Implementing project management in a decentralized manufacturing environment
Competition in the food industry today is fierce. New product introductions, increased market share, and higher margins (read lower manufacturing costs) are the ingredients for survival and success. To support this strategy, manufacturers are rolling out product innovations at a dizzying pace and demanding higher returns on capital investment. At the same time, the in-house staffs necessary to support these efforts are shrinking. The “do more with less” mandate that fell out of this new reality spawned the need to find better ways of managing scarce time and resources. The old remedies of “throwing” money and human resources at projects that are well over budget or behind schedule are no longer viable solutions. Campbell Soup Company began addressing this need in earnest about 10 years ago.
Introducing management techniques associated with unique, one-time endeavors such as capital projects into an organization whose primary function is the planning and maintenance of ongoing manufacturing operations can be a daunting task. At Campbell’s in 1989, this situation was further complicated by a highly autonomous, decentralized, manufacturing organization. Individual plants were responsible for managing their own capital budgets, often with limited engineering resources. In time, Campbell’s Corporate Engineering Department realized that disciplined scope, budget, and schedule management was desperately needed to better achieve corporate objectives. But how can this new discipline be developed; and how can it be “sold” to corporate executives, skeptical plant engineering staffs, and many engineers in corporate engineering itself whose work to date had been oriented toward technical tasks? This paper describes the process used by Campbell’s corporate Engineering Systems Department in successfully introducing project management techniques to the implementation of significant capital projects, and how that discipline is now being applied in areas outside the engineering and construction realm. It is a story of overcoming the political and territorial hurdles inherent in such an organization to effect a fundamental change in how projects are executed. It is also an ongoing story of the changing roles of project managers in a manufacturing organization regardless of their technical background.
Campbell Soup Company is a global manufacturer and marketer of convenience foods, with annual sales of approximately $6.5 billion. Today, its major domestic brands include condensed and ready-to-serve soups, Stockpot refrigerated soups, Swanson broths, Pepperidge Farm, Franco American pastas and gravies, Pace salsas, V8 beverages, and Godiva chocolates. There are currently more than 20 manufacturing plants in North America with varying levels of engineering capability. In 1989, the beginning of this story, Vlasic pickles and Swanson frozen foods were also part of Campbell’s. The plants manufacturing those products, as well as some miscellaneous ingredient operations, contributed an additional 20+ plants to the system. Then, as now, the annual capital budget was between $250MM and $350MM. The corporate engineering office in Camden, NJ was approximately 200 engineers strong in 1989. The organization was broadly grouped along developmental, business unit liaison, and plant support lines. The overall objective of corporate engineering was to develop new, and improve existing, manufacturing technologies and provide technical support on plant capital projects. Projects at the larger plants were managed by plant engineering personnel, while those at smaller facilities were usually managed by a corporate engineer. Few engineers in the corporate office or at the plants had any formal training in project management.
The Catalyst for Change
In late 1989, a $50MM capital project was undertaken at one of the major thermal process (i.e., soup) plants. It involved the construction of a 225,000-square-foot building addition and installation of a new spaghetti sauce production line (Project STAR). The plant initiated the project, and assumed overall management responsibility. Corporate engineering assistance was requested to oversee design and construction of the building infrastructure by a design-build contractor, and to design the production line itself. The production equipment would be purchased directly by Campbell Soup. While the corporate engineers were also expected to help control the project scope, budget and schedule, ultimate control over these elements clearly rested with the plant managers.
Exhibit 1. Implementation Timeline
Campbell Soup Company had not undertaken a single project of this magnitude in many years, and none at this particular plant since its construction in the 1960s. The informal project management techniques used by many corporate and plant engineers to date focussed on technical rather than cost and schedule issues. Those techniques quickly proved to be inadequate to address the degree of cost and schedule risk that characterized this particular project. Examples of their failure include the following.
