Dr. Morris has worked extensively in the Middle East, North Africa, Europe, South America and North America in the management of a wide variety of projects in the areas of steel, telecommunications, utilities, petrochemicals, and city government planning. He received his Ph.D. from Manchester in 1972. Dr. Morris presented papers at both the 1977 and 1978 PMI conferences.
Interface management is essentially the project manager’s job: planning, coordinating, and controlling the work of others at project interfaces. The following are all example of interface management principles:
- Tight control of dynamic interfaces is essential to achieving project cost, schedule and scope targets.
- Static project interfaces should be kept clearly defined through the life of the project.
- Organizational factors should not be allowed to inhibit required project integration.
- Project organization structures generally need to change as the project develops.
- Early, firm control of design is essential for effective project control.
- The design/production interface is the most critical project interface; it is also the most difficult to manage.
- The required amount of project management effort is a function of project size, speed, and complexity.
These interface management principles clearly use a jargon — a special language — a language more or less intelligible to most project managers, depending on their exposure to systems and organization thinking.
This paper reviews the relevance of modern systems and organization theory to project management, and then discusses these principles in detail.
I. Organization Theory and Project Management
Organization theory explains why organizations take on the structures and characteristics they do. It helps explain why project organizations have certain similar characteristics.
1. A systems approach is fundamental to thinking about project organizations.
Systems theory is an approach to bringing order to complex, poorly structured situations (4)(8). It does so by emphasizing certain core concepts, applied in a strict sequence, as outlined in figure 1. Using the systems approach, it is possible to clarify the design of a system (9), thereby permitting its overall objective to be more easily achieved.
Systems concepts, such as focus on objectives, subsystems’ definitions, and boundary control, have proved helpful to project managers in a number of ways. Project breakdown structures, organization and responsibility charts, information systems, and project specifications are all examples of project work that have been affected by the systems approach.
Systems management is a key systems concept (19). It focuses on two major activities:
- Goal achievement — getting the required results from the system
- Boundary management, which covers monitoring transactions at the system’s boundary and ensuring that the boundary is properly maintained and developed (16).
A major part of a project manager’s job involves boundary interactions. His task consists largely of monitoring the interrelationships between functional groups to achieve the overall project scope, cost, and schedule goals. Such monitoring is both technical (boundary monitoring) and managerial (maintenance and development).
2. Boundary management is especially important in the now prevalent Contingency School of Organization.
The major impact of the systems approach on organization theory has been the realization that organization structures vary depending on their environment (24). This view has become known as the Contingency School of Organization — because the organization’s appropriate structure is seen as contingent upon the task to be achieved (2) and the principal behavioral and structural factors affecting the organization (see figure 2). The primary tenets of the school are:
- Recognize there is no single, right structure.
- Accept change as often inevitable, necessary, and useful.
- Recognize the importance of subsystem interdependency.
In the contingency view, attention to organizational behavior at the boundary is very important. This is where change takes place, as a result of both the organization’s own performance and the organization’s interdependencies and environment.
3. The organization’s objectives are the single most important factors affecting its structure.
For organizations as for people, the prime goal is the most important factor influencing structure and behavior. Work in corporate strategy has explored how an organization can better adapt to achieve its primary goals and objectives (2)(3).
4. “Environment” is also a basic determinant or organization structure …
Environmental rate of change and uncertainty directly affect an organization’s structure and the level of management effort required (20)(31)(7). A flexible, organic type of structure is required in conditions of fast rate of change and high uncertainty, whereas a more rigid one is appropriate in steadier environments. Research by the author has confirmed the critical importance of project rate of change to the amount of coordination required in project organizations (29).
5‥‥ as is technology.
The type of technology being applied affects an organization’s structure. The more rigid the production technology, the more rigid the organization; and vice versa (35). In addition to the actual characteristics of the technology itself, the rate of change and uncertainty associated with technological forces are also important (7). Thus it is not just the type of work to be done (36) (design compared with production, for example), but the amount of management control needed that affects organization structure (37).
Evidently, environment and technology are closely interwoven. Indeed a dominant school of organization theorists — the socio-technic school — holds that organization structure is a product of both environment and technology, together with the attitudes of the people in the organization (11) (figure 2 again).
6. The degree of difference (“differentiation”) between subsystems partially determines the amount of work required to integrate them.
Differentiation is a term used by organization theorists to describe and measure both the relative differences in orientation between groups’ members and the relative differences in the groups’ formal organization structures.
Using the concept of differentiation, it is possible to identify where major subsystem interfaces — points of differentiation — should be located on a project. It is also possible to assess the difficulty likely to be encountered in coordinating the actions of groups across these interfaces.
Research has shown that for a system to be successfully managed both its boundaries and its areas of control must be clearly defined (25). Furthermore, the boundaries should be located where there are clear differences in technology, territory, or time (24).
