" is itself a field that has spawned substantial scholarship. See, for example, Robert Burgelman and Leonard Sayles, Inside Corporate Innovation (New York: The Free Press, 1986); and Zenas Block and Ian MacMillan, Corporate Venturing (Boston: Harvard Business School Press, 1993).
20. The idea that one of the most critical jobs of a manager is to define and create the appropriate context within which an organization's processes can function effectively has been articulated by Robert Burgelman in "A Model of the Interaction of Strategic Behavior, Corporate Context, and the Concept of Strategy," Academy of Management Review (3), 1983, 61-69; and in Edward H. Wrapp, "Good Managers Don't Make Policy Decisions," Harvard Business Review, September-October, 1967.
C H A P T E R S I X
Match the Size of the
Organization to the Size
of the Market
Managers who confront disruptive technological change must be leaders, not followers, in commercializing disruptive technologies. Doing so requires implanting the projects that are to develop such technologies in commercial organizations that match in size the market they are to address. These assertions are based on two key findings of this study: that leadership is more crucial in coping with disruptive technologies than with sustaining ones, and that small, emerging markets cannot solve the near-term growth and profit requirements of large companies.
The evidence from the disk drive industry shows that creating new markets is significantly less risky and more rewarding than entering established markets against entrenched competition. But as companies become larger and more successful, it becomes even more difficult to enter emerging markets early enough. Because growing companies need to add increasingly large chunks of new revenue each year just to maintain their desired rate of growth, it becomes less and less possible that small markets can be viable as vehicles through which to find these chunks of revenue. As we shall see, the most straightforward way of confronting this difficulty is to implant projects aimed at commercializing disruptive technologies in organizations small enough to get excited about small-market opportunities, and to do so on a regular basis even while the mainstream company is growing.
1 2 6 | MANAGING DISRUPTIVE TECHNOLOGICAL CHANGE
A R E T H E P I O N E E R S REALLY T H E O N E S
W I T H A R R O W S I N T H E I R B A C K S ?
A crucial strategic decision in the management of innovation is whether it is important to be a leader or acceptable to be a follower. Volumes have been written on first-mover advantages, and an offsetting amount on the wisdom of waiting until the innovation's major risks have been resolved by the pioneering firms. "You can always tell who the pioneers were,"
an old management adage goes. "They're the ones with the arrows in their backs." As with most disagreements in management theory, neither position is always right. Indeed, some findings from the study of the disk drive industry give some insight into when leadership is critical and when followership makes better sense.
Leadership in Sustaining Technologies May Not Be Essential One of the watershed technologies affecting the pace at which disk drive makers have increased the recording density of their drives was the thin-film read/write head. We saw in chapter 1 that despite the radically different, competence-destroying character of the technology, the $ 1 0 0 million and five-to-fifteen year expense of developing it, the firms that led in this technology were the leading, established disk drive manufacturers.
Because of the risk involved in the technology's development and its potential importance to the industry, the trade press began speculating in the late 1 9 7 0 s about which competitor would lead with thin-film heads.
H o w far might conventional ferrite head technology be pushed? Would any drive makers get squeezed out of the industry race because they placed a late or wrong bet on the new head technology? Yet, it turned out, whether a firm led or followed in this innovation did not make a substantial difference in its competitive position. This is illustrated in Figures 6.1
and 6 . 2 .
Figure 6.1 shows when each of the leading firms introduced its first model employing thin-film head technology. The vertical axis measures the recording density of the drive. The bottom end of the line for each firm denotes the maximum recording density it had achieved before it introduced a model with a thin-film head. The top end of each line indicates the density of the first model each company introduced with a thin-film head. Notice the wide disparity in the points at which the firms felt it was important to introduce the new technology. I B M led the industry,
Match the Size of the Organization to the Size of the Market | 1 2 7
Figure 6.1 Points at Which Thin-Film Technology Was Adopted by Leading Manufacturers, Relative to the Capabilities of Ferrite/Oxide Technology at the Time of the Switch
Source: D a t a are f r o m v a r i o u s i s s u e s of Disk/Trend Report.