•Scope Management—A very rough, order-of-magnitude, estimate was prepared by plant and corporate engineering to determine the feasibility of this project. Unfortunately, that estimate became the basis for a decision by management to proceed and effectively became the project budget. Immediately, the plant showed its commitment to the project by selecting a local design-build contractor and, based on sketchy design concepts, directed him to begin excavation work on a time-and-material basis. It was at this point that corporate engineering was asked to help develop the building infrastructure and production line designs. In effect, the contractor had begun work with very little conceptual design information from Campbell’s and no written scope of work. We were now in the position of trying to define project scope and final project cost while incurring costs that were irretrievable.
Another failure in scope management centered around the lack of a formal authorization procedure for changes in the work. As mentioned above, corporate engineers were enlisted to guide the design work and control costs. However, as the “owner,” plant engineering felt it had the right to make changes in the design as necessary. With no change of scope procedures in place to monitor these changes, the cost impacts did not always come to light until after the fact. There was no mechanism for corporate and plant engineers to jointly evaluate a change before its implementation. Also, written direction was rarely given to the contractor to proceed with a change. As might be expected, the result of this deficiency was a loss of cost control and confusion among the parties concerning who had approval authority.
•Cost Management—Compounding the difficulty of controlling costs through scope management was the lack of a useful cost model and cost tracking system. The design-build contractor had never undertaken a project of this magnitude. His cost tracking methods were crude and basically being developed on the fly. Early on, it was difficult to assign the time and material costs to specific work areas since no detailed work breakdown structure had been established. The initial work breakdown structure was very general in nature, with associated scope of work descriptions that were so broad as to make effective cost control impossible. Finally, the most important obstacle to meaningful cost management was the absence of a guaranteed maximum price contract with the design-builder until nine months into construction.
•Schedule Management—The contractor’s inexperience with this type of project manifested itself again with respect to the schedule. Used to using simple bar chart schedules for smaller, less complex projects, his staff had very little experience with critical path scheduling. Hence, the building construction schedule he produced often lacked sufficient detail, contained flawed logic, and was not properly updated. This had two related effects. First, it proved to be a worthless document to the contractor as a plan for executing his work—many work milestones were missed with little or no warning; second, Campbell Soup could not rely on it to schedule its purchase and delivery of production equipment.
The lack of project management experience among all parties, and the failure to implement sound project management practices, resulted in a project that exceeded its budget by 15% and its scheduled completion by three months.
The headaches this project caused firmly convinced senior engineering managers that a project management discipline must be developed to meet the new cost and schedule challenges on future capital projects. Reinforcing this need was a substantial downsizing of corporate engineering in mid-1990, making more efficient management of remaining resources an imperative. Their strategy for implementation centered on five steps: (1) leverage the skills of the few on-staff engineers who already had experience in implementing project management methodologies, (2) evaluate the compatibility of existing engineering processes with good project management practices, (3) create a dedicated project management function within corporate engineering, (4) provide formal project management training across the engineering community, and (5) establish sound project control systems. Of course, the ultimate success of this effort rested on recognition of project management’s value by Campbell’s senior management, in particular operations management, and their support of the changes in procedures and relationships necessary to make it work. Exhibit 1 shows the time relationships of some of the actions taken.
Shortly after Project STAR, in early 1991, a project review session was held to discuss lessons learned. Two needs surfaced from this review. First, a more thorough review process must be instituted to insure that a project contains the appropriate scope of work, and cost and schedule control mechanisms prior to funding approval. Second, some sort of written document outlining the elements of a properly planned and executed project was needed to introduce Campbell’s engineers to project management concepts. At this point engineering management turned to Campbell’s own in-house expertise. Two corporate engineers were assigned the task of developing the planning and execution guidelines. One engineer had prior training and work experience in project management principles, and the other extensive hands-on experience with the engineering development process at Campbell’s. The result of their efforts was the publication of an elementary, “how-to” manual titled “Project Management Guideline.” The Guideline was structured in outline form, similar to a checklist, to make it seem less imposing to project management novices and increase the chances of it actually being used. It took corporate engineering two years to finally issue the document to all corporate and plant engineers, as the necessary support was garnered from senior operations management. Three nearly simultaneous events helped to achieve that support: (1) Preparation of a project management/project controls needs study by an outside consultant, (2) the undertaking of a $100MM+ capital project utilizing a more disciplined project management approach, and (3) execution of a $20MM capital project under the old methods.
