Precluding the Centralization Versus Decentralization Argument
In an industry where productivity is often measured by cpu cycles, a different sort of cycle has emerged, one which is defined by the vacillation between centralization and decentralization. This vacillation, or cycling, back and forth has caused some industry analysts to propose a middle ground, one that seeks to provide a balance of the best of both centralization and decentralization. While achieving such a balance is seemingly possible, it requires strong leadership at the top of the organization and ultimately requires the ability to align business goals with the information technology function.
In his 1990 article, “The ‘Centrally Decentralized’ IS Organization,” Ernest M. von Simpson makes the case for a hybrid model that incorporates the best of both the centralized model and the decentralized model. Speaking largely from the perspective of a period of decentralization, von Simpson (1990) states that decentralization successfully “minimized turf battles over budget allocations and ensured closer connections to users – but too often with the results of creating a rudderless IS staff” (p. 3). In an effort to provide the staff with some direction, he argues that an organization should centralize its information technology functions, while still allowing business units to maintain their flexibility provided by the decentralized model (von Simpson, 1990). Centralizing the information technology functions would enable an organization to take advantage of lower licensing fees, volume discounts, and increased support options while providing structure to companywide efforts such as managing the network infrastructure, staff training and recruitment, and standard establishment (von Simpson, 1990). To achieve this sort of structure, von Simpson (1990) argues that an organization must operate in a bidirectional fashion regarding the interaction between the corporate community and the information technology community.
This sort of bidirectional interaction appears thirteen years later, when Barbara T. Bauer essentially argues the same points as von Simpson in her article “Is a Centralized or Decentralized IT Organization Better?” She describes a movement toward centralization, which was brought on by the need for Year 2000 conversions and the emergence of “new network and database technologies” that required “centralized IT organizations for development and operation” (Bauer, 2003). Acknowledging that these arguments have been made in the past, Bauer indicates that the same issues von Simpson highlighted still remain and are now accompanied by the addition of business’ almost total reliance on information technology (2003). In addition, the effective use of information technology has become a “life or death matter” in her opinion (Bauer, 2003).
Bauer highlights the major advantages and disadvantages of each model before proposing her hybrid/federated model. Specifically, she states that the benefits of centralization are in software and hardware procurement, redundancy elimination, and resource integration (Bauer, 2003). The disadvantages are the inability to be responsive to users and the perception of information technology as a huge cost center to the business (Bauer, 2003). Regarding decentralization, she states that ease of integrating various systems using the client/server model and the ability to respond rapidly to change are advantages. The disadvantages are higher operating cost due to redundancy, a lack of unified presence, and the lack of a single point of accountability (Bauer, 2003).
In her hybrid model, “functions that require consistency across the entire company are centralized” (Bauer, 2003). These functions include procurement, operations, architecture, standards and processes, and ERP application development. Unique applications, however, are left to each business unit, as well as unit specific service level agreements, resources, and strategies (Bauer, 2003). In essence, business units maintain “budget control over the decentralized functions” (Bauer, 2003) and technical resources remain the responsibility of the centralized organization.
Bauer’s arguments for the hybrid/federated model are very similar to those of von Simpson, perhaps speaking ironically to the cyclical nature of the centralization versus decentralization argument itself. Notable, however, is the way Bauer branches off her argument into a discussion of alignment by stating, "The corresponding challenge, and possible disadvantage to a federated model, is that it requires strong, collaborative leadership in both headquarters and individual business units. The CIO, in particular, must be an executive who can lead a complex technical organization as well as understand the business needs and strategy of each business unit. It also requires effective strategy, planning and resource allocation processes in the company so that centralized resources continue to meet the needs of specific business activities" (2003).
Indeed, this task is quite daunting, but it is exactly one of the skills that define successful leadership by the CIO and other executive level leaders. As Ben Worthen (2003) notes, tradition places information technology and business units on different pages. It therefore becomes necessary for the CIO to practice alignment, which is the art of anticipating business needs and translating them into successful information technology systems (Worthen, 2003). While numerous alignment techniques exist, the importance of the goal remains the same. The coordination between the information technology staff and the business units is paramount, and is perhaps the most important function of executive leadership.
Worthen (2003) states that the seemingly difficult task of executive leadership can be accomplished and illustrates it with two examples. The first example details the business “playbook” developed by Christine Modie, CIO of the Massachusetts Mutual Fund Financial Group. She developed this playbook by interviewing users, assessing their business needs, and prioritizing them as a series of goals for the year. Implementing the initiatives of the playbook saved the company 12.6 million in one year alone and her information technology group became more aware of the business as a result (Worthen, 2003). In addition, by saving money, the idea of information technology as a cost center was reduced and the credibility of her group increased in the eyes of the business (Worthen, 2003).
