Project categorization systems and their use in organizations

an empirical study

Dr. J. Brian Hobbs,
University of Quebec at Montreal

Dr. J. Rodney Turner,
Erasmus University, Rotterdam

Proceedings of the PMI Research Conference 11-14 July 2004 – London, UK


This paper presents a summary of the methodology and the results of an investigation of project categorization1 systems initiated and supported by the Project Management Institute2. The primary focus of this research project has been the study of project categorization systems as they are used in practitioner organizations.

In any organization that undertakes a large number of projects, the description of the portfolio of projects requires the identification of types of projects undertaken and the use of labels to name the different types. These labels are the project characteristics that form the basis of a project categorization system. All organizations that manage an appreciable number of projects have a system of categories for describing and managing their set of projects, whether the system is explicit and formalized or implicit and informal.

Two distinct aspects of project categorization systems have been investigated and form the backbone of this paper: 1) the organizational purposes served by such systems and 2) the attributes or characteristics used by organizations to divide their projects into groups or categories. The study revealed that project categorization systems are used for a large number of different purposes that include the grouping of projects in order to identify the level of approval they require, the competencies and training needs of project management personnel, the methods and techniques that will be appropriate to apply to their management, which budget they will be funded from, and their alignment with organizational strategy. The study also revealed that a great variety of attributes is used to divide projects into categories. These include size, level of complexity, geographical location, and technical discipline. All the organizations studied have complex project categorization systems based on multiple attributes that they use for a large number of different purposes.

The primary product of this study is a model that is comprised of two “maps.” By map, we mean a hierarchically ordered presentation resembling a decision tree. The first map presents the multiple organizational purposes served by such systems. The second map presents the many different attributes or characteristics that organizations use to divide their projects into groups or categories.

Principles of Categorization System Design and Management

What is categorization and what are its implications? It is such an entrenched part of our lives, and indeed of the way we think, that it is very often taken for granted. Writers such as Taylor (1999) and Bowker and Star (1999) even go as far as to suggest that the need to categorize and organize is an innate part of human nature. Very little of the world around us is not categorized and many of the categorizations we make are an unconscious part of our thought processes. Kwasnik (1992) states that we “create classificatory schemes to organize our knowledge of the world in such a way as to be useful” (p. 63).

Designing a categorization system presents numerous challenges, the first of which is where to begin. Categorization systems seek not only to organize a field, but also to do so in a way that is of use to those whom it effects. In short, categorizations serve a purpose and it is this purpose that is responsible for shaping the system. The scope of the system, its hierarchy and what is considered significant in distinguishing between its entities, all derive from the aims that the system hopes to meet. Similarly, the success of a categorization system can be measured by how well it achieves these ends.

Bowker and Star (1999, p. 231) identify three key areas of challenge in developing a categorization system for a work setting: comparability, visibility and control. First, a categorization scheme must provide comparability. To do this, there needs to be some standardization of the language used to describe work activities. This greatly aids communication, ensuring understanding among users. A standardized vernacular also means that people can move between projects, even internationally, without having to learn new terminology. Also, comparability makes it possible for practitioners to draw on the lessons learned from similar projects, facilitating knowledge management, and increasing the likelihood of success. Researchers and professional associations also benefit from standardization, as it brings comparability. Standardization does, however, pose its own set of challenges. Reducing the variety and complexity of reality to a small set of categories requires a considerable degree of simplification. In addition, rigid adhesion to a set of categories and rules for categorizing can lead to inappropriate decisions.

The second challenge is visibility. Categorizing something greatly enhances visibility, and, similarly, entities that are not identified by the categorization scheme can be ignored and become invisible. There are several issues associated with visibility. What territory is covered by the categorization system? In a project context, this means deciding which activities will be included or excluded from the system. Will non-project activities such as operations and maintenance also be included? Which projects should be included or excluded? Which types of projects are sufficiently different to merit identification within the system? Which attributes should be used to identify these differences? The system needs to identify projects that are different and attributes that differentiate, in other words, to identify differences that make a difference.