The needs study was initiated in mid-1991. An internationally recognized engineering/construction/procurement (EPC) firm was contracted to analyze Campbell’s current capital project processes and assess the project management skills level of corporate engineering staff. The methods used to gather data included distributing questionnaires to all corporate engineers and interviewing a select sample of engineers. The study concluded that although the corporate engineering staff’s technical expertise was excellent, there were notable deficiencies within the organization in the areas of project scope management, and cost and schedule control systems. It also recommended that a formal project management training program be instituted.
Within six months of the EPC firm’s report, a $100MM project (called “D&I Project”) was proposed to upgrade steel can manufacturing facilities at the four U.S. soup plants. As a fairly unique Campbell’s project, with regard to dollar size, type of technology, standardization requirements, and the involvement of multiple plants, it presented itself as a good candidate to test the project management methodologies corporate engineering was touting. The characteristics of the project also dictated that it would best be managed out of corporate engineering. Just prior to the start of this project, a new Engineering Project Management group was created within corporate engineering. A project manager was assigned from this group. He, as well as the personnel chosen to oversee the building construction aspect of the project had prior project management experience. Drawing from that experience and the ideas presented in the needs study and Project Management Guideline, they prepared an initial execution plan that addressed two of the more glaring problems on previous projects—project organization and scope definition.
The organization challenge was to find a way to meld the long-held practice of plants managing all capital projects on their property themselves with the need to manage this particular project from corporate, where the project management expertise lay. The project manager and senior corporate engineering managers met with plant managers and engineering staffs at each plant to present their proposed execution plan. Support of the execution plan by senior manufacturing operations managers at corporate headquarters was certainly a factor in the plants’ acceptance of the more disciplined project management approach advocated by corporate engineering. This did not, however, preclude disagreements over the details of how the proposed plan would be implemented. All recognized the need to control the design, budget, and schedule; who would have that responsibility was the question. After some negotiation, a structure was finally arrived at that balanced the desires of all parties. Although there were some slight differences among the plants, the basic structure was the same. Each plant would have a “plant project manager” who would oversee the technical aspects of the production line and building infrastructure design, and also manage the construction. The “corporate project manager” would be responsible for developing and implementing the project controls systems necessary to manage budgets and schedules. Although the plant project managers were still responsible for budget and schedule (as the plants were still responsible for managing their own annual capital budgets), the corporate project manager had substantial influence over the decision-making process. Corporate engineers would provide technical support to the plant project managers, but play a greater role in the development of the construction strategy. The responsibilities, relationships, and communications requirements were finally set forth in a project procedures manual for each plant. They were also presented to senior corporate operations managers in an alignment meeting at Campbell’s headquarters, where team members from all four plants and corporate were brought together to review the entire project execution plan. The result was an organization that, although complicated by compromise, functioned very well and significantly improved communication between plant and corporate team members.
Defining and controlling the scope of work, one of the weaknesses surfaced by Project STAR and emphasized in the 1991 needs study, was given special attention on this project. An order-of-magnitude estimate was prepared by in-house engineers at the project’s inception to assess feasibility. However, it was agreed early in the development of the execution plan that a design consultant would be brought on board to prepare a preliminary engineering study that would establish the scope of work and increase the cost estimate accuracy. Only after this study was completed would a project authorization be requested by management. Once the project was authorized and entered the final engineering phase, all changes to the scope of work had to be approved by both the corporate and plant project managers. Another important factor in increasing scope control was better contracting methods. Although the work at all four plants was “fast tracked” using phased construction, construction documents were brought to 100%, or near 100%, completion before bidding. Lump sum or guaranteed maximum price contracts were used exclusively; no time-and-material. Change order procedures were clearly spelled out in each contract, and all significant change orders were reviewed by a corporate engineer and approved by the plant project manager.
