Market Management to Transform the IT Organization

In recent years, there have been calls and prescriptions for business transformation.1 There has also been growing pressure for radical change in how the IT function is organized and managed. Rockart et al. captured the consequences of an agenda for reform in the IT function, and Cross et al. proposed a possible model of a transformed IT organization.2 However, there have been no robust models or frameworks developed to guide executives in the process of transformation, that is, how to do it. We suggest that this is an important practical omission, not only because transformation of the IT organization is on many business agendas, but because it is an area that is probably always undergoing change.

Another reason for developing a process model for transformation is that organizations often seem to recognize crises in IT management, but their responses frequently only worsen the crisis or achieve little until the next crisis. Consider the scenario we witnessed in a well-known multinational manufacturing corporation:

In the 1980s, a large corporate IT department built a group infrastructure with common transaction processing systems for the company worldwide. In terms of best practice, technical robustness, and economies of scale, there was little to criticize. However, not unusually, the business units claimed that the IT department was not responsive to business needs, and that the common systems were a lowest common denominator solution that never quite matched local requirements. In response, the IT department personnel said that, in the absence of good business-led information systems strategies, they had been forced to provide the lead and that the business units did not recognize the value of a standard architecture and common systems in enabling business across divisions and by regions. When the whole company went through rapid decentralization at the end of the 1980s, the corporate IT department was downsized and the business units were given full responsibility for most aspects of IT.

Soon the units were identifying their own systems requirements but had no one to supply their needs and little experience in commissioning such work. Furthermore, they often did not want to be responsible for running an IT department. The company then decided to create an information services business from the old corporate IT department, which would be a profit center. It could supply services inside or outside the parent company. Not long afterward, the new information services business decided it did not have enough critical mass in the marketplace and merged with another IT services company.

By the mid-1990s, top management sensed that the business units were not developing the information systems they required and, in buying systems development from the information services business, they were not giving enough thought to the group’s priorities. Corporate management instigated a moratorium on systems development, while it assembled a task force to review IT management in the group.

The moral of this case history is threefold. First, u-turns in IT management rarely succeed; they only consume resources and executive attention as a new scheme is implemented, leaving behind the same problems. Second, management of IT is concerned with two fundamental issues: the supply side, namely, the provision of IT operations, systems development, and user support, and the demand side, namely, the identification and prioritization of application requirements and of opportunities to exploit emerging technologies. Planning and delivering are a distinguishing characteristic of the IT function. Third, firms get into trouble when they concentrate only on supply or only on demand. In the case history just described, the decentralization phase emphasized demand and overlooked supply. The outsourcing phase emphasized supply and overlooked demand. The moratorium and review phase has returned to the demand side.

Because both supply and demand have to be managed, we refer to the process as market management. Drawing on a longitudinal case study we have been researching for nearly five years, we have conceptualized a four-stage model in which the balance of attention allocated to supply or demand shifts over time, but in which neither side is ever ignored. The case study company, BP Exploration (BPX), transformed its IT organization during a six-year period.3 We develop and explain the process of market management in the next section, building on BPX’s experiences (see Figure 1).

Market Management

At stage one, “recognizing disequilibrium,” a company first articulates, explores, and analyzes the crisis or loss of confidence in IT. We posit that normally there are both supply and demand problems. For example, the IT organization’s delivery performance is causing anxiety, or there is a perception that IT is not keeping up with best practice and new technologies. In short, the supply side is not well managed. At the same time, the demand side is causing concern. There is a sense that the business is not developing and using the systems it needs or that general and line management are not leading the IT effort as required. To pursue the economic analogies further, the business is no longer operating efficiently or making optimal decisions because what is both technologically feasible and desirable from a business viewpoint has shifted. To understand this change, the company examines the supply and demand sides and takes early initiatives to attack both sides of the problem, but does not tackle one at the expense of the other. This stage is typified by questioning the value of IT and the IT function.

At stage two, “emphasizing supply management,” the company seeks radical performance improvement of the supply side by setting new delivery goals and beginning to rebuild the technology platform. At the same time, it audits current and planned applications to prioritize and rationalize systems demand. This stage focuses on releasing and realizing value from the inherited IT situation.