The third area of challenge in constructing a categorization scheme is control of the application of the system. Having control of the system means being able to exercise discretion with respect to the interpretation of the rules of categorization. A categorization system is a representation of reality. As such, it is necessarily a simplification. Designing and using a categorization system always requires some degree of judgement, first in the identification of the categories and the rules for categorizing. The rules will never be perfectly unambiguous. Some judgement or discretion will be necessary in their interpretation and application. During use of the system, judgement must also be exercised to account for unforeseen or changing circumstances. In short, some discretion will be necessary. Exercising discretion in turn will require both judgement, which is often based on experience and training, and organizational power.

These three challenges can also be represented as a triangle, analogous to the triple constraint in project management, with trade-offs being necessary along the sides as shown in Exhibit 1.

The Three Challenges of Categorization Systems

Exhibit 1: The Three Challenges of Categorization Systems

In a perfect scheme, these three areas are balanced in such a way that the benefits of each are enhanced without any one area yielding ground to another. However, such a balance is a practical impossibility as each of the three areas "tradesoff" against the others and cannot be advanced simultaneously. High levels of comparability and visibility negatively impact on user discretion, thus affecting the manageability of the system. Increased user control takes away comparability, as it introduces variance among users. Increasing the number of categories and the number of attributes reduces standardization and comparability. The interplay between control and visibility is complex. For anything to be controlled, it must be visible. Making things visible makes them potentially controllable. Part of the game of control is played out by selectively making other people's activities visible and by making some of one’s own activities visible while hiding others. An understanding of the interactions and trade-offs among comparability, visibility and control is essential to the design and management of a categorization system. The effectiveness of the system will depend to a large extent on the interplay of these issues.

A categorization system must be accepted by those whom it affects for it to work effectively. Implementing or changing an existing project categorization system is an organizational change. As such, its success depends on issues of ownership, process, and perceived interests. The system must accurately reflect the participants’ real-world experiences. This is not always easy as different users may have different perceptions of the system and its fit with work practice. “The messy flow of bodily and natural experience must be ordered against a formal, neat set of categories” (Bowker & Star, 1999, p. 68). A categorization system must use terminology and categories that are present in the organizational culture. There is often a tension between clear, consistent, scientifically-based categories on the one hand, and intuitive, commonsensical, well-accepted terminology on the other.

By nature, categorization systems seek to describe something as it exists at a particular moment in time. As such, a scheme designed to categorize a complex and evolving field runs the risk of excluding future changes and developments. In a work context, this is of particular concern as the nature of the whole set of projects and their context evolve, and as new tools and techniques are introduced to address current managerial needs. If the scheme is not relevant to current work practices, it will not be used.

The Politics of Categorization

The purpose of a categorization system is a primary factor in the shaping of the system. A system with multiple users with different needs can result in conflicts that, if not resolved, can become embedded in the system and therefore affect its functionality. In the development of any categorization system, decisions have to be made with regard to choosing categories, which in turn determine what is to be a visible part of the system. Too strict a categorization can lead to the loss of user discretion, particularly in a work setting. This loss of autonomy is generally not evenly distributed across the organization as “the loosest [categorization] of work is accorded to those with the most power and discretion who are able to set their own terms” (Bowker & Star, 1999, p. 46).

Interests and biases are at work. Bowker and Star note, “the spread or enforcement of categories … involves negotiation or force” (1999, p. 44). What is considered significant in differentiating between the entities of the system is subjective and dependant on one’s perspective, on the need one has for the system, and on the impact the system will have on one's position. People responsible for financial management, strategic planning, human resource management, and project management methodology will all have different points of view and different needs. “Each category valorizes [sic] some point of view and silences another” (Bowker & Star, 1999, p. 5). Therefore, without proper negotiation and consideration, exclusion can occur, rendering some things wrongly invisible and, as such, negatively impacting on the effectiveness of the system.

Categorizing Projects

Literature on categorization within the field of project management has for the most part focussed on tailoring management style to suit project type (Shenhar & Wideman, 1996, 1997; Dvir, Lipovetsky, Shenhar, & Tishler, 1998; Shenhar, 1998; Payne & Turner, 1999; Youker, 2002). An organization adopting one of these project categorization systems would do so for the purpose of identifying the project management practices that are best suited to each of the different types of projects.