In the end, thanks to a well-coordinated project organization, rigorous scope control process, formal execution plan, better scheduling process, and countless other improvements to its project management methods, Campbell’s finished this complex project on schedule and on budget at all four plants.
Concurrent with the D&I Project, a $20MM project to install a new production line was under way at one of the frozen food plants. This project did not employ the same project management methodologies used on the D&I Project and suffered a fate similar to that of Project STAR. The cost overruns were significant.
The immensely successful implementation of good project management techniques on the D&I Project, especially in light of the difficulties of the frozen food plant project, convinced senior managers in the operations sector of the company that the recommendations of the needs study must be carried out, and that corporate engineering’s efforts to “spread the gospel” to the plants should receive their full support.
With solid support now secured from the operations sector, corporate engineering moved forward in early 1994 with the development of a project management training program for both corporate and plant engineers. The Engineering Project Management (EPM) Group, through the recently created project controls subgroup, lead the search for an outside firm to conduct the training. Several engineering firms were evaluated before selecting a local capital project management consulting firm. The EPM group director and project controls personnel worked with the consultant to tailor its standard training program to Campbell’s typical engineering process. The training was conducted over a five-day period. Corporate engineers were trained first, then senior plant engineers. The program was structured to address the following phases in the life cycle of a project:
•Final Design and Construction
On this model the execution plans for all future Campbell’s projects of significant scope or risk would be based. The model was also employed to educate project stakeholders in other sectors of Campbell Soup on the engineering process, in particular the need for up-front planning and what cost estimate accuracy can be expected at various stages of a project. Finally, they would provide the framework for an expanded version of the recently issued Project Management Guideline.
Project Control Systems
Another needs study recommendation vital to the successful implementation of a project management discipline was development of effective project control systems. Although Primavera scheduling software had been introduced at the end of Project STAR, only one or two engineers knew how to use it. Even at that, their knowledge of the basic principles of critical path scheduling was sketchy. Cost tracking tools were primarily manual, difficult to keep current, and seldom had a cost forecasting element. The director of EPM set about to rectify these deficiencies with the help of engineering’s computer support group. He and some of his project managers met with the EPC firm that conducted the needs study to discuss in detail a strategy for developing a strong project control function. The result was a decision to formulate an integrated system for managing costs, schedule, and project documentation, and to create a project controls subgroup in EPM to oversee its implementation.
As a first step in the process, members of the support group underwent CPM training on Primavera scheduling software to improve scheduling proficiency. The EPM project managers were then consulted to determine the type of information required from a cost management system, and the scope of project documentation that must be maintained. Finally, a consultant was brought on board to develop an integrated cost, schedule, and document management system. One of the desired features of the system was a linking to Campbell’s purchasing function so that cost information would be updated instantaneously as purchase requisitions were issued, a substantial improvement over the then current system of tracking dozens, or hundreds, of requisitions manually. Questionnaires were distributed to all of the corporate engineers, followed by group interviews, to both solicit their input and sell them on the benefits of the system.
Approximately one year passed before what was known as the Integrated Project Management System (IPMS) was rolled out. Although the integrator had done a good job in defining the system requirements, the actual programming work had to be performed by Campbell’s engineering’s computer support group. The heart of the system, the cost management piece, was successfully integrated with the scheduling software. Because of technology limitations at the time, integration of document management was limited to just providing access to the software at the IPMS screen. The only major disappointment with the new system’s capabilities was the failure to tie directly into the purchasing department’s purchase order system. Campbell’s was developing a new accounting system at the time, and engineering was unable to muster the support needed to integrate the systems. Nevertheless, IPMS did have a purchase requisition writing feature, where the cost of a requisition written by an engineer was immediately displayed in the project cost report.