The integration that is needed to overcome differentiation is a measure of the unity of effort required between subsystems (21). In project management terms, integration is precisely the project manager’s function: planning, coordinating, and controlling the functional groups’ efforts to achieve the project scope, cost, and schedule targets.
7. Other factors affecting the amount of integration required include the objectives of the organization, the speed of decision making required, and the degree of uncertainty.
The basic objective of the organization is a major factor determining the need for integration between subgroups. It determines both the groups that need integrating (marketing with pharmaceuticals, computing with hospital records, etc.) (3) and the amount of integration required (23).
The need for speedy management action is often overlooked as a factor affecting the integration required. Decisions required quickly often involve more integration than slower paced ones. As will be seen, project speed is a major factor determining the amount of project management effort required.
Uncertainty refers to both technical and environmental factors. Managers require information to reduce uncertainty. This invariably calls for integrative effort (12).
8. Every organization has both a degree of differentiation appropriate to it, and an appropriate degree of integration.
Subsystems have various patterns of differentiation depending on their objectives, environment, and technology. These differences should be clearly recognized (9):
- Differences between subgroups should not be smoothed over.
- Boundaries and their interfaces should be clearly recognized and appropriately managed.
- If additional integration is required to manage this differentiation, it should be provided.
Projects which move through a very definite life cycle, with different groups at work either simultaneously or sequentially, are organizations that typically require clear subsystem differentiation.
9. There is a range of integrating mechanisms available, ranging from the relatively simple to the distinctly complex.
There are three basic types of interface coordinating mechanisms: rules and standards, plans and programs, and personal contact (33).
Personal contact is the most complex and covers:
- Direct contact between managers
- Creation of liaison roles
- Creation of task forces
- Use of teams
- Use of permanent integrators (coordinators)
- Expansion to a full managerial linking role (with more authority and responsibility than the integrating role)
- Establishment of the matrix organization form (12)( 13).
10. The type of integration required is a function of the relationship between the interfacing subsystems.
The type of integration required at art interface is a function not just of the differentiation at the interface, but also of the interdependence between the interfacing subsystems. For projects the two key types of interdependence are sequential and reciprocal (figure 3). Sequential interdependence is where one subsystem follows straight on from another — construction from a completed design, for instance. Reciprocal interdependence is when each gives and each takes; this is the hardest type to manage. Integration by plans and programs may be sufficient for sequential interdependence but for reciprocal cases the more complex mechanism of personal contact is needed (33).
II. Project Interfaces and Their Integration
Project organizations are relatively new and not extensively documented, yet we are now reaching the point where it is possible to say with some confidence where the principal organizational interfaces in a project will be, and to say (as we will in part III), with slightly less confidence, how they should be managed.
1. The project manager’s principal concern is the successful management of the project’s process interfaces.
The project manager is ultimately responsible for the on-time, in-budget, to-scope completion of his project. It is therefore his responsi-b-ility to control and coordinate the work of the functional groups to achieve these project scope, cost and schedule targets.
Functional managers have responsibility for:
- Accomplishing work in their functional area on schedule, in budget, to scope
- Providing functional policy and procedural guidelines
- Providing trained staff
- Maintaining technical excellence
The project manager, on the other hand, has responsibility for:
- Planning, coordinating, and controlling functional groups’ work to achieve project budget, schedule, and scope targets
- Instigating and guaranteeing corrective action where necessary to ensure achievement of these targets (9).
Functional work is primarily product-oriented, whereas the project manager’s work focuses more on the process of accomplishing the work. Process interfaces can be distinguished from product interfaces, as shown in figure 4.
The extent of project manager/functional responsibility varies, as illustrated in figure 5, depending on the size of the project, the availability of experienced project managers, and the tradition within the company (30).
2. The project manager’s prime role is that of interface manager.
Interface management is the integration of different groups’ activities across (project) interfaces. An understanding of the two key concepts in that definition — integration and project interfaces — is crucial to any understanding of what project management is today.
As already discussed, integration consists of bringing the necessary unity of effort to bear. Effective integration implies:
- Contacts with all principal functional groups
- Emphasis on equalizing power differences and increasing trust
- Emphasis on influencing and guaranteeing decision making rather than taking the decisions themselves (13).
Project interfaces can be divided between static interfaces — basic organizational interfaces that do not change with the evolution of the project — and dynamic interfaces — process interfaces that mark key transitions in the project’s evolution.
3. Static project interfaces arise from the basic project structure.
Static project interfaces arise from permanent subgroups on the project. These groups may exist for special company or project reasons, but at least three and sometimes four subgroups should exist on most projects. These are:
- The independent planning and control group. Planning and control should be an organizationally independent group. It should not be a part of engineering, construction, project management, or any other group. Only in this way can one ensure the reporting of reliable project status data, free from the bias of groups with their own ax to grind.