The literature on project portfolio management casts quite a different light on the subject. The grouping of projects into categories is an essential step in the project portfolio management process (Cooper, Edgett, & Kleinshmilt, 1997; Dye and Pennypacker, 2000; Archibald, 2003; Aalto, 2001). Managing a portfolio of projects requires the grouping of projects into categories. The purpose of these groupings is quite different from the groupings in the literature mentioned above. Portfolio management focuses on project selection, allocation of financial and other resources, the alignment of the portfolio with organizational strategy, monitoring and controlling the attainment of strategic goals, balancing the portfolio, maximizing value to the organization, and providing visibility to upper management.


There were six phases in the life cycle of this research project. The phases were as follows:

  1. Literature review
  2. Nine focus groups to investigate project categorization systems in organizations
  3. Web-based questionnaire to validate and expand on results from phase 2 (119 responses)
  4. Development of a preliminary model based on a synthesis of the results to date
  5. Validation of the preliminary model with organizations from phase 2
  6. Analysis of the results, modification of the preliminary model, and the writing of the final report

The following sections of the paper present many of the research findings. These lay the foundation for the model that follows.

Project Categorization Systems are Not Immediately Visible

In response to inquiries as to the existence of a project categorization system in many organizational contexts, practitioners often respond in the negative, even after a short explanation of what is meant by the term. Paradoxically, after further examination, accompanied by examples, all the participants in this study could clearly identify the project categorization systems in use in their organizations. We argue that any organization with a significant number of projects must have some way of labelling the types of projects they do and that this is, in fact, a categorization system, though possibly an informal one. There are two explanations for this difficulty in recognizing the systems in place: 1) categories in use are taken for granted and 2) categorization is an abstract concept.

Within organizations, people invent, adopt and modify terminology in an effort to make sense of their organization, its activities, and its environment. Through usage over time, the terms used to identify and describe the organization’s projects become an artifact of the organizational culture. Visiting an organization with which one is not familiar, one is very often struck by the “peculiar” use of terms which, once explained, make perfect sense. Just as one example, the expression “dusk-to-dawn-projects” in a transportation-infrastructure company refers to projects in which execution activities must be done at night while the transportation system in not in use, and that must return the system to operational status by dawn each day. The term is widely used and understood by organizational members, but when asked whether they have a project categorization system, they may not think immediately of this term as a category of projects. It is simply a term that is used in everyday life in the organization. It may be formalized in organizational policies, procedures, and documentation, or it may not.

There is a dynamic interplay between the language used in each organization or organizational unit, and the language used in the industry and the geographical location in which the organization operates. Both emerge over time through usage. And both can become part of the culture and not come immediately to mind when one is queried on the presence of a project categorization system.

The other reason that people have difficulty identifying the categorization systems in use in their organizations is that the concept is quite abstract. An inquiry as to what types of projects are done in an organization can be answered quite readily. But the term “project categorization system” is a term not often used by practitioners in their everyday lives and has no immediate and clear meaning for them.

The Formalization of Project Categorization Systems

The respondents to the questionnaire were queried as to the nature of their systems. Overwhelmingly (97%), organizations develop their categorization systems internally. However, there is considerable variation in the extent to which categorization systems are consistent throughout the organization or local and variable, and in the extent to which they are formalized or not. The research identified three factors that were associated with the level of formalization of the project categorization systems found in organizations: size of the organization, public vs. private, and level of centralization of decision-making.

Size Contributes to Formalization

The relationship between organizational size and level of formalization is the longest-standing and most-often-verified relationship found in organizational research (Mintzberg, 1979). Not surprisingly, the same association was found during this research project. Of the eight organizations that participated in the focus groups, seven readily identified project categorization systems that were all relatively formal. These organizations all had more than 1,000 employees involved in project work. The other organization with which a focus group was organized had more difficulty identifying the way in which projects were categorized and, when the system as such was identified, it was very informal. This is an entrepreneurial firm with 180 employees spread out across several cities. In this organization, the categorization of projects was not seen as important. Project categorization is a tool for dealing with multiple projects. As the number of projects increases and as the set of projects becomes more stable, the use of categorization becomes more formalized.

Public vs. Private

There is some indication of a relationship between the public/private nature of the organization and the formalization of its project categorization system. There seems to have been a phenomenon of self-selection among respondents to the web-based questionnaire. The percentage of respondents from the public sector was high (79%). This may mean that a greater proportion of employees of public sector organizations felt strongly enough that the issue of categorization was important enough to merit responding to the questionnaire. However, no systematic variation was found between respondents for public and private organizations. There is a general perception that public organizations are more formalized. This may be true of project categorization in public organizations. However, the evidence of a relationship between the public/private nature of the organization and the formalization of its project categorization system is inconclusive.

