Avoiding large-scale information system project failure
the importance of fundamentals
Concerns of Project Managers
THIS & THAT
Erwin V. Martinez, CSC Consulting, Waltham, Massachusetts
During the past 20 years, business and government have experienced their fair share of large-scale information systems (IS) projects. A large-scale IS project is big in all measurable terms: duration is measured in years, total team size numbers in the hundreds (or more), and work effort is tracked in tens of thousands of workdays. Furthermore, large-scale projects directly affect—and significantly alter—critical mainstream business functions. They therefore involve a broad cross-section of the business organization, uncover and address complex cross-functional issues, and fundamentally change core business operations.
Unfortunately, however, large-scale projects are extremely difficult and challenging to manage. Many, in fact, fail to meet their basic objectives of implementing broadly scoped information systems. The popular press and the computer trade journals chronicle a steady stream of failed large-scale IS projects—which span the spectrum of technologies, development methods, industries, project leadership philosophy, and computing platforms. Despite their differences, however, these projects all share a common characteristic: they failed because there was a breakdown in project management fundamentals.
Inattention to the fundamentals of project management will inevitably lead to project failure. This is especially true in more ambitious efforts, because large-scale projects allow little room to recover from fundamental mistakes. For instance, a delay of a month or two on a 300-person project may have serious consequences. Alienation and distrust among 600 or 6,000 users is very difficult to deal with.
Table 1. Large-Scale Project Management Essential Functions Model
ANALYST AND DOER LEVEL
Some industry professionals claim that the best way to avoid large-scale project failure is to simply stop doing large-scale projects. This suggestion, however attractive, is impractical because it ignores a basic reality about corporate and government systems in the 1990s: that is, many organizations have allowed their information processing capability to deteriorate so much that large-scale projects are imperative to their competitiveness—even their survival. As a result, large-scale projects are necessary to help these organizations make up for years of inadequate response to competitive pressures, customer demands, changing business conditions, and growing business volumes. Like it or not, large-scale projects will not go away.
BACK TO THE FUNDAMENTALS
To end this cycle of failure and improve the chances for project success, IS professionals must rededicate themselves to paying attention to the basic functions of project management, and must adopt methods to ensure that the essential project management functions are identified and earnestly performed. The following section reviews a multi-level model of these essential functions, and discusses in more detail the functions associated with the model's executive and project levels.
Each large-scale project embraces a set of essential function.s that must be actively performed for the project to be successful. These functions can be grouped in four basic levels: executive, project, team, and analyst/doer (Table 1). The redundancy of stating that these functions must be “actively performed” is deliberate. Too often, functions are not actually performed, but addressed in name only. Perhaps they are merely appended to someone's job title, added to the duties of an already overloaded team or team member, or ignored outright by the people to whom they are assigned.
These essential functions are interdependent yet distinct, and define a complete framework that addresses all activities critical to a project's success. Ignoring, understaffing, de-emphasizing, or circumventing any of these functions invites failure.
Project management provides the structure that holds a project together up and down this hierarchy and across it to business and information technology constituencies (Figure 1). The essential functions at each level coordinate and manage the creative build of the end products, connect the team with its constituencies, and administer and control project processes.
THE KEY FUNCTIONS
The need for fundamental project management, and the consequences if the fundamentals are ignored, are most critical in the model's first two levels, executive and project.
Organizations must overcome the anxiety toward change through executive support and sponsorship. Executive involvement and guidance helps ensure that the business vision is articulated and the project is managed toward attaining this vision. Without sufficient and substantive executive involvement, a large-scale project's focus frequently wavers, as the strategic business mission is supplanted by project or system objectives. A common example of loss of focus is when meeting an interim milestone date for finalizing the design becomes more important than designing a system that enables the organization to attain the business vision.
Figure 1. Project Management Communications Structure
As is evident in Table 1, several executive-level functions address the need to communicate frequently and comprehensively with business and IS constituencies not included in the project team, and to build and maintain positive relationships between all interested parties. Without active and willing participation by leaders within the organization that will inherit the system when the project is finished, building a workable solution is extremely difficult. Regardless of how good the system is, it will often be rejected if key constituencies were ignored or alienated along the way.
In sum, executives must assume direct oversight of the project, managing the effort in much the same way that they oversee business operations and maintain P&L responsibility and accountability. They must ensure that a discipline of administration and controls is actively in place to plan, manage, and report on project activities, and they must help identify and resolve the most severe resource problems and business and technical issues. Ultimately, the responsibility for whether the project succeeds or fails rests with executive management.
The project level includes the traditional “stuff” of project management, which takes place on a project—wide basis and constitutes the project manager's role.
The skills needed by a project manager are much different from those required by the system architect. A project manager must be adept at planning, controlling, orchestrating, and monitoring activities and resources, as well as at collecting information and building consensus. Technical skills, such as in-depth knowledge of the latest software packages and hardware platforms, are less important than the ability to direct team activities, track results, and anticipate potential roadblocks. Conversely, a system architect must be able to build creative business and technical solutions. Instead of focusing on the logistics and orchestration of the project, the architect is charged with identifying, designing, and assembling the correct combination of technical components that solve the business problems at hand. Because these tasks are fundamentally different in large projects and too consuming to combine, placing both project management and system architect responsibilities in the hands of a single person assuredly invites failure—not to mention burnout on the part of the unfortunate individual.
There should always be a focal point for project responsibility and account-ability—whether it is a few leaders or a single person—so that there is no confusion about who is in charge and to whom those playing other project-wide roles must report.
THE CHALLENGE OF LARGE-SCALE PROJECTS
The high incidence of failure among large-scale projects is a huge threat to modern commercial and government organizations. Technology innovations, expanding markets, increased regulation, growing customer demands, and new sources of competition are rapidly changing the business environment. Government is straining to expand its scope and volume of services while reining in costs. Information technology represents a source of power that, if properly managed, can enable organizations to meet these challenges. But while the opportunities presented by information technology are unmatched in world history, our frequent inability to manage the introduction of solutions that significantly leverage technology is standing in the way of our capitalizing on these opportunities.
The solution to large-scale project management lies within our grasp. We must get better at executing what we already know, and at mastering and applying the fundamentals of successful large-scale IS project management.
Editor's Note: For more detail, see the author's companion article on page 17 of the June issue of the Project Management Journal. ❑
Erwin Martinez is a partner with CSC Consulting, the consulting and systems integration unit of Computer Sciences Corporation. He has worked in systems development for 17 years, the past 12 managing large-scale information technology projects. Mr. Martinez's experience spans numerous vertical industries and functional areas, includes work with a variety of technologies, and encompasses all aspects of the systems development life cycle. He is a graduate of Yale University and a PMI member.
PMNETwork • June 1994