- The independent project manager. The project manager should be the focus of responsibility for project success. As the integrator ensuring balanced effort by the other groups and instilling trust, he must be organizationally independent of all other groups.
- The functional groups. These include such groups as engineering, logistics, procurement, construction, quality assurance, etc. For large new-company creation projects such groups as marketing, sales, operations planning, maintenance planning, recruiting, and training may also be included.
- In multiproject situations, or large single projects having a number of subproject groups, a further major static interface is that between the subprojects.
Figure 6 shows an owner’s organization for a capital expansion proj-e-ct which illustrates these four static interface types.
4. Dynamic interfaces arise from the process structure of the project.
The process structure of the project is a function of the process technology, the speed of project execution required and the organizational constraints acting on the project. (For example, the British building process is severely constrained organizationally by the Royal Institute of British Architects’ (RIBA) stranglehold on the building process and their process definition: the RIBA Plan of Work) (32).
Examples of interfaces arising from process technology are:
- Acceptance of project scope definition/brief (systems/configuration analysis)
- Acceptance of basic engineering/design
- Design: manufacturing interface
- Design: construction interface
- Plant start-up/product delivery quality check
Figure 7 gives a model of a typical project showing the basic process interfaces.
Speed and organization factors give rise to managerial interfaces. They also affect the type of integration required. Examples include:
- Contracting policy definition
What is to be contracted out
Type of contract
- Contract letting process
- Degree of overlap between design and production (manufacturing and construction)
III. Research on Interface Management
The following are principles of interface management. All are based on actual cases; most are documented research. In no way do they represent all there is to know of interface management, but they do highlight a number of project interface management’s core concepts.
1. Tight control of dynamic interfaces is essential to achieve project cost, schedule, and scope targets.
Control of dynamic interfaces has a major impact on project cost, schedule, and scope performance.
Basic engineering determines almost totally the scope, cost, and schedule parameters of the project. Design exerts a strategic influence on the management of a project (26). Incomplete or late definition of engineering or poor understanding of the implications of what is being designed has led to many project overruns. Concorde is probably one of the most famous examples; the project cited in item four below probably one of the least. Effective control of brief and engineering is a major testing area of effective project management.
Poorly planned plant start-up can have a very large budgetary impact. At 12 percent p.a., a one-month delay can cost 1 percent in interest charges, let alone in opportunity costs of delayed production and sometimes, company-wide inefficiencies of underutilized personnel and equipment.
The interface between the project proper and operations start-up — predominantly a scheduling one — is a very major one, requiring concentrated senior management attention in effectively integrating the two systems’ work. Project management has, to date, generally confined itself to providing the unified management of the many work efforts required to deliver a product. Focus on integrating the similarly many efforts needed to plan and prepare for start-up and operation in parallel with the project delivery process has not yet received sufficient attention in practice, though some theoretical work is just beginning to be published (5). The challenge of providing trained staff, ready at the precise moment the facility is completed, is a very major one. More work is needed in this area.
Contracting policy and letting of contracts is critical. They determine the extent of owner liability and the degree of project management and control that will be needed on the project. Many owners in the developing world, for example, have in the past had their facilities constructed on a fixed-price, turnkey basis. Nowadays, with developing nations generally seeking a more active role in the management of their projects, fixed-price, turnkey projects are becoming less popular. As a result owners have become involved in real control of contractors’ costs as well as more extensive management of engineering, purchasing, expediting, transport, warehousing, and scheduling (14).
A major technique for control of dynamic interfaces is change control. Change control, covering such techniques as the change control board, change orders, and the new budget amendment concept (15), is a classic boundary management technique.
2. Static project interfaces should be kept clearly defined through the life of the project.
Poor definition of the interfaces between principal static groups — project management, functional, planning and control, subprojects — can lead to organizational problems of blurred definitions of responsibility and the reporting of unreliable data.
Because of the temporary nature of most project organizations, there is a tendency, especially in matrix organizations (30), toward confusion of roles and conflict of decision-taking power between the functional groups and the project managers (38). A major concern of senior management should thus be to ensure that roles are kept clear by maintaining clearly defined project management/functional interfaces.
Static interfaces are nearly always those between established, relatively long-lived groups. Often substantial competition exists between these groups. In such cases, care must be taken to ensure that the control data used to measure performance across the interfaces is as free from bias as possible (22).
Many construction organizations place the cost and schedule group under the control of the construction manager. No construction manager likes to show his cost and schedule trending to overrun too vigorously, too soon. Planning and control must be independent of the functional groups; if they are not the data may become fudged.
In multiproject situations a further concern of senior management is the integration of subproject work. This means paying careful attention to ensuring effective intercommunication between project managers (by implementing regular meetings and often by using teambuilding techniques for the project managers) and integrating project performance within functional groups. The project shown in figure 6 — a $3.5 billion steel mill project — required management emphasis on both these points (30).