Centralization of Decision-making

In both the financial institutions participating in this study, the project categorization systems were put in place as part of a move to centralize investment decisions. In each there was a clear link between project categorization and investment decisions, with decisions being made at a very high level. These organizations were using categorization to manage the portfolio of projects and to align investment decisions with strategy. When the researchers returned to one of the organizations after several months, a new CEO had been appointed who was emphasizing decentralization of investment decisions. The project categorization system was being used in a much more ad hoc fashion and was perceived as less important than had previously been the case. These are illustrative examples and nothing indicates that this phenomenon is restricted to the financial sector. These examples are indicative of a relationship between the centralization of project approval and of budgeting and investment decisions, and the increased formalization and visibility of project categorization.

Construction of Complex Categorization Systems

None of the organizations studied had a simple project categorization system based on a small number of attributes. The project activity of organizations is multi-dimensional and complex, and organizations develop project categorization systems that match important elements of this complexity. Increasing the complexity of systems is certainly not seen as desirable. On the contrary, the complexity of the systems can be a source of difficulty, as will be discussed later.

The analysis of several project categorization systems has led to the identification of several ways in which organizations introduce complexity into their systems using a set of attributes as building blocks. Four dimensions have been identified and are discussed below: hierarchical systems, parallel systems, composite attributes, and exceptions.

Hierarchical systems

Many of the categorization systems that were analyzed are hierarchical. The organization's projects are divided into groups using one attribute and then each of the groups is further divided into sub-groups. The sub-groups are often each divided up using different sets of attributes. The hierarchical breakdown can continue for several levels, often in an asymmetrical fashion. Of the organizational responses to the web-based questionnaire that consistently recognized having a project categorization system (n=100), 54% claim that the project categorization is hierarchical.

Parallel or overlaid systems

The majority of organizations participating in the focus groups have hierarchical categorization systems. However, they also use other attributes to categorize. These additional attributes were overlaid on the primary system in a complementary fashion. Each of the “additional attributes” dealt with an important aspect of the projects and their management. Many organizations have more than one overlay on their primary categorization system. This is very difficult to illustrate graphically. The “additional attributes” produce an overlaid or multi-dimensional matrix system. However, not all of the attributes used to describe and group the organization’s projects are organized into one coherent and comprehensive system. Quite to the contrary, multiple parallel systems are in place in organizations. Respondents do not see this as dysfunctional. It is seen rather as necessary to manage the organizations’ projects effectively.

Composite attributes

Some organizations use composite variables to categorize projects. The most common example encountered was the attribute of complexity. Some organizations use complexity as a single attribute, but others define complexity using multiple characteristics to create composite attributes. In responding to the web-based questionnaire, those that stated that their organizations used complexity to categorize projects (n=57) indicated that they used between one and twelve attributes to measure complexity with an average of five attributes. The frequencies with which different attributes were reported as being used as components of complexity are reported in Exhibit 2.

One of the participating organizations has a scale for measuring the complexity of its systems modification projects. The ten-point scale is built up from the following attributes: number of design elements involved, the procurement method (fixed price or not), the uncertainly of system requirements, the reliability of the data on the current operations of the system, the presence of specific technologies deemed as sources of risk, and the reliability of the contractors being used. The use of such composite attributes is itself a means of simplifying the representation of complex situations. However, their usage is more complex than that of simple single-dimension attributes.

Attributes used to characterize complexity

Exhibit 2: Attributes used to characterize complexity


Most organizations have systems that are flexible enough to allow for exceptions. The systems are built to handle most projects, but the managers reserve the right to make exceptions where appropriate3. This is often done as a way of making a simpler system work most of the time, but it adds an element of complexity to what would otherwise be a simpler system.