Worthen’s second example details Sheleen Quish’s “rankings game,” which she developed for U.S. Can. The goal of the rankings games was to “mesh the corporate goals with divisional goals” (Worthen, 2003). Using a series of rankings, corporate goals and divisional goals are mapped out and prioritized. Once the priorities are established, each goal is labeled as a “must-do, should-do, or a nice-to-do” (Worthen, 2003). In the end, the goal is to ensure that the information technology staff and the company’s business units see “eye-to-eye” (Worthen, 2003). Seeing eye-to-eye is exactly Bauer’s point regarding the hybrid/federated model’s capability to allow centralized resources to meet business needs in a decentralized environment. From Worthen’s examples, one can see that the CIO must play an integral role in bringing together both the information technology staff and the business units, whether thorough interviewing techniques, prioritization activities, or other alignment procedures.
Such examples illustrate that alignment procedures are not limited to the hybrid/federated model. While Bauer places this extra responsibility on the CIO only in the hybrid/federated model, one could argue the conceivability of these techniques in any organizational structure. Alignment, then, becomes the issue, and the centralization versus decentralization argument is seemingly precluded. In fact, applying the alignment techniques could aid in overcoming the disadvantages of the other models. Worthen’s examples illustrate these applications. For example, responsiveness and cost were labeled as disadvantages of the centralized model. If a CIO in a centralized environment were to incorporate the techniques Modie developed, these issues could be alleviated to some degree, due to the heightened awareness that results from interaction with the users.
Alignment techniques have implications for the decentralized model as well. For example, redundancy, accountability, and a lack of a unified presence were labeled as disadvantages. Quish successfully addressed these issues in her rankings analysis by requiring that every submitted project include an ROI figure, and a statement of how the project increases revenue, reduces cost, improves customer service, or stabilizes the work environment. In order for the project to proceed, it must have buy-in from everyone. In fact, her largest goal was to eliminate the “everyone-for-himself affair” (Worthen, 2003) that largely characterized the yearly allocation of information technology projects.
While the debate about centralized and decentralized models may continue to cycle for years to come, these examples indicate that alignment strategies can mitigate their respective disadvantages and bring them closer to the balance sought by the hybrid/federated model. Perhaps it is the nature of the information technology industry and the ever-changing technologies involved are the cause for such cycling between models. In any event, The importance of the executive leadership’s ability to align resources with business goals will remain the key to successfully implementing information technology in an effort to advance the company’s future.
References
Bauer, Barbara T. (2003, October). Is a Centralized or Decentralized IT Organization Better? Darwin. Retrieved October 28, 2003, from http://www.darwinmag.com/read/100103/question54.html
von Simpson, Ernest M. (1990). The ‘Centrally Decentralized’ IS Organization. Harvard Business Review, 1-6.
Worthen, Ben. (2003, April). Line Up the Team. Retrieved October 29, 2003, from
http://www.cio.com/archive/040103/hs_team.html
In his 1990 article, “The ‘Centrally Decentralized’ IS Organization,” Ernest M. von Simpson makes the case for a hybrid model that incorporates the best of both the centralized model and the decentralized model. Speaking largely from the perspective of a period of decentralization, von Simpson (1990) states that decentralization successfully “minimized turf battles over budget allocations and ensured closer connections to users – but too often with the results of creating a rudderless IS staff” (p. 3). In an effort to provide the staff with some direction, he argues that an organization should centralize its information technology functions, while still allowing business units to maintain their flexibility provided by the decentralized model (von Simpson, 1990). Centralizing the information technology functions would enable an organization to take advantage of lower licensing fees, volume discounts, and increased support options while providing structure to companywide efforts such as managing the network infrastructure, staff training and recruitment, and standard establishment (von Simpson, 1990). To achieve this sort of structure, von Simpson (1990) argues that an organization must operate in a bidirectional fashion regarding the interaction between the corporate community and the information technology community.
This sort of bidirectional interaction appears thirteen years later, when Barbara T. Bauer essentially argues the same points as von Simpson in her article “Is a Centralized or Decentralized IT Organization Better?” She describes a movement toward centralization, which was brought on by the need for Year 2000 conversions and the emergence of “new network and database technologies” that required “centralized IT organizations for development and operation” (Bauer, 2003). Acknowledging that these arguments have been made in the past, Bauer indicates that the same issues von Simpson highlighted still remain and are now accompanied by the addition of business’ almost total reliance on information technology (2003). In addition, the effective use of information technology has become a “life or death matter” in her opinion (Bauer, 2003).
Bauer highlights the major advantages and disadvantages of each model before proposing her hybrid/federated model. Specifically, she states that the benefits of centralization are in software and hardware procurement, redundancy elimination, and resource integration (Bauer, 2003). The disadvantages are the inability to be responsive to users and the perception of information technology as a huge cost center to the business (Bauer, 2003). Regarding decentralization, she states that ease of integrating various systems using the client/server model and the ability to respond rapidly to change are advantages. The disadvantages are higher operating cost due to redundancy, a lack of unified presence, and the lack of a single point of accountability (Bauer, 2003).
