What email address or phone number would you like to use to sign in to Docs.com?
If you already have an account that you use with Office or other Microsoft services, enter it here.
Or sign in with:
Signing in allows you to download and like content, and it provides the authors analytical data about your interactions with their content.
Embed code for: a Multi-phase research program investigating
Select a size
MARCH 2007 PROJECT MANAGEMENT JOURNAL74 Over the last decade, the project man- agement office (PMO) has become a prominent feature in many organiza- tions. Despite the proliferation of PMOs in practice, our understanding of this phenomenon remains sketchy at best. No consensus exists as to the way PMOs are or should be structured nor as to the functions they should or do fill in organi- zations. In addition, there is no agree- ment as to the value of PMOs. Despite the importance of this phenomena and the lack of understanding, there has been very little research on this topic. A three-phase research program has been undertaken in order to develop a better understand of PMOs. This paper pres- ents the research strategy, the overall program, and the results of the first phase of the research. Keywords: project management office (PMO); survey results; research methods ©2007 by the Project Management Institute Vol. 38, No. 1, 74-86, ISSN 8756-9728/03 Introduction I n recent years, many organizations have established PMOs. Dai and Wells (2004, p. 526) showed that PMOs first started to become popular in 1994 and that their number has been growing significantly since. Many books and arti- cles on PMOs have been published in recent years, with the vast majority of the literature produced by practitioners and consultants promoting the implementa- tion of PMOs. This literature is rational, self-evidently correct and normative, as is much of the project management literature (Williams, 2005). Observations of PMOs in organizations contrast quite sharply with the image portrayed in the literature. The population of PMOs is characterized by very sig- nificant variation in: • The structure of PMOs • The roles assumed by PMOs • The perceived value of PMOs. Prior to the undertaking of the present research program, a reliable portrait of the population of PMOs was not available. In addition, an adequate explana- tion of the great variety has yet to be found. The Definition of a PMO A Guide to the Project Management Body of Knowledge (PMBOK® Guide) defines a PMO as: An organizational body or entity assigned various responsibilities related to the centralized and coordinated management of those projects under its domain. The responsibilities of the PMO can range from providing project management support functions to actually being responsible for the direct management of a project. (PMI, 2004, p. 369) This definition is very close to the definition the authors adopted during this investigation. It highlights that PMOs are organizational entities and that their mandates vary significantly from one organization to the next. However, the present study makes a distinction between the multi-project PMO and the sin- gle-project PMO or “project office,” which has responsibility for the manage- ment of one large project. The PMBOK® Guide definition and much of the literature on PMOs include both, and both are important phenomena worthy of ABSTRACT A MULTI-PHASE RESEARCH PROGRAM INVESTIGATING PROJECT MANAGEMENT OFFICES (PMOS): THE RESULTS OF PHASE 1 BRIAN HOBBS, PMP, University of Quebec at Montreal, Canada MONIQUE AUBRY, University of Quebec at Montreal, Canada MARCH 2007 PROJECT MANAGEMENT JOURNAL MARCH 2007 PROJECT MANAGEMENT JOURNAL 75 investigation. Multi-project PMOs and entities responsible for the manage- ment of a single project are quite dif- ferent and can best be investigated separately. The scope of the present investigation includes only PMOs with mandates that cover many projects or “multi-project PMOs.” In part because of the great variety found among PMOs in different organizations, and in part because of the lack of both a consensus among practitioners and adequate descrip- tions in the literature, discussions on this topic tend to be characterized by diversity of opinion and confusion. Many people have been exposed to a limited number of PMOs and have concluded inappropriately that all PMOs are similar to the ones they have observed. The lack of consensus is understandable given (1) that the PMO is a relatively recent phenome- non, (2) that PMOs take on a great variety of forms and functions, and (3) that there has been a lack of systematic investigation. The present investiga- tion employs a rather large definition of the PMO in order to capture the variety of form and function. For the purposes of this investigation, it is not necessary that the organizational unity be called a PMO. PMOs in the Literature Several books and papers have been published on PMOs in recent years. The descriptions of PMOs in the litera- ture are often summarized in typolo- gies comprised of a small number of models. Dinsmore (1999) introduced the earliest typology of PMOs with four types, starting with a single proj- ect entity in which project manage- ment services are developed and used