Difficulties with the Use of Project Categorization Systems in Organizations

Both the focus groups and the web-based questionnaire identified problems with the use of project categorization in their organizations. Exhibit 3 presents the list of problems with scores indicating the percentage of respondents to the web-based questionnaire that identified that this item was a potential problem. As can be seen in this figure, the respondents were almost unanimous in reporting potential problems with the use of project categorization systems. This indicates that there is some problems associated with project categorization, but does not necessarily indicate that it is a problem area in organizations. In several cases, the participants in focus groups were unanimous in their portrayal of their organization's project categorization system as long-standing, widely used, well-adapted, clear, legitimate, and creating value. Even in these organizations, the participants could readily identify problems associated with the use of their categorization systems. Therefore, the high scores for percentages of respondents reporting problems does not mean that a high percentage of respondents feel that the system in their organization needs to be changed significantly or removed. Those designing and using project categorizations systems need to be aware of these potential problems.

Percentage of respondents indicating that this item is a potential problem

Exhibit 3: Percentage of respondents indicating that this item is a potential problem

Purposes and Attributes

The present study makes a clear distinction between the purposes served by or uses made of project categorization systems in organizations and the attributes that organizations use to sort projects into groups. Examples of usages include the grouping of projects in order to identify the level of approval they require, the competencies and training needs of project management personnel, the methods and techniques that will be appropriate to apply to their management, which budget they will be funded from, their alignment with organizational strategy, and the like.

By attributes, we mean the characteristics that are used to sort projects into groups or categories. These include: size, level of complexity, geographical location, technical discipline, etc. Here again, there are many different characteristics or attributes that are used to sort projects into groups or categories.

The difference between purposes and attributes is important and forms the basis of the model presented below. However, this distinction is again an abstraction, one that practitioners find difficult to use in a consistent fashion while discussing their project categorization systems. The confusion between the organizational purposes served by categorization systems and the attributes used to sort projects adds to the difficulty in clearly identifying an organization’s system.

Attributes Not Linked Directly to Purposes

One of the hypotheses of this research project was that there would be a relationship between the purposes an organization is pursuing with the use of its project categorization system and the attributes it would or should use to sort its projects into categories. In other words, an organization pursuing purpose A would choose attributes X, Y or Z, but not N, M, or P. Inversely, organizations categorizing their projects with attributes X, Y or Z would be using the categorization for purpose A but not B. Although the idea is intellectually appealing, it is unsupported by our investigation. Within a specific organizational context, a clear link exists between organizational purpose and the categories in use. However, attempts to map purposes to attributes and vice versa across several organizations have failed. Just with the small sample from the nine focus groups, many examples were found of the same purpose being pursued by different organizations using different attributes to categorize their projects. Likewise, many examples were found of different organizations using the same attributes for different purposes.

For example, the common attribute of geographical region is used by many organizations for many purposes. Some use it to indicate which regional office will do the work, others use it to adapt to differences in regulatory frameworks, and others again use it to align their market penetration strategy. And these are but a few examples. Likewise, many organizations group projects into categories in order to develop specific tools and methods for each category. Many different attributes are used for this purpose. It all depends on the relevant sources of variation among the organization’s projects. For some organizations, the product type or the technology are the primary source of variation; for others, it is again the geographical division with international projects being managed differently from domestic projects; and for yet other organizations, the variation may be primarily by contract type or complexity or level of risk or size.

As can be seen from these examples, there is no direct and systematic relationship between the organizational purpose served by the project categorization system and the attributes that have been chosen to group projects into categories. The relationship between the organizational purposes served by the systems and the attributes that are most relevant is very context-specific. Two organizations pursuing the same objective in different contexts will use different attributes to categorize their projects. This divorce between organizational purposes and attributes makes model building in this field more complex. The model will need to make considerable allowance for adaptation to the organization's specific context.

The lack of a simple relationship between organizational purpose and the attributes used to categorize projects has led to the development of a model with two separate components: one for purposes and another for attributes. We have called these the Organizational Purposes Map and the Attribute Map, respectively. The fundamental idea behind the model is that the design, analysis or modification of a categorization system would require that these two distinct but interrelated aspects be examined. We find that difficulties are encountered when only the system of attributes for sorting the projects into categories is considered. Difficulties are also encountered if the two aspects are confused.

