ALIGNING TECHNOLOGICAL AND ORGANIZATIONAL CHANGE:WHY THE TOPIC IS CRITICAL TO INDUSTRIAL ENGINEERS

1. WHY THE TOPIC IS CRITICAL TO INDUSTRIAL ENGINEERS

Failures of Implementation of New Technology

Accumulated evidence indicates that the implementation of computer-automated technology has not achieved as much success as originally anticipated. The American Production and Inventory Control Society and the Organization for Industrial Research have estimated the failure rate of these tech- nologies to be as high as 75% (Works 1987). In a study in which 55 managers in 41 organizations supplying or using CAM were interviewed, half of the CAM installations were reported as failures (Ettlie 1986). In a study of 95 flexible manufacturing systems in the United States and Japan, the FMSs in the United States were found to be so ineffectively used as to yield little of the flexibility had been achieved in Japan (Jaikumar 1986). Kalb (1987) reports a 30–70% failure rate of comput- erized manufacturing technologies. One new product-development manager of a large computer man- ufacturer reported that ‘‘Inadequately implemented new technologies cost our plants up to $1 million a day in unexpected losses.’’ A major study of 2000 U.S. firms that had implemented new office systems revealed that at least 40% of these systems failed to achieve the intended results (Long 1989). Gibbs (1994) reports that for every six new large-scale software systems that are put into operation, two others are cancelled, with the average software development project overshooting its schedule by half. The Standish Group in 1995 reported that only 16% of information systems projects were judged to be successful, with 31% outright cancelled (Wall Street Journal 1998b). The Standish Group conducted another survey in 1997 of 360 information system professionals and found that 42% of corporate information technology projects were abandoned before completion and 33% were over budget or late (Computerworld 1997).

Examples of these failures abound. For example, after spending more than $17 million on a long- anticipated overhaul of Los Angeles’s computerized payroll system, the city controller scrapped it (Los Angeles Times 1999). The London Ambulance Service computer-aided dispatch system deployed in 1992 was intended to provide an automatic vehicle-locating system, telephone call processing, and allocation buttons for crew to report on current status. The system was pulled because crews couldn’t accurately indicate their status and dispatchers couldn’t intervene to get the crews to the needed locations (Flowers 1997). The State of California cancelled deployment of an automated child-support system for automatically tracking parents across counties who do not have primary custody of their children after spending $100 million (Los Angeles Times 1997). Fox Meyer, once a $5 billion drug- distribution company, was unable to process the huge volume of orders from pharmacies after in- stalling a $65 million ERP system. As a result, it filed for bankruptcy in 1996, was bought in 1997 for just $80 million, and filed a $500 million lawsuit against Andersen Consulting, the implementers of the ERP system (Information Week 1998; Wall Street Journal 1998b). Oxford Health Plans lost $363 million in 1997 when their new claims-processing system delayed claims processing and client billing (Wall Street Journal 1998a; Champy 1998). Computer systems were blamed for delaying the scheduled opening of the first deregulated electricity market in the United States (Information Week 1997). Hershey, the nation’s largest candy maker, installed a $110 million ERP system in July 1999. Glitches in the system left many distributors and retailers with empty candy shelves in the season leading up to Halloween (Wall Street Journal 1999). Whirlpool reported that problems with a new ERP system and a high volume of orders combined to delay shipments of appliances to many distributors and retailers.

These failures are expensive. In an internal document of September 15, 1997, the information systems research firm MetaFax calculated an average yearly loss of $80 billion from a 30% cancel- lation rate and a $59 billion loss from a 50% over-budget rate. In 1997 alone (before the Y2K inflated IT expenditures), companies spent $250 billion on information technology; a 30–70% failure rate clearly means that billions of dollars are spent with disappointing results (Wall Street Journal 1998b). Aside from a disappointing return on investment, the impacts of failed technology investments in- clude:

• Harm to the firm’s reputation (where poor implementation gets blamed on the technology vendor or designer)

• Broken trust (where workers are unwilling to go the extra mile the next time)

• Reduced management credibility (because management can’t deliver on promises)

• Slower learning curve (leading to crisis management as problems increase with implementation rather than decrease)

• Reduced improvement trajectory (since there is no time to explore opportunities for new tech- nology or new business opportunities for existing technology)

Why These High Failure Rates?