3. Organizational factors should not be allowed to inhibit project integration.
Organizational factors have a significant impact on the nature of dynamic interfaces. Unfortunately there is often a danger that they may inhibit rather than stimulate the type of integration required.
Research by the author showed how this often happens in the British building industry. Figures 8 and 9 — based on figure 10, a model of the British building process — compare two basically well-run, large, complex projects. The first (figure 8) was managed by architects working under the prevailing concepts of the traditional RIBA Plan of Work (32). The Plan of Work is overly architect-oriented and excludes any production input into the early stages of the project. Thus neither scheduling nor ways of getting the required production expertise into the project in its early stages were considered.
The second project, however (figure 9), being a large factory built for a large, experienced company employing a well-known U.S. A/E firm, did not use the RIBA Plan of Work. Instead basic “good” principles of project management organization were employed (29). As a result, (i) a management contractor joined the project at the outset; (ii) there was thus integration of detailed schedule studies into the early design work; (iii) careful consideration was given to organizing the project to get maximum speed of completion plus effective control and coordination; (iv) careful cost and buildability evaluations were made of the initial design implications; (v) there was a well-controlled, evolutionary transition across the design/construction interface.
4. Project organization structures generally need to change as the project develops.
As projects develop, the relative influence of various groups changes. Project organization structures should reflect this change. For instance, on electronics/telecommunications/weapons projects, early systems work is fundamental; on large capital expansion projects, especially in developing-world countries, infrastructure work is paramount. Later both these assume a greatly reduced role. Similarly, in construction as one moves from earthmoving to structure to equipment erection and finishing work there is a change in organizational influence. Andreotti, for instance, showed how the project structure changed as a housing project developed on a U.S. Corps of Engineers project in Saudi Arabia (1).
5. Early, firm control of design is essential for effective project control.
The interface management of design is exceptionally difficult. Both the major design interfaces are poorly defined. The basic/detailed design interface has no clearly recognized changes in either technology, territory, or time (25); nor are there any major organizational changes. Yet, as seen in 111 (1), firm control on finalizing design is fundamental to project success. Definition and subsequent management of this interface is thus hard as well as important.
The project shown in figure 8, for instance, suffered from inadequate recognition and poor control of the basic design/detailed design interface. Many of its later construction problems arose from the poorly appreciated implications of the basic design concept.
6. The design/production interface is the most critical project interface; it is also the most difficult to manage.
The design/production interface covers design/manufacturing and/or design/construction. This is the most difficult project interface to manage, precisely because the differentiation across it is so big. Characterized by an amazing conjunction of dimensions, the design/production interface is differentiated between systems having various technologies and often between different organizations or organizational units. It is an information interface, often with time and territory differences, and it is oftert vague and ill-defined.
Especially important in determining the difficulty of managing this interface is the complexity of the design relative to the project’s size and speed of completion. Where the design is complex, organizational problems should be minimized; otherwise the interface may become too complex to manage. A more useful generalization of this rule is that if there must be technological, timing, geographical, or organizational complexity at an interface, make sure it is only one of these at a time (27)(28). Do not complicate matters by having organizational complexity — for example, contractual complexity or responsibility uncertainty — at the same time as technical complexity. Thus, at the design/production interface: identify it well; assess schedule and organization integration requirements; and set up clear and effective integrating mechanisms for coordinating and controlling progress across the interface.
7. In projects, the amount of project management effort required is a function of the project size, speed, and complexity.
As already noted, the degree of differentiation in an organization, its basic objective, speed, and degree of uncertainty are generally held to be key factors determining the amount of integration it will require. Similarly, research by the author has shown that for projects the key factors determining the amount of project management effort required are the project size, speed, and complexity (27)(28).
The span of control concept has long been a wellknown management principle. The bigger the group, the more management effort will be required to control it. Rate of change was seen in Part I of this paper to be a major factor determining an organization’s structure.
Complexity is a compound term covering the degree of uncertainty; the project’s technical complexity; the pattern of technical, contractual, geographical and schedule differentiation; and sequential or reciprocal interdependence.
In addition, as already seen, the integration required at an interface is a function both of the degree of difference and type of interdependence across it. For projects, it holds that the type of integration and personnel makeup of integrating groups is a function of the degree of differentiation and the type of interdependence between the interfacing subsystems. Also, the frequency of the integration is a function of its rate of change and degree of uncertainty (28).
The structural perspective of organization theory gives project managers a useful overview on the management of key interfaces on their projects. Basic project interfaces are now well defined. Organization theory can explain why these are so important and can give guidelines as to how they should be managed.
The three critical factors affecting the degree of integration required at a project interface are the size, speed, and complexity of the project. The degree of differentiation and type of interdependence across an interface determine the type of integration it requires. Figure 11 summarizes these conclusions on integration.