At stage three, “emphasizing demand management,” the focus shifts from supply to demand, but not exclusively. The company sets a more visionary direction for applications development and installs continuous self-sustaining processes and policies for demand management or systems planning. On the supply side, it moves beyond just a focus on performance to planning and building the future infrastructure. This stage is concerned with building IT capabilities and creating future value.

At stage four, “maintaining equilibrium,” the company completes the transformation process by implementing final radical changes in both demand and supply sides. However, the most important task at this stage is for both the business and the IT function to recognize that IT management requires continual improvement. So, on the supply side, the company addresses governance of the new IT organization and puts new skills in place. On the demand side, the company develops partnerships between IT and both business managers and external suppliers in order to henceforth identify new application needs and relevant emerging technologies. This stage is concerned, therefore, with sustaining value.

Transformation at BPX

BPX, a $13 billion division of British Petroleum (BP), explores for and produces oil and gas. Like many oil companies in the mid-1980s, BP expected the price of crude oil to continue rising. Projections had forecast $100 per barrel by the end of the century, and company business plans had been based on $30 per barrel. Thus BP had established an expansion campaign to take advantage of this apparently benign environment, growing in value by 25 percent between 1986 and 1989. It achieved this growth mainly by acquiring Standard Oil, Britoil, and Lear Petroleum. BP left these acquisitions largely autonomous because the primary managerial focus was on growth through the discovery of new fields.

Unfortunately, the plan received a major blow when the price of crude oil plummeted to $14 per barrel. The CEO of BPX, John Browne, who had been appointed in May 1989, decided that radical change was the only way to survive. He envisaged a flatter, global, yet responsive organization capable of competing in a commodity marketplace. He also realized the need for a substantially new, lower cost structure.

Such a change program had massive implications for the information technology function of BPX, known as XIT. It would have to be part of the radical transformation that Browne was orchestrating, both changing policies and practices as every other business unit was doing and supporting the company’s new modus operandi with appropriate applications and technologies. By the end of the 1980s, IT itself was changing dramatically. Computer hardware shifted from mainframe computing to client-server architecture. The software industry was developing more off-the-shelf application packages. Also, the maturing information services marketplace meant that it was possible to outsource IT.

With these challenges in the background, XIT set out, during a six-year period, to redesign and reinvent itself (see Table 1).4 But how does a company start the process of transformation and then sustain it? We believe that the BPX example clearly shows the four stages in the process.

Stage One: Recognizing Disequilibrium

Crises are not unusual in IT management. Indeed, Gibson and Nolan and Nolan based a stages theory of DP management on the concept of anticipating and smoothing the crises that were a necessary outcome of experiential learning.5 Here, we suggest that when there is a loss of organizational confidence in IT, it is likely due to both demand-side and supply-side problems.

At BPX, there was a business or demand discontinuity: the business had to reinvent itself in order to operate with prices of $14 per barrel (or less) rather than $30 per barrel. Browne saw two complementary challenges for the IT organization. First, it had to reduce its own cost base to realign the function with a new cost structure in the business. Second, the application of IT had to enable a new, flatter, global, and responsive organization.

However, there was also a supply-side discontinuity. Browne felt that the IT function had not been serving the business well. XIT had been building state-of-the-art applications that satisfied the IT professionals in terms of sophistication but delivered little competitive advantage to the business. Indeed, Browne felt that the company could buy many applications from the marketplace and that if it didn’t outsource, it was giving away competitive advantage.

Once a company recognizes demand and supply issues, the business and the IT function share the transformation or turnaround problem. Browne’s next step, therefore, was to get senior executives to recognize what they probably knew already — that the old way of doing business, including the management of IT, was no longer valid. In June 1989, Browne held a meeting of the top 150 managers to deliver two clear, dramatic messages.

1. Browne wanted merged businesses to reduce total costs by 20 percent. BPX had gone through two significant mergers (and a smaller acquisition). Browne reasoned that at least a 40 percent reduction in costs was possible.
2. Browne announced that he was the only person in the room with a job. Those invited to another meeting in six weeks would have a job in the new organization (approximately fifty of the managers returned).

These messages reinforced the transformation of the IT function. Any radical change program in XIT was not insular but an essential part of the wider business transformation. Conversely, there could be no excuses for failing to achieve change. Browne believed that the IT function could be a laboratory and litmus test for the transformation that he and the rest of the organization could observe.