In one of the first major studies on this problem of implementation, the Congressional Office of Technology Assessment concluded: ‘‘The main stumbling blocks in the near future for implemen- tation of programmable automation technology are not technical, but rather are barriers of cost, organization of the factory, availability of appropriate skills, and social effects of the technologies’’ (OTA 1984, p. 94). A few years later, the Manufacturing Studies Board of the National Research Council conducted a study of 24 cases of the implementation of CAM and CIM technologies and concluded: ‘‘Realizing the full benefits of these technologies will require systematic change in the management of people and machines including planning, plant culture, plant organizations, job de- sign, compensation, selection and training, and labor management relations’’ (MSB 1986). In a 1986 Yankee Consulting Group marketing survey of CAM and CIM users, the users reported that 75% of the difficulties they experienced with the technologies could be attributable to issues concerned with planning the use of the technology within the context of the organization (Criswell 1988).

Recent evidence continues to support the conclusion that a significant component of the com- plexity of technological change lies in the organizational changes often experienced. C. Jackson Grayson, Jr., then Chairman of the American Productivity and Quality Center in Houston, Texas, analyzed the 68 applications for the Malcolm Baldrige National Quality Award for 1988 and 1989 and found that a major reason for failing to meet the examination criteria was the neglect of and failure to integrate human and organizational aspects with technology investments (Grayson 1990). Peter Unterweger of the UAW Research Department, after extensive case study visits in the United States and abroad, concluded that the successes of technological implications can be attributable to:

(a) hardware playing a subordinate role to organizational or human factors and (b) developing the technical and organizational systems in step with one another (Unterweger 1988). In a study of 2000 U.S. firms implementing new office systems, less than 10% of the failures were attributed to technical failures; the majority of the reasons given were human and organizational in nature (Long 1989). The MIT Commission on Industrial Productivity concluded from their extensive examination of the competitiveness of different American industries: ‘‘Reorganization and effective integration of human resources and changing technologies within companies is the principal driving force for future pro- ductivity growth’’ (Dertouzos et al. 1989). More recently, in a 1997 survey by the Standish Group of 365 IT executive managers, the top factors identified in application development project failures were poor management of requirements and user inputs (Computerworld 1998a). The 1997 MetaFax survey found the reasons for IS failures to include poor project planning and management. In a 1998 Computerworld survey of 365 IT executives, the top factors for software development project failures were the lack of user input and changing requirements (Computerworld 1998a). A careful study of six failures of information technology projects found that those projects that devoted more effort to the technology rather than to the organizational issues (such as awareness, training, and changes to organizational procedures) were more likely to fail (Flowers 1997). In short, these failures can be attributed to the inadequate integration of technical with social and organizational factors during the introduction of the technological change, called sociotechnical or TOP (for Technology, Organization, and People) integration. This recognition has led The Wall Street Journal to write: ‘‘What’s emerging here is a search for a better balance between manpower and computer power’’ (1998b, p. 1).

Several cases of failures directly attributable to poor alignment of technology and organizational change can be cited. In one such example (Ciborra and Schneider 1990), a U.S. aircraft instruments plant implemented a computerized MRP system. Ten months into the implementation process, none of the expected gains in efficiency had materialized, despite clearly defined goals and plans, a sound economic evaluation, and a structured implementation plan. The major problem was that there was so much emphasis on following the rules created by the MRP system that clerks often hesitated to override the system’s commands even when they knew that the commands did not make sense. Even useful localized innovations with the system, such as shortcuts and new rules of thumb, remained private know-how because localized practices were not sanctioned by management. Learning from mistakes was limited because effective job performance for the system designers was measured by adherence to best technical practice, not to shop-floor reality, and thus the system designers were not willing to have their competence questioned.