1 2 8 | MANAGING DISRUPTIVE TECHNOLOGICAL CHANGE
introducing its new head when it had achieved 3 megabits (Mb) per square inch. M e m o r e x and Storage Technology similarly took a leadership posture with respect to this technology. At the other end, Fujitsu and Hitachi pushed the performance of conventional ferrite heads nearly ten times beyond the point where I B M first introduced the technology, choosing to be followers, rather than leaders, in thin-film technology.
W h a t benefit, if any, did leadership in this technology give to the pioneers? There is no evidence that the leaders gained any significant competitive advantage over the followers; none of the firms that pioneered thin-film technology gained significant market share on that account. In addition, pioneering firms appear not to have developed any sort of learning advantage enabling them to leverage their early lead to attain higher levels of density than did followers. Evidence of this is displayed in Figure 6 . 2 . The horizontal axis shows the order in which the firms adopted thin-film heads. Hence, I B M was the first, M e m o r e x , the second, and Fujitsu the fifteenth. The vertical axis gives the rank ordering of the recording density of the most advanced model marketed by each firm in 1 9 8 9 . If the early adopters of thin-film heads enjoyed some sort of experience-based advantage over the late adopters, then we would expect the points in the chart to slope generally from the upper left toward the lower right. The chart shows instead that there is no relationship between leadership and followership in thin-film heads and any subsequent technological edge.1
Each of the other sustaining technologies in the industry's history present a similar picture. There is no evidence that any of the leaders in developing and adopting sustaining technologies developed a discernible competitive advantage over the followers.2
Leadership in Disruptive Technologies Creates Enormous Value In contrast to the evidence that leadership in sustaining technologies has historically conferred little advantage on the pioneering disk drive firms, there is strong evidence that leadership in disruptive technology has been very important. T h e companies that entered the new value networks enabled by disruptive generations of disk drives within the first two years after those drives appeared were six times more likely to succeed than those that entered later.
Eighty-three companies entered the U.S. disk drive industry between 1 9 7 6 and 1 9 9 3 . Thirty-five of these were diversified concerns, such as
Match the Size of the Organization to the Size of the Market | 1 2 9
Figure 6.2 Relationship between Order of Adoption of Thin-Film Technology and Areal Density of Highest-Performance 1989 Model
Source: C l a y t o n M. C h r i s t e n s e n , " E x p l o r i n g t h e L i m i t s of t h e T e c h n o l o g y S - C u r v e . P a r t I: C o m p o n e n t T e c h n o l o g i e s , " Production and Operations Management 1, n o . 4 (Fall 1992): 3 4 7 . R e p r i n t e d by p e r m i s s i o n .
Memorex, Ampex, 3 M , and X e r o x , that made other computer peripheral equipment or other magnetic recording products. Forty-eight were independent startup companies, many being financed by venture capital and headed by people who previously had worked for other firms in the industry. These numbers represent the complete census of all firms that ever were incorporated and/or were known to have announced the design of a hard drive, whether or not they actually sold any. It is not a statistical sample of firms that might be biased in favor or against any type of firm.
The entry strategies employed by each of these firms can be characterized along the two axes in Table 6 . 1 . T h e vertical axis describes technology strategies, with firms at the bottom using only proven technologies in their
Table 6.1 Disk Drive Companies Achieving $100 Million in Annual Revenues in at Least One Year Between 1976 and 1994
New
%
Sales
%
Sales
Technology
Type of Firm
S
F
N
T
Success ($millions)
Type of Firm
S
F
N
T
Success ($millions)
Start-ups
0
7 3
10
0%
$235.3
Start-ups
3
4
1
8
37%
$16,379.3
Related-Technology
0
1 0
1
0%
0.0
Related-Technology
0
0
0
0
—
—
Related-Market
0
3
0
3
0%
1.4
Related Market
0
0
0
0
—
Forward Integrators
0
1
0
1
0%
0.0
Forward Integrators
0
0
0
0
—
Technology
Total
0
12 3
15
0%
$236.7
Total
3
4
1
8
37%
S 16,3/9 3
Strategy