Browne’s next step was to appoint a new general manager for IT, John Cross, who had a track record in management, change, and IT. The two began to work closely together. They both sponsored a two-day workshop with 500 senior managers on IT’s capabilities and business impact. This was followed by a workshop with a new BPX management team to explore which areas of IT were strategic to the business, useful, or of no business value. They used a strategic grid derived from McFarlan et al.’s information systems planning framework.6 One manager remarked that the framework helped establish where, how, and why IT was important to the business and where it offered little benefit. The framework also helped Browne emphasize the need for reducing complexity, whereas he felt that IT had increased it.

In this sequence of events, there were four important factors: First, the business units and IT became owners and executors of assessment and prescription. Second, workshops and education internalized and solidified the change message. Third, frameworks and language helped in managing IT by providing a common language. Fourth, a strong CEO-CIO relationship was an asset in managing and, especially, transforming IT.7

One further critical initiative backed up the process investments. The company introduced a new budgeting and cost control system for IT. It evaluated project sponsors and leaders against cost targets to deter overruns. It centralized all IT costs under Cross not only to control duplication and redundancy, but also to identify costs and encourage cost justifications. As one IT manager observed, “The new budget had teeth!”

Once the company recognized the disequilibrium and accepted the need for and shape of transformation, the next, more aggressive stage began.

Stage Two: Emphasizing Supply Management

Cross quickly assembled a global management team for XIT. The team soon realized that many inefficiencies had been inherited, largely due to the recent acquisitions. Also, technological advances in desktop computing, client-server computing, and packaged application software suggested that further economies were possible. The implied mandate to save costs, given Browne’s challenge to the business as a whole, was not threatening. “Our initial budget was almost $360 million, and we set a target of saving $100 million in 1990,” recalled Cross. As a member of his team observed, “The first general direction for IT was to eliminate redundancy, rationalize, and consolidate.”

Simplification and retrenchment are often early tasks in models of business transformation. In this case, the team found many ways to effect dramatic cost savings. It was important to stop the hemorrhaging, so the team set $100 million as a target to approximate Browne’s call for a 20 percent reduction. It strongly signaled that things had to change quickly. Some of the “excess” savings could fund future redesign. Ambitious goals might encourage radical thinking.

“Saving $100 million was simple, not difficult,” according to Cross. The team consolidated two data centers into one in both the United Kingdom and the United States. It rationalized and integrated three networks. It retained only one supercomputer, with most high-capacity work transferred to Sun workstations. In addition, 600 personnel left the company, “even though there was only a rudimentary understanding of the current disposition of skills or the required capabilities in the future,” as a member of the XIT management team commented. During a three-year period, IT head count dropped from 1,400 to 390.

Cross then suggested a further $100 million savings in the next year. Between 1989 and 1992, the team delivered $460 million savings to the business “without any visible loss of value,” according to Cross. The shift from mainframe computing to a client-server platform helped. The percentage of MIPs (million instructions per second) on the desktop in 1989 was only 20 percent; by 1995, it was 99 percent. This shift not only took advantage of the downward sloping curve of costs per MIP, but also released savings from reduced support costs.

The team used the analysis at stage one to prune the application portfolio. In addition, it switched off any applications without sponsors. Such systems were terminated in batches of fifty. If anyone complained, the team assessed the business value of the application and decided whether to reinstate it. One application, a drilling data system, had run on a dedicated VAX, with drillers entering local data, for three years. No one used it, so the application was switched off, the VAX closed down, and the data input process discontinued. Between 1989 and 1992, the number of applications was cut by 35 percent, from 170 to 110.

When BPX reflected on this stage, it saw a period of releasing value from underdeployed, outdated, or unnecessary assets. However, this was not entirely a short-term tactical exercise. Simplification was also a starting point for rebuilding IT architecture to enable Browne’s vision of a global, interdependent, and flat organization. The move to a desktop orientation and client-server platform would enable common access to, and processing of, information. If BPX could rationalize data structures while coordinating and sharing activities in oil exploration and production, not only internally but with all the partners in the sector, cost savings would be useful and business process benefits high. So XIT worked with the business to advocate for and help found the Petrochemical Open Software Corporation (POSC).

POSC, an international nonprofit membership organization, started with 5 founding members in 1990 and grew to 115 by 1995. Its mission was to define, develop, and deliver an industry standard, open systems software integration platform. The effort initially concentrated on developing a common data model, followed by a focus on application development. By 1995, the POSC common data model had created estimated savings of $60 million to $100 million for BPX by reducing the conversion efforts necessary to move data across different functions, business units, and companies.