As another example, in 1997 Chrysler Financial tossed out a sophisticated financial package bought for the company’s financial team. The problem was that the system was incompatible with the company’s e-mail system. So the company adopted a less sophisticated approach that was more closely aligned with the way the financial staffers worked: instead of monitoring dealer activity with a 100% computerized system, the company instructed clerks to obtain information from dealers the old-fashioned way—over the phone—and enter the information quickly to make it available to fi- nancial staffers who wanted to know which dealers were moving a lot of cars or taking bad loans. According to the project director, the purely computerized solution would have cost many millions of dollars more and taken years to install, but ‘‘by adding some people into the equation, we could get 95% of what we needed’’ and take only 90 days to set it up (Wall Street Journal 1999, p. A26).

Another example of how advanced technology without correct organizational alignment in the automotive industry failed is presented by The Economist:

[T]he giant Hamtramck plant in Detroit, which makes Cadillacs, is just five years old and heavily automated but ranks among the least competitive plants in the United States. Hamtramck is typical of GM’s early efforts to beat the Japanese by throwing truckloads of cash into a new technology. Hamtramck had what is politely called a ‘‘very rough start-up’’. Its robots ran wild. Although the problems have now largely been tamed, GM learnt in a joint venture with Toyota that what really mattered in manufacturing was people. (Economist 1990).

As another example, British Airways put in a system at airport gates in which the screen was mounted horizontally, low on a desktop. Ticket agents looked down when checking in passengers; as a result, passengers saw only the top of the agent’s head. The consultant on the project reported that they did this deliberately so there would be less eye contact and less schmoozing and the lines would be shorter. However, after installation, passengers complained; apparently fliers are naturally anxious and often need a little schmoozing, according to the consultant. The airline moved the screens to eye level (Computerworld 1998b).

Similarly, according to a survey of the artificial intelligence industry by The Economist, blind introduction of computers in the workplace by an American airline (which prefers to remain nameless) proved that people do not like taking orders from a machine when an expert system was installed to schedule the work of maintenance engineers (Economist 1992). The engineers simply rejected the system’s plans and the computer system had to be withdrawn. But when, after suitable delay, the airline reintroduced more or less the same system for engineers to use when and if they wanted, it was much better received.

A final example of a project devoting too much attention to the technology side and too little to the organizational side is the London Ambulance system failure. In the formal inquiry on the failure, it was noted that the initial concept of the system was to fully automate ambulance dispatching; however, management clearly underestimated the difficulties involved in changing the deeply in- grained culture of London Ambulance and misjudged the industrial relations climate so that staff were alienated to the changes rather than brought on board. (Flowers 1997).

While much of this information supporting the important role of aligning technology and orga- nizations is anecdotal, there have been several econometric studies of larger samples supporting this claim. A growing body of literature has established strong empirical links among such practices as high-involvement work practices, new technologies, and improved economic performance (MacDuffie 1995; Arthur 1992). Pil and MacDuffie (1996) examined the adoption of high-involvement work practices over a five-year period in 43 automobile assembly plants located around the world, their

technologies (ranging from highly flexible to rigidly integrated), and their economic performance and found that the level of complementary human resource practices and technology was a key driver of successful introduction of high-involvement practices. Kelley (1996) conducted a survey of 973 plants manufacturing metal products and found that a participative bureaucracy (i.e., group-based employee participation that provides opportunities to reexamine old routines) is complementary to the produc- tive use of information technology in the machining process. Osterman (1994) used data on 694 U.S. manufacturing establishments to examine the incidence of innovative work practices, defined as the use of teams, job rotation, quality circles, and total quality management. He found that having a technology that requires high levels of skills was one factor that led to the increased use of these innovative work practices.

To conclude, it should be clear that technological change often necessitates some organizational change. If both organizational and technological changes are not effectively integrated and managed to achieve alignment, the technological change will fail.