IPMS, in its original form, was used for three years to manage projects. Although it was readily accepted by the project managers, there was some resistance to it at first from the other corporate engineers. Many believed the effort required to input data and documents provided no real benefit to them. Excel bar chart schedules and budgets were satisfactory for their purposes. It was ultimately left to the project managers to enforce the new discipline. As it slowly began to take hold, IPMS became a valuable cost tracking tool. Unfortunately, IPMS effectively dissolved when the cost management feature was eventually abandoned. The reason for abandonment was companywide adoption of the new accounting and requisitioning system mentioned above. Engineering had to use the new system, instead of IPMS, to track project costs. This new system, primarily an accounting tool, does not perform the project budget tracking task as well as IPMS. One positive aspect of it, however, is that a project manager now has the ability to monitor all purchase order costs in real time, or approve them if he chooses to establish that level of control. The scheduling and document management pieces of the old IPMS continue to survive and are now used extensively by all corporate engineers.
Into the Future
By the mid 1990s, Campbell Soup Company’s engineering function had come a long way in project management since Project STAR. Corporate engineering made the commitment to find out how to do projects better. It called on its own in-house talent and outside experts to help define its weaknesses and create the processes to correct them. It reorganized to develop project management as a core competency. Formal project management training was provided. The new methodologies were tested on a major, high-risk project, and proven valid.
Yes, project management had gained a strong foothold in engineering, but there was still a lot of work to do. There still were skeptics. Senior management in other areas of the company was usually difficult to convince of the benefits of project management, particularly with respect to up-front planning and the cost of adequate project controls. Some plants still resisted changes in the way they had been running projects for years, and also saw no value in these new methods being proposed by corporate. Even some in corporate engineering, used to focussing on technical issues, considered the “paperwork” required to be of little or no value.
Three things have helped advance the project management discipline and increase its acceptance. First is the reduction in engineering staff over the years, both at corporate and the plants. This has forced engineering to employ more outside resources to provide engineering and construction management services. The project management skills needed to develop the scopes of work and necessary controls to effectively manage these resources have become more important. Second, Campbell’s now has a more centralized capital budgeting process than it had in the past. To help meet the company’s objectives for manufacturing standardization and more focussed capital management, large or high-risk projects are usually executed out of corporate headquarters. This has heightened the awareness of corporate engineering’s project management expertise. Finally, and most importantly, proof of project management’s value has been reflected by the superior execution of over $500MM worth of capital projects, including the D&I project. Budget and schedule overages have been reduced substantially. On the few projects where they have occurred, they have been relatively minor and were forecasted well in advance, allowing sufficient time to take mitigating action.
As a discipline, project management is today considered critical to Campbell’s corporate engineering department’s success in providing quality, cost-effective, timely technical support to the corporation. The high level of expertise in this area that corporate engineering has achieved gains more recognition every day. Certainly those inside Campbell’s who work directly on project teams appreciate its value. Likewise, outside engineering consulting firms have been impressed with our level of project management maturity and, because of it, work more effectively with us. Perhaps the clearest example of the degree to which project management has been accepted by senior Campbell’s management is the recent assignment of two engineering project managers to the role of program manager for new business development initiatives. This is the first time engineering project managers have been made responsible for overseeing more than just the engineering, construction, and start-up phases of a project. They are now being asked to lead teams of people from marketing, R&D, packaging, operations, and other Campbell’s organizations in executing all aspects of selected new product programs. This is a great opportunity for engineering to become more involved in the broader scope of such projects. But it is also a step toward greater understanding of project management methodologies by other organizations in the company, who will hopefully apply them in conducting their own business endeavors.
Proceedings of the Project Management Institute Annual Seminars & Symposium
September 7–16, 2000 • Houston, Texas, USA
Hurricane Katrina decimated thousands of buildings in New Orleans, Louisiana, USA, in 2005, including a U.S. Department of Veterans Affairs (VA) medical facility that served approximately 40,000…