In her hybrid model, “functions that require consistency across the entire company are centralized” (Bauer, 2003). These functions include procurement, operations, architecture, standards and processes, and ERP application development. Unique applications, however, are left to each business unit, as well as unit specific service level agreements, resources, and strategies (Bauer, 2003). In essence, business units maintain “budget control over the decentralized functions” (Bauer, 2003) and technical resources remain the responsibility of the centralized organization.
Bauer’s arguments for the hybrid/federated model are very similar to those of von Simpson, perhaps speaking ironically to the cyclical nature of the centralization versus decentralization argument itself. Notable, however, is the way Bauer branches off her argument into a discussion of alignment by stating, "The corresponding challenge, and possible disadvantage to a federated model, is that it requires strong, collaborative leadership in both headquarters and individual business units. The CIO, in particular, must be an executive who can lead a complex technical organization as well as understand the business needs and strategy of each business unit. It also requires effective strategy, planning and resource allocation processes in the company so that centralized resources continue to meet the needs of specific business activities" (2003).
Indeed, this task is quite daunting, but it is exactly one of the skills that define successful leadership by the CIO and other executive level leaders. As Ben Worthen (2003) notes, tradition places information technology and business units on different pages. It therefore becomes necessary for the CIO to practice alignment, which is the art of anticipating business needs and translating them into successful information technology systems (Worthen, 2003). While numerous alignment techniques exist, the importance of the goal remains the same. The coordination between the information technology staff and the business units is paramount, and is perhaps the most important function of executive leadership.
Worthen (2003) states that the seemingly difficult task of executive leadership can be accomplished and illustrates it with two examples. The first example details the business “playbook” developed by Christine Modie, CIO of the Massachusetts Mutual Fund Financial Group. She developed this playbook by interviewing users, assessing their business needs, and prioritizing them as a series of goals for the year. Implementing the initiatives of the playbook saved the company 12.6 million in one year alone and her information technology group became more aware of the business as a result (Worthen, 2003). In addition, by saving money, the idea of information technology as a cost center was reduced and the credibility of her group increased in the eyes of the business (Worthen, 2003).
Worthen’s second example details Sheleen Quish’s “rankings game,” which she developed for U.S. Can. The goal of the rankings games was to “mesh the corporate goals with divisional goals” (Worthen, 2003). Using a series of rankings, corporate goals and divisional goals are mapped out and prioritized. Once the priorities are established, each goal is labeled as a “must-do, should-do, or a nice-to-do” (Worthen, 2003). In the end, the goal is to ensure that the information technology staff and the company’s business units see “eye-to-eye” (Worthen, 2003). Seeing eye-to-eye is exactly Bauer’s point regarding the hybrid/federated model’s capability to allow centralized resources to meet business needs in a decentralized environment. From Worthen’s examples, one can see that the CIO must play an integral role in bringing together both the information technology staff and the business units, whether thorough interviewing techniques, prioritization activities, or other alignment procedures.
Such examples illustrate that alignment procedures are not limited to the hybrid/federated model. While Bauer places this extra responsibility on the CIO only in the hybrid/federated model, one could argue the conceivability of these techniques in any organizational structure. Alignment, then, becomes the issue, and the centralization versus decentralization argument is seemingly precluded. In fact, applying the alignment techniques could aid in overcoming the disadvantages of the other models. Worthen’s examples illustrate these applications. For example, responsiveness and cost were labeled as disadvantages of the centralized model. If a CIO in a centralized environment were to incorporate the techniques Modie developed, these issues could be alleviated to some degree, due to the heightened awareness that results from interaction with the users.
Alignment techniques have implications for the decentralized model as well. For example, redundancy, accountability, and a lack of a unified presence were labeled as disadvantages. Quish successfully addressed these issues in her rankings analysis by requiring that every submitted project include an ROI figure, and a statement of how the project increases revenue, reduces cost, improves customer service, or stabilizes the work environment. In order for the project to proceed, it must have buy-in from everyone. In fact, her largest goal was to eliminate the “everyone-for-himself affair” (Worthen, 2003) that largely characterized the yearly allocation of information technology projects.
While the debate about centralized and decentralized models may continue to cycle for years to come, these examples indicate that alignment strategies can mitigate their respective disadvantages and bring them closer to the balance sought by the hybrid/federated model. Perhaps it is the nature of the information technology industry and the ever-changing technologies involved are the cause for such cycling between models. In any event, The importance of the executive leadership’s ability to align resources with business goals will remain the key to successfully implementing information technology in an effort to advance the company’s future.
References
Bauer, Barbara T. (2003, October). Is a Centralized or Decentralized IT Organization Better? Darwin. Retrieved October 28, 2003, from http://www.darwinmag.com/read/100103/question54.html
von Simpson, Ernest M. (1990). The ‘Centrally Decentralized’ IS Organization. Harvard Business Review, 1-6.
Worthen, Ben. (2003, April). Line Up the Team. Retrieved October 29, 2003, from
http://www.cio.com/archive/040103/hs_team.html
Labels: Academic Papers
posted by The Voice at 2:34 PM
0 Comments:
Post a Comment
<< Home