Coping with new applications development is a demand-side task that cannot be overlooked at stage two. Businesses do not stand still while an IT organization chooses to sort out the supply side; this was a problem in the multinational manufacturing corporation cited earlier. Indeed, requests for new systems exceeded the desired 1991 budget by 40 percent. The BPX IT management team could find only 15 percent in cutbacks by searching for technical efficiencies and synergies. So the team returned the requests to the application areas in the business, and asked the new regional management teams installed by Browne to justify and prioritize each application.

One particular application considered at this time was a new database package. As XIT began to realize the importance of business flexibility in its new environment, it sought to separate data from applications to facilitate information access and not constrain the economies of information processing and business responsiveness. Cross and his senior management team chose Sybase, then a small, relatively new, unproven company with only $20 million in sales. Various parts of XIT and the business resisted the decision. However, Cross and his team felt that this software was key to breaking down the “functional strategies” in their own function, as well as in the business. In an unusual move, Cross took this policy decision on a relational database management system to the board and explained Sybase’s potential advantages, as well as its costs and risks. In the end, the company selected Sybase, but, more importantly, it set a tone and philosophy at the most senior levels of balancing IT and business benefits and risks.

In addition to pursuing ambitious performance targets, starting to rebuild the technology platform, and prioritizing application development, two less tangible factors are important at stage two: (1) sending substantive signals about the change, which have more impact if early wins are evident; and (2) building a cohesive IT management team and working with business management teams on business issues.

Stage Three: Emphasizing Demand Management

At stage three, BPX focused more on demand than supply, in particular, designing and implementing demand management processes and turning its attention to value creation rather than value realization. Demand management involved three issues: understanding current and future business directions and needs, assessing where new technologies and ideas could provide competitive advantage for BP, and planning and prioritizing application requirements in a routine, coherent manner. Creating value involved investment in systems, information, and processes that would yield future business returns.

By 1994, BPX was a federation of “assets” or business units with management teams measured by contribution and added value. Cross made the units responsible for their own demand management, in particular, identifying opportunities to create value, prioritizing necessary IT support and applications on the basis of value added, agreeing on IT budgets with the global XIT team, and specifying service levels on a cost-benefit basis. An IT executive helped the units with these tasks. In addition, the global XIT group developed consultancy skills to help the units’ management teams. The company retrained former systems analysts as business consultants.

This shift of orientation to the demand side forced XIT to rethink its purpose and develop a set of value propositions. The XIT management team developed, articulated, and agreed on a value model to guide policies, practices, and resource allocation. Known as the “IT value chain,” it explicated the differences between creating value and realizing value (see Figure 2). In particular, the team saw value-adding potential in reengineering business processes and exploiting information, both firm-specific capabilities. Thus most XIT personnel were to be deployed in these areas, once retrained. Applications and infrastructure were more a commodity, often available in the marketplace; this is where the team pursued efficiency and low cost through outsourcing.

At stage two, XIT had experimented to determine whether outsourcing worked in practice and how to manage it. Now it moved to widespread outsourcing of applications development and infrastructure. It initiated a policy of buy, not build, in software development and chose outsourcing vendors to work together, in basic infrastructure operations and support.8 Outsourcing was not the grand knee-jerk, cost-saving exercise that might be expected in a supply-side turnaround. It was a considered strategy aimed at building in-house capability for value creation (and demand management) and exploiting a maturing market for value realization (and supply management). As we shall see later, the company subsequently brought third-party vendors into the demand side of the transformation.

At this time, BPX also reinforced the budgetary control introduced earlier. There was only one centralized IT budget, and Cross’s signature was required for any expenditure of more than $100. This may seem unusual, given the distribution of IT resources and decentralization of IT management that was typical in the early 1990s. However, it not only facilitated the identification, monitoring, and control of costs, but forced prioritization of applications development and enabled the XIT management team to ensure it invested in value-adding activities. With a decentralized IT budget, 70 percent of applications development had been in back-office systems; after centralizing the IT budget, 70 percent of applications development was in support of business operations, including exploration data analysis and seismic modeling. One force for change was an XIT initiative known as 20/20, which aimed to select and fund IT projects that would reduce costs by 20 percent and increase quality by 20 percent or more.