Organizational Purposes Served by Categorization Systems

The focus groups and the web-based questionnaire identified a large number of different purposes that categorization systems serve in organizations. The organizational purposes and the frequency of their identification by respondents to the web-based questionnaire are indicated in Exhibit 4 below. In further support of the use by organizations of project categorization systems for various purposes, 81% of respondents to the web- based questionnaire indicated that personnel are chosen according to project type; 48% indicated that their organization provides training programs to develop skills relating to different types of projects, and over 75% indicated that projects are executed and financial resources are allocated to projects according to their type.

The many purposes that project categorization systems serve in organizations, as identified in the focus groups and the web-based questionnaire, have been used to develop an Organizational Purposes Map. The aim in producing the map is to present a set of organizational purposes that is both complete and logically organized. If the set is complete, an organization could identify all the possible uses it might make of a project categorization system. Organizing the set in a logical fashion makes the information easily available. The information has been organized in a hierarchical fashion as shown in Exhibit 5.

As can be seen in the model, the organizational purposes have been sorted into two primary high-level purposes: strategic alignment and capability specialization. Many more detailed usages have been grouped under these two headings. Usages grouped under strategic alignment are focused on assuring that the organization is doing the right set of projects in that the projects undertaken are aligned with the organizational strategy and with the organization’s capacity to undertake and complete projects. Usages grouped under capability specialization are focused on doing projects right. The focus is on identifying the groups of projects the successful management of which requires similar capabilities and practices. In other words, organizations create categories of projects where it is felt that there is a difference that makes a difference to project performance. Specialization of capabilities can be seen both as an immediate short-term need to ensure that current projects are being appropriately managed, and as a longer-term requirement to develop the capabilities to appropriately manage the organization’s projects in the future and to promote these capabilities. This is represented by the breakdown of this branch of the model into capability alignment and capability development.

Organizational purposes served by categorization systems

Exhibit 4: Organizational purposes served by categorization systems

In addition to the two important high-level organizational purposes under which most other uses can be grouped, another perspective was identified that was best dealt with separately. This is the third branch of the map: promoting a project management approach in the organization. None of the organizations undertook to implement a project categorization system primarily for this purpose. However, several organizations implemented a project categorization system as a part of a larger initiative to implement or reinforce a project management approach in the organization. 55% of respondents to the web-based questionnaire reported that a project categorization system formed part of their organization’s project management methodology.

The Organizational Purposes Map

Exhibit 5: The Organizational Purposes Map

Definitions and discussions of the many organizational purposes have been inserted into The Organizational Purposes Map. We have decided not to present this detail in the body of the paper. This information is integrated into the Map in the form of notes attached to each branch and sub-branch. The Map is presented in upcoming publication by PMI (Crawford, Hobbs & Turner, 2004).

Variation of Usage with Organizational Context

The results of the focus groups and the web-based questionnaire revealed wide variations in the usages made of project categorization systems in different contexts. Very significant differences were observed between, on the one hand, supplier organizations that execute projects that have been selected either through a bidding process in the market or that are requested of them by other divisions of their parent organization; and, on the other hand, organizations that finance their own projects. A contractor, a consultant and an engineering and construction division of a utility or manufacturing organization would be examples of the first of these. Any organization making decisions relative to the allocation of scarce resources to projects to meet its own needs would be an example of the second. The financial institutions in our sample are examples of the second.

In the financial institutions, the project categorization system is very much focused on the usages grouped under strategic alignment. The usages grouped under capability specialization were also present, but as a secondary consideration. Supplier organizations, on the other hand, see project categorization primarily as an issue of capability specialization, both in the short-term view of appropriate management of current projects and in the longer-term view of capability development.

The distinctions among strategic alignment, capability alignment, and capability development are not watertight. For supplier organizations, and to some extent for other organizations, strategy is very much about appropriate allocation of scarce resources including key personnel and the choosing of which capabilities to develop and to market. Supplier organizations do use project categories for purposes of planning, tracking and reporting. However, the context of each organization dictates to what extent project selection is an issue. In the engineering and construction division of a large organization, the project selection decisions may be largely outside of their area of responsibility. In a contractor organization, selection may be largely a question of opportunistically bidding on jobs where the market determines which projects are actually allocated to the firm. However, in one contractor organization in our sample, the issue of exposure to risk was critical. This issue came into play in two ways. First, the project management procedures specify what constitutes a high-risk project that requires senior management approval. And second, the company very carefully manages its exposure to risk by restricting the total amount of exposure it will accept at any one time. This is achieved by controlling the projects for which it has contracts that are fixed price with performance guarantees. As can be seen by these examples, the organizational purposes served by project categorization systems are greatly influenced by the organization's context and strategy.