within this single project. The three other models in Dinsmore’s typology are multi-project entities: project sup- port office, project management center of excellence, and program manage- ment office. The Gartner Research Group’s 2000 study (cited in Kendall & Rollins, 2003) proposed one of the most influential typologies of PMOs. The Gartner Group typology is com- prised of three types of PMOs: (1) project repository, (2) coach, and (3) enterprise. Several authors have pro- posed typologies since the publication of the Gartner report, some of whom explicitly reference the earlier work. Within the space restrictions of the present paper, it is not possible to sum- marize these typologies. Table 1 pres- ents a listing of some of the types of PMOs described in the literature, iden- tified only by their names. Some of the typologies identify the single-project entity of “project office,” which is outside the scope of the present study. Each of the typolo- gies proposes two, three, or four multi- project PMOs, organized in an ascending hierarchy. Different authors use different properties to characterize the passage from one level to the next within their hierarchy. The following are among these properties: • Staff functions or line functions with project managers included within the PMO • Organizational scope: covering larger portions of the organization • Level within the organizational hier- archy: from the lower operational level to the top level • Influence and authority: from pas- sive to supportive to enforcing stan- dards to empowered • Operational issues to strategic issues, often associated with a progression from project management to pro- gram and/or portfolio management • Process-driven to business-driven • Project management maturity (culture) within the organization: from non-sup- portive to fully-supportive culture. Each type presented in these typologies is a model of a PMO. Any model is necessarily a simplification and a reduction of the complexities of organizational reality. Models are very useful, even necessary, to support both research and practice. However, the reduction of all or even most multi- project PMOs to two, three, or four types is a radical reduction. The pres- ent investigation does not use the models found in the literature as a starting point. The authors believe that MARCH 2007 PROJECT MANAGEMENT JOURNAL Table 1: Typologies of PMOs in the literature MARCH 2007 PROJECT MANAGEMENT JOURNAL76 it is useful and necessary to put these models aside and to investigate organi- zational reality directly in order to cap- ture the diversity and the complexity of PMOs in practice. The Multi-Phase, Multi-Method Research Program The objectives of the research program are two-fold. The first objective is to pro- duce a reliable description of the present population of PMOs. The second objec- tive is to develop a better understanding of PMOs, of why they take on such a variety of forms, and of the dynamics surrounding their creation, transforma- tion, and action in organizations. It would be difficult to start from the present state of knowledge in which no reliable description of the phenom- ena is available to develop an adequate understanding of PMOs and their roles in organizations. The investigation of PMOs has, therefore, been organized into a three-phase research program. Each phase is a separate project with its own methodological approach. Successive phases build upon the find- ings of previous phases. This approach is motivated by the present lack of knowledge, by the great variety of forms and functions observed, and by the complexity of the organizational phenomena under investigation. The authors adopted the approach suggest- ed by Van de Ven (in press) on engaged scholarship, where the complexity of the subject merits looking at the prob- lem from various angles. The program has been organized into the following phases: 1. A descriptive survey of 500 PMOs aimed at providing a realistic por- trait of the population of PMOs in organizations (2005). 2. The development of a rich concep- tual model to guide further investi- gation (2006). 3. Four in-depth case studies aimed at understanding the dynamics sur- rounding PMOs in their organiza- tional context (2006). 4. A confirmatory study to validate the understanding that will emerge from the previous two years work and modification of the model pro- duced in phase 2 (2007). At the time of this writing, the sur- vey in phase one has been completed. The results are reported in the present paper. The conceptual model has been developed (Aubry, Hobbs, & Thuillier, in press). Data collection for the four case studies has been completed and is in the early stages of analysis. The fol- lowing paragraphs present each phase in more detail. Phase 1: A Descriptive Survey of 500 PMOs Each of the 500 survey responses describes one PMO and its context. Each is a snapshot of a PMO as it was at the time of the survey investigation. The 500 snapshots were analyzed to: 1. Provides a description of the total population and variations in PMO structure, role, and perceived value. 2. To identify common configurations or models that describe significant numbers of PMOs. 3. To identify relationships between the variability of PMOs and the vari- ability of their contexts. 