The company did not ignore the supply side. It further developed the directions, which had been initiated at stage two. The emergence of a common platform, the discipline on technical solutions enabled by the central budget, and the POSC initiative all raised the importance and visibility of infrastructure planning and control. Browne had asked for “a unified platform between the IT function and the business, so that we can change business practices easily”; and the value chain emphasis on information exploitation and process innovation implied that information should be shared and exchanged freely. So XIT concluded that its primary purpose should be designing, planning, and overseeing an information infrastructure. As a result, the company added a supply management responsibility — conformance with global infrastructure standards — to the responsibility of demand management allocated to the business units.

At stage three, a company invests in demand management and refines some of the supply management directions set in stage two. The business and the IT organization are able to agree on value propositions for IT and formulate new policies to pursue them.

Stage Four: Maintaining Equilibrium

The vital fourth stage is concerned with sustaining value from IT and ensuring that both the demand and supply sides are well managed and in equilibrium. Of course, problems and issues may arise as business and technology change and the company makes occasional mistakes. One BP executive commented, “There is always another hill to climb.” By late 1994, XIT described the transformation as a journey and issued a map of successive summits to climb. This helped management understand the concept of continual change (and avoidance of major crises). Four sets of actions typified the first two years of this stage:

1. Creating an adaptive, learning organization. Within XIT, departments became affinity groups, and managers became group leaders. Individuals selected their own coaches and had responsibility for selecting projects and tasks and defining their own roles. Managers no longer controlled the resources or worked in a prescribed organization.

These changes helped break down the functional silos inherent in traditional IT departments. IT personnel managed their own careers and arranged for the development of new skills and competencies. XIT also defined the new skills expected of personnel, emphasizing a balance among business, technical, and people skills. It removed job titles to create a consulting organization in which people would not cling to specialist roles and worry about lines of responsibility. The goal was to foster an attitude of willingness to do whatever it takes to get the job done.

2. Browne and Cross’s frequent sharing of ideas. They admit that many ideas did not work, but part of the new philosophy at BPX was “to try lots of new things as pilots and only keep what works,” according to Browne.

3. XIT’s responsibility for initiating innovative and revolutionary investments in IT. Most of the units’ ideas for exploiting IT would support ongoing business and were somewhat evolutionary. So Browne and Cross sponsored a radically new project, the “virtual team” project providing desk-to-desk videocon-ferencing and information exchange using the latest technologies.

4. Examining governance structures and establishing a global management council for IT. Cross suggested that the council’s job was to prevent XIT from going too far in cutting costs. In a sense, the council maintained a balance between supply and demand.

Crucial in these four actions was clarification of the responsibilities and expectations of the different stakeholder groups associated with IT. Maintaining equilibrium requires all parties to work together and be clear that they share responsibility for IT’s success. While it is fashionable to blame IT for all problems, in many ways, businesses get the IT they deserve. On the other hand, the IT organization usually has to design and nurture the processes required to achieve effective management of IT’s supply and demand sides.

Thus, a critical success factor at stage four, although perhaps a cliché, is partnership. In BPX’s case, one type of partnership is between the IT function and the business, so that the business manages demand but is aided by XIT, and XIT manages supply but is supported by the business. Partnership between XIT and its vendors is also a factor, so third parties must fully understand the levels of service and cost structures that BPX requires and capture innovative ideas in either the demand or supply sides from the marketplace.

Conclusion

We have suggested that a distinguishing characteristic of IT is that two sides need managing: demand and supply. Firms often go from crisis to crisis when they tackle the demand side or the supply side in isolation. BPX saw that a successful transformation process had to tackle both. However, the balance of emphasis changes over time, as suggested by the four stages of transformation. Our framework is a prescription for transforming IT departments that are either in crisis or need to be reinvented for new business or technological environments. In the tradition of process models, the framework suggests how to make radical performance improvements.

In summary, there are four components of transformation at each stage (see Table 2). We propose that stage four is an ongoing state of equilibrium between supply and demand. However, if business or technological discontinuities occur and the company does not deal with them, it can initiate a new transformation process by returning to stage one.

In using this model for transforming the IT organization, it is vital not to forget the role of IT. IT has the difficult task of understanding the past and predicting the future of both the technological advances and business needs. IT senior management must balance these supply and demand management perspectives to keep the transformation on course.