The Attributes used to Categorize Projects

Systems of project categories in organizations can be very complex. Exhibit 6 presents 37 attributes that were identified from the literature review and focus groups to be used to categorize projects. The data provided in this Table is from the web-based questionnaire. The number of attributes identified varied considerably among respondents, with a mean of eight and a mode of five.

The attributes that organizations use to categorize projects can be seen as building blocks from which elaborate constructions can be built. The authors have attempted to identify these fundamental building blocks from which organizations can and do build their categorization systems. These are presented in Exhibit 7 as 14 major building blocks. The detailed model that is presented in the upcoming publication by PMI gives many concrete examples of labels used to operationalize each attribute (Crawford, Hobbs & Turner, 2004).

Attributes uses to categorize projects

Exhibit 6: Attributes uses to categorize projects

A Guide to the Use of the Model by Organizations

An organization could use this model to analyze its existing project system in order to better understand its functioning. The model could then be used to guide the redesign of the system. This would be the most common usage because almost all organizations already have a project categorization system, in whatever state of formalization it might be. The model could also be used for the design of a new system in a greenfield situation. A complete description of a categorization system would need to cover both the organizational purposes served by the system and the attributes used to group projects into categories. The two would, of course, need to be integrated together in a coherent fashion.

The design or redesign of a system would logically start with the identification of the organizational purposes the system was to serve. This would naturally start at a high level and work down to the detailed level. The organizational purposes map provides a structured set of possible uses and is designed to be used in this way. First, the organization would need to choose from the map the primary uses it wished to make of the system. The next logical step would be to select the attributes that would be the most appropriate for the intended usage, given the specific context and objectives of the organization.

The Attributes Map

Exhibit 7: The Attributes Map

Because project categorizations are used for many different purposes in different parts of the organization, and in different stages of the project life cycle and the organization’s planning and reporting cycles, it is important to verify that the attributes that have been chosen will be useful for these various purposes. It is difficult to imagine that this might be done in isolation from user groups. Quite to the contrary, experience with focus groups during the data gathering and validation phases of this research project indicate that focus groups are a powerful tool for gathering information, coming to common understandings, and validating designs.

The validation of the proposed system should proceed in two steps. First, a conceptual verification by the design team and, second, a validation with organizational stakeholders that will use the system or be impacted upon by its use. Here again the focus group is a powerful tool.

The analysis of an existing system might well start from the opposite side of the model. It could start with a search for the ways in which projects are currently categorized within the organization and the uses that are currently being made of the categorizations. Both the current usages of categorization and the attributes that are currently being applied will most likely be at the detailed level. Moving up the hierarchy in either the organizational purposes or the attributes maps requires a certain amount of analysis and abstraction. The description and analysis will require many cross-validations between the purposes and the attributes, in order to complete the description of the existing system. Once the description of the system has been completed, it can then be evaluated with the intention of making minor improvements or redesigning the system.


There is a clear and important distinction between the purposes that a categorization system is meant to serve and the categories that the system sorts things into. This distinction is clear in the literature on categorization in general (Bowker and Starr, 2000), but not in the project management community. To a large extent, the project management community centers what little attention this topic receives on the identification of appropriate categories into which projects should be sorted, without addressing the issue of what purposes the sorting into categories might serve. This approach is fundamentally flawed and unlikely to produce useful results. Our intention is that the present document be used to sort out the distinction between purposes on the one hand, and categories and attributes on the other. This distinction is necessary if progress is to be made in this area by both researchers and practitioners.


  1. The terms “classification” and “categorization” are very often used interchangeably. The term “categorization” has been used in this document.
  2. Preliminary results of this research project were presented at the 2nd PMI Research Conference (Crawford, Hobbs & Turner, 2002). A complete report of this research project is being published by PMI (Crawford, Hobbs & Turner, 2004)
  3. This is an important part of the control dimension discussed previously and represented in Exhibit 1.


Aalto, T. 2001. Strategies and Methods for Project Portfolio Management. in K. Artto, M. Martinsuo, & T. Aalto (Editors), Project Portfolio Management: Strategic management through projects (pp. 23-60). Helsinki: Project Management Finland.