4. To identify correlations between the characteristics of PMOs and their perceived value. A descriptive survey with a large sample is an adequate methodology for describing a population. The primary result of phase 1 was the production of such a description of the population, a description characterized by extreme variety. If the majority of the population can be described by a small number of configurations or models, the analysis of survey data can identify these. However, the present investigation was unable to reduce the population to a small number of configurations. The survey instrument collected contextual data, which was analyzed to identify statistical associations between PMO characteristics and con- textual variables. Intuitively, PMOs can be thought to vary in different contexts and the context can provide at least a partial explanation for the variability found within the population of PMOs. However, no such statistical associa- tions were found. The survey data was largely descriptive. However, the survey instru- ment did include questions as to the perceived value of each PMO. Statistical analysis revealed that some characteristics of PMOs are associated with more highly valued entities. The statistical associations are quite strong and provide some insight into the dynamics surrounding highly valued PMOs. However, they provide only a partial explanation of the performance of PMOs and their contributions to organizational performance. There are advantages and incon- veniences with any methodological approach. Several possible explana- tions can be provided as to why the analysis of survey data does not provide an adequate explanation of such phe- nomena, as is the case here. First, the survey data is limited to the questions on the survey instrument. It is possible that statistical associations exist between PMO characteristics and con- textual variables not included in the survey instrument. Second, the survey only provides descriptive snapshots. It does not reveal the dynamics surround- ing the PMO and its evolution over time. The analysis of survey data gath- ered at one point in time is also limited in that it is very difficult, if not impos- sible, to determine the underlying causal relationships. In the present sur- vey, it is not clear why PMOs with cer- tain characteristics are perceived better than others. Phases 2 and 3 were designed to improve our understand- ing of the PMO phenomenon. Phase 2: The Development of a Rich Conceptual Model to Guide Further Investigation Phase 1 provided a description of the population of PMOs but did not pro- vide an adequate understanding of the dynamics surrounding the PMO nor did it identify the major sources of variability. Phases 2 and 3 of the research program are designed to over- come the shortcomings of a large sam- ple descriptive survey. Here, the PMO is not considered as a standalone enti- ty but rather as an important structural element of the organization in which it is implemented. The unit of analysis passes thus from the PMO to the organization that encompasses it. In MARCH 2007 PROJECT MANAGEMENT JOURNAL MARCH 2007 PROJECT MANAGEMENT JOURNAL 77 this perspective, the PMO is seen as the gateway into the organization in order to study the dynamics of project man- agement in the organizational context and the role of the PMO in these dynamics. “The critical task is to adopt and use the models, theories, and research methods that are appropriate for the research problem and question being address” (Van de Ven, in press). Given the complexity and the rich- ness of the subject being studied and the exploratory nature of the investiga- tion, a constructivist ontology in which the PMO is conceptualized as a dynamic constructed entity has been adopted for phases 2 and 3 of the research program. It is not the purpose of the present paper to provide a com- plete description of the rich conceptu- al model that has been developed (Aubry, Hobbs, & Thuillier, in press). However, it is important to understand the basic premises upon which the model is based, as these condition the balance of the research program. The conceptual model is based on the following elements: • Organizational structures are con- ceptualized as the result of a dynam- ic strategizing/structuring process (Pettigrew, 2003). • An historical and contextual perspec- tive (Hughes, 1987) is adopted for the examination of both: ° The host organization ° The PMO or PMOs. • The dynamic relationships between the PMO and its host organization are conceptualized as co-evolution- ary (Van de Ven & Garud, 1994). • Network structure approach (Hagström & Hedlund, 1999) and actor network theory (ANT) (Callon & Law, 1989) are borrowed from the field of sociology. Both are used to depict the PMO as a network, the for- mer in its structural aspect, and the latter in examining the relationships among the actors involved. • The conceptualization of the organi- zational contribution of the PMO is based on a “competing values approach.” In this approach, organi- zational contribution is seen as a subjective construct rooted in values and preferences of stakeholders (Cameron & Quinn, 1999; Morin, Savoie, & Beaudin, 1994; Pettigrew, 2003). The model includes four rep- resentations intended to provide an overall view of organizational project management performance. The rational goals representation integrates economic value to measure profit, project management efficiency, and return on investment. The open system representation contains variables that measure adaptation and growth. The human relations representation intro- duces considerations of human resource development, cohesion, and morale that are almost invisible in corporate evaluation. The internal process representation captures the measures related to corporate processes linked to project manage- ment such as program and portfolio processes and knowledge manage- ment processes. The theoretical model developed in phase 2 has been used as the basis for phase 3. Phase 3: In-Depth Case Studies Aimed at Understanding the Dynamics Surrounding PMOs in Their Organizational Context In-depth case studies are particularly well adapted to subjects as complex as the one under investigation here, espe- cially when the study is exploratory, as is the case here. Both the survey results from phase 1 and the conceptual model developed in phase 2 were drawn upon in the design of the research instruments for phase 3. These included both interview guides and questionnaires. Extensive data was gathered in each of four organizations in order to produce a rich description of the organization, its PMO or PMOs, and their joint evolution. Three types of data were gathered: 1. Company documents were collected. 2. In-depth, semi-structured interviews were conducted, recorded, and tran- scribed with multiple respondents with different organizational roles. Each interview gathered both factual and perceptual information. 3. Two questionnaires were developed and used. These included a question- naire based on the survey instrument from phase 1 to which several ques- tions were added, and a questionnaire addressing the issue of the organiza- tional contribution of PMOs. The analysis of data from multiple sources provides a rich, detailed, and reliable description of each organization and its PMO or PMOs in their specific context as they evolved together over time. At the time of this writing, the data collection activities have been complet- ed and the analysis is under way. Phase 4: The Confirmatory Study to Validate the Understanding That Will Emerge From the Previous Two Years Work and Modification of the Model Produced in Phase 2 The rich data and in-depth analysis in phase 2 is expected to produce a better understanding of PMOs in these four organizations. The strategy for phase 4 is to draw upon the results of the first three phases and to conduct investiga- tions to both complete and to validate the understanding that will emerge from the analysis of both the survey and case study results analyzed togeth- er. It is too early to be able to describe in detail the exact nature of the confir- matory study that will be carried out in 2007. The balance of the present paper is devoted to the presentation and dis- cussion of the methodology and the results of the survey in phase 1. A more complete presentation of the results can be found in Hobbs (in press). Detailed Methodology for the Survey in Phase 1 Because there has been very little empirical research on PMOs, a reliable portrait of the population of PMOs is not available. The objective of this research is to provide such a portrait. Providing a descriptive portrait is typi- cally an objective of exploratory research into a previously unexplored topic. In this sense, the present research should be considered as descriptive and exploratory. Phase 1 of the research program is a project in and of itself. This project took place in four steps over a two-year period. MARCH 2007 PROJECT MANAGEMENT JOURNAL MARCH 2007 PROJECT MANAGEMENT JOURNAL78 Step 1 was to undertake a prelimi- nary and systematic investigation of 30 PMOs in different organizations and different industries. This was done in 2004. The objective was to provide a preliminary validation of the hypothe- sis that the structures, roles, and legiti- macy of PMOs vary significantly from one organization to the next, and to gather data that would contribute to the production of a richer and more reliable portrait of the reality of PMOs. To this end, a preliminary version of the survey questionnaire was devel- oped and tested. Feedback sessions were held with informants from the organizations to validate and discuss these preliminary results. The prelimi- nary investigations produced an image of PMOs characterized by extreme vari- ety in structures, roles, and legitimacy, while at the same time validating and significantly enriching the question- naire, which became the survey instru- ment. The results from step 1 were enlightening but the sample is small. It does not lend itself to statistical analy- sis and it is impossible to judge how representative this sample is of the general population. Step 2 was undertaken to validate and further enrich the preliminary results from step 1. A web-based survey instrument was designed and tested. The questionnaire had already been validat- ed and tested in step 1, however three respondents from different industries tested the web-based version and a small number of minor adjustments were made. The instrument was designed so that each respondent describes one PMO. The questions were descriptive until the end of the instrument, where a small number of more evaluative ques- tions completed the instrument. Step 3 was the data collection phase. The invitation to participate was available on the Project Management Institute (PMI) website. The authors solicited respondents through several project management networks including the PMI Montreal Chapter’s Community of Practice on PMOs, the PMI Southern Ontario Chapter, PMForum, the American Society for the Advancement of Project Management, and the firms Human Systems and Valence—and with the collaboration of colleagues from the University of Limerick, Athabasca University, University of Technology Sydney, and ESC-Lille. The authors wish to thank all of those whose collaboration made this proj- ect possible. A total of 500 usable responses were received. The respondents were distributed among organizational roles as follows: • Project managers 38% • Managers of PMOs 23% • Professionals in PMOs 11% • Executives and other managers 10% • Consultants 8% • Others 10% The geographical distribution of respondents was as follows: • Canada 43% • United States 26% • Europe 16% • Other 15% The respondents work in a very wide variety of industries. The largest proportions came from the following: • IT/IS 14% • Financial services 14% • Telecommunications 10% Step 4 consists of data analysis and presentation of results, of which this paper is a part. The Survey Results The Name of the Entity The majority of entities described in this study were called “project management offices.” However, many of these orga- nizational entities were given a great variety of other names. The distribution of names is presented in Table 2. Some of the labels used to describe these organizational entities deserve comment. Interestingly, 2% of respondents described entities that exist in their organizations but that have no official label and, therefore, do not appear on the organizational chart. It is quite plausible that these entities have been created to fill a real need, but that their existence has not yet been made official. It is also plau- sible that, because of a previous failed attempt to implement a PMO, or for some other reason, some PMOs are maintaining a low profile. The number of entities bearing the title “project office” is certainly much greater than these results indicate. This label is often used to name an entity responsible for the management of a single large project. The survey instruc- tions asked specifically that informants not refer to this type of unit in respond- ing to the questionnaire. An examina- tion of the 2% of responses describing entities with this label indicates that these were multi-project entities similar to those labeled PMO. They have, therefore, been included in the sample. A total of 12% of responses described entities labeled as program management offices. This group of responses was compared to those labeled as project management offices and no statistically significant differ- ences were found between the two. The program management function is more important for those labeled pro- gram management office, but the dif- ference is not statistically significant. Program management is, therefore, very often part of the role of the PMO, whether it is labeled a project or a pro- gram management office. The analysis that follows is, therefore, based on the entire sample, including both labels. One or Several PMOs Each respondent to the survey described one particular PMO. However, some organizations have more than one. In 53% of the cases, the respondents indicated that the PMO described is the only one in the organization. Of these, 30% were described as central PMOs and 23% as located in a business, functional, or regional unit. Another 25% reported that other PMOs exist but have no rela- tionship with their PMO or its man- date. Finally, 22% described a PMO that is related to at least one other PMO in their organization. The Age of PMOs Most PMOs have two characteristics in common; they tend to be young and to have a small staff. Apart from these two points in common, PMOs vary enor- mously one from the other. PMOs have MARCH 2007 PROJECT MANAGEMENT JOURNAL MARCH 2007 PROJECT MANAGEMENT JOURNAL 79 been popular since the mid- to late- 1990s. Surprisingly, 54% of PMOs in existence today were created in the last two years, according to data from 2005. The Interthink Consulting survey (2002, p. 12) showed the same result: half of the PMOs were less than two years old in 2002. Two phenomena are at work producing this result. First, new PMOs are being created at a relatively high rate. Second, PMOs are being shut down or radically reconfigured at almost as fast a rate. The result is a population dominat- ed by PMOs that have only been in exis- tence in their present form for a few years, as shown in Figure 1. PMO Staff Most PMOs have very little in the way of staffing. Figure 2 shows the staffing levels of PMOs expressed in full-time equivalents, including the person responsible for the PMO, but exclud- ing the project managers. This staff is overhead, and organizations are very reluctant to create overhead expenses. The issue of the cost of overhead is a key issue for PMOs, creating a some- what paradoxical situation in which the PMO is asked to take on many functions with few resources. As this data shows, the vast majority of PMOs have been recently created or restructured. Most have very little staff other than the proj- ect managers. PMOs have very little else in common. Quite to the con- trary, great variety characterizes the population of PMOs described in this survey. On some characteristics, the population displays distribu- tions that are close to being either normal or skewed toward one extreme. In many cases the variance is high. On other variables, the dis- tributions are almost bipolar, with most PMOs at one extreme or the other of the distribution and few in the middle ground. The Decision-Making Authority of the PMO The distribution of decision-making authority is close to a normal distribu- tion, but with very high variance, as shown in Figure 3. PMOs in a passive or supporting role with little or no decision-making authority make up 41% of the sample. At the other extreme, 29% have considerable or very significant authority to make deci- sions to allocate resources, set priori- ties, or initiate, change, or cancel projects. This illustrates the great vari- ety of roles different organizations assign to their PMOs. The Allocation of Projects and Project Managers to PMOs The variation among PMOs as to the per- centage of projects and project managers found within their structures is even more extreme. These distributions, shown in Figures 4 and 5, respectively, show bipolar distributions with more PMOs at each extreme than in the middle ground. In different organizations, the answer to the question “Are project managers grouped within the PMO?” received radically different responses: 31% of organizations reported that they group 100% of the project man- agers in the PMO, while 29% of PMOs had no project managers. These two extremes corresponded to PMOs that are either strictly a staff function with no project managers, or a line function with responsibility for the active management of projects in the hands of their project managers. MARCH 2007 PROJECT MANAGEMENT JOURNAL Table 2: Names of organizational entities 35% 30% 25% 20% 15% 10% 5% 0% 1 year or less 1 to 2 years 3 to 5 years More than 5 years Figure 1: Age distribution of PMOs MARCH 2007 PROJECT MANAGEMENT JOURNAL80 An alternative way of expressing this extreme variation is to note that 46% of PMOs had less than 25% of the project managers within their struc- ture, 40 % had more than 75%, and only 14% had between 25% and 75%. The percentage of projects that are within the PMO’s mandate is also extremely varied. Figures 4 and 5 show the extreme variety in the way organiza- tions structure their PMOs. This variety contrasts with the literature on PMOs that tends to oversimplify reality. Given the extreme variety of forms that PMOs take on in reality, any general statement claiming to describe the decision-mak- ing authority of PMOs or the allocation of project managers or projects to PMOs should be viewed critically. The Organizational Roles of PMOs PMOs fill many different roles or func- tions in different organizations. The interchangeable terms “role” and “function” are used here to identify the content of the PMO’s mandate within the organization. A list of roles or func- tions that are part of the mandates of PMOs was derived from preliminary investigations of a smaller sample of PMOs and from a review of the litera- ture. A large number of different func- tions were identified. The final list contained 27 functions. Several of these functions were added during the process of pre-testing the question- naire. Within the survey, respondents were asked if their PMO filled any functions not included in this list. An analysis of the responses did not iden- tify any functions important for more than a very small number of PMOs. A large number of respondents indicated that the list is complete, a result that can be seen as a validation of the list of 27 functions of PMOs. The respondents to the survey reported the importance of each of these functions for their PMO using a scale ranging from 1 (not important at all) to 5 (very important). Table 3 shows the percentage of PMOs in which each function was scored either of considerable importance or very important. In the minds of many practitioners, PMOs were associated with particular roles or functions. It was not uncom- mon to hear statements such as, “A PMO is an entity that develops and implements a standardized project management methodology.” Table 3 confirms that 76% of PMOs are heavily involved in this function. But to define PMOs by associating them with a par- ticular function or group of functions is out of line with organizational reality. All 27 functions are important for sig- nificant numbers of PMOs, and 21 of the 27 are important for at least 40% of PMOs. This result again illustrates the extreme variety found among different PMOs in different organizations, and the difficulty in providing a simple and accurate description of what they are and what roles they fill. It may seem surprising that 50% of PMOs consider monitoring and controlling the performance of the PMO itself as important. However, this result is consistent with, and likely a MARCH 2007 PROJECT MANAGEMENT JOURNAL Figure 2: Personnel of PMOs excluding project managers (full-time equivalents) Figure 3: Decision-making authority of PMOs MARCH 2007 PROJECT MANAGEMENT JOURNAL 81 consequence of, the fact that the value of PMOs and the justification of the expenses they generate are often brought under scrutiny and ques- tioned. Many PMOs are under pressure to justify their expenses and show value for money. Members of the project manage- ment community recognize most of the functions listed in Table 3 very eas- ily. However, some functions have only recently come into prominence. Program management (48%) and portfolio management (49%) are shown as quite important despite the fact that they only recently became the focus of much attention with the development of “enterprise or organi- zational project management.” Benefits management (28%) is an even more recent phenomenon in the proj- ect management community and liter- ature. Many members of the commu- nity are as yet unfamiliar with this practice, which may explain why it is considered relatively less important. Groups of Functions Analyzing 27 different functions is quite a detailed task. Identifying groups of func- tions greatly simplifies interpretation and use of this data. This can be done con- ceptually by identifying practices that are logically related. For example, reporting project status to upper management requires that project performance be monitored, which can best be done with a project information system and a proj- ect scorecard. These four functions are thus logically related. One would expect to find that PMOs that fill one of these functions would also have a tendency to fill the others. The tendency to fill functions in groups can also be identified and measured through statistical associa- tions. Factorial analysis was used to identify such groupings. Functions that are grouped together through factorial analysis are tightly associated statisti- cally with each other, and statistically independent from the other functions and groups of functions. These inde- pendent groups constitute the dimen- sions of the fundamental underlying structure. The factorial analysis identi- fied five groups of functions. Each group was examined to ensure that it was internally consistent in both con- ceptual and practical terms. These groups show the structure underlying the many roles filled by PMOs in organizations. Identifying groups of functions that are both conceptually and statistically sound has very practical consequences. The long and disorganized list of functions is replaced by a simple structure of underlying high-level roles or functions. These are pre- sented next in decreasing order of the average importance of the func- tions included in the group, which are indicated on a scale of 1 to 5. The average importance is indicated in parentheses for each group. Within each group, the functions are presented in decreasing order of average importance. MARCH 2007 PROJECT MANAGEMENT JOURNAL Figure 4: Percentage of projects within the mandate of the PMO Figure 5: Percentage of project managers within PMO MARCH 2007 PROJECT MANAGEMENT JOURNAL82 Group 1: Monitoring and Controlling Project Performance (3.82) The group of functions related to the monitoring and controlling of project performance is the most important group. This group includes both the monitoring, controlling, and reporting of project performance and the man- agement of the computer-based tools to do these tasks. PMOs with these functions are providing for the infor- mation managers’ needs to maintain visibility and control the performance of projects for which they are responsi- ble. In so doing, the PMO is support- ing project governance functions. The interrelatedness of these functions was previously discussed. • Report project status to upper man- agement • Monitoring and control of project performance • Implement and operate a project information system • Develop and maintain a project scoreboard. Group 2: Development of Project Management Competencies and Methodologies (3.54) The group of functions most tradition- ally associated with PMOs includes functions dealing with tools and methodologies and with competency development. This group is composed of the following functions: • Develop and implement a standard methodology • Promote project management within the organization • Develop competency of personnel, including training • Provide mentoring for project managers • Provide a set of tools without an effort to standardize. The development and implemen- tation of tools and methodology and the provision of project management training and mentoring are the func- tions most people associate with PMOs. The PMO with these functions is often in the role of promoting the use of the methodology, the develop- ment of competencies, and project management in general. This group thus constitutes a coherent set of func- tions that reinforce one another. This reinforcement is the practical reality behind