Archibald, R. D. 2003. Managing High Technology Programs and Projects (3rd ed.). New York: Wiley.

Bowker, G. C., & Star, S. L. 1999. Sorting Things Out: Classification and Its Consequences. Cambridge, MA: MIT Press.

Crawford, L., Hobbs, B., & Turner, J. R. 2002. Investigation of Potential Classification Systems for Projects. Proceedings of the 2nd PMI Research Conference, Seattle, Washington (pp. 181-190). Newtown Square, PA: Project Management Institute.

Crawford, L., Hobbs, B., & Turner, J. R. 2004. A Complete Research Report on the Project Categorization Project, to be published by PMI.

Cooper, R., Edgett, S., & Kleinshmilt, K. 1997. Portfolio Management in New Product Development: Lessons from the leaders II. Research Technology Management, 40 (6), 43-52.

Dvir, D., Lipovetsky, S., Shenhar, A., & Tishler, A. 1998. In Search of Project Classification: A Non-Universal Approach to Project Success Factors. Research Policy, 27 (9), 915-935.

Dye, L. D., & Pennypacker, J. S. 2000. Project portfolio management and managing multiple projects: two sides of the same coin? Proceedings of the PMI Seminars & Symposium, Houston, Texas. Newtown Square, PA: Project Management Institute.

Kwasnik, B. H. 1992. The role of classification structures in reflecting and building theory. Proceedings of the 3rd ASIS SIG/CR Classification Research Workshop: Advances in Classification Research (Vol. III, Chap. October 25, pp. 63-81). Pittsburgh, PA.

Mintzberg, H. 1997. The Structuring of Organizations: a synthesis of the research. Englewood Heights, NJ: Prentice-Hall.

Payne, J. H., & Turner, J. R. 1999. Company-wide project management: the planning and control of programmes of projects of different type. International Journal of Project Management, 17 (1), 55-59.

Shenhar, A. J. 1998. From theory to practice: toward a typology of project management styles. IEEE Transactions on Engineering Management, 45 (1), 33-48.

Shenhar, A., & Wideman, R. M. 1996. Improving PM: Linking Success Criteria to Project Type. Project Management Institute Symposium: Creating Canadian Advantage through Project Management . Calgary: Project Management Institute.

Shenhar, A., & Wideman, R. M. 1997. Toward a fundamental differentiation between projects. D. F. Kocaoglu, & T. R. Anderson (Editors), Innovation in Technology Management - The Key to Global Leadership. PICMET '97: Portland International Conference on Management and Technology: 27-31 July 1997 . Hoboken, NJ: Stevens Institute of Technology.

Taylor, A. G. 1999. The organization of information. Englewood, Colorado: Libraries.

Turner, J. R., Crawford, L. H., & Hobbs, J. B. 2004. Aligning capability with strategy. Proceedings of PMI Europe 2004, Prague . Newtown Square, PA: Project Management Institute.

Youker, R., 2002. The Difference between Different Types of Projects (Revised), Original version (1999) PMI 30th Annual Seminar and Symposium, Philadelphia, Pennsylvania, USA.

Author Contact Information

Dr Lynn Crawford


Department of Project Management

Faculty of Design Architecture and Building

University of Technology, Sydney

Postal Address:

10 Amaroo Crescent

Mosman NSW Australia 2088

Tel: +612 9968 3644

Fax: +612 9514 8875

[email protected]

Dr J Brian Hobbs


Professor in Project Management

University of Quebec at Montreal Business School

Department of Management and Technology

Postal Address:

PO Box 6192,

Downtown Station

Montreal, Quebec H3C 4R2 Canada

Tel: +1 514-987-3000

Fax: +1 514-987-3343

[email protected]

Dr J Rodney Turner


Professor of Project Management,

Faculty of Economic Sciences

Erasmus University, Rotterdam

Postal Address:

Wildwood, Manor Close,

East Horsley, Surrey, KT24 6SA, UK

Tel: +44-(0)1483-282 344

Fax: +44-(0)1483-281 281

E-mail: [email protected]

This material has been reproduced with the permission of the copyright owner. Unauthorized reproduction of this material is strictly prohibited. For permission to reproduce this material, please contact PMI or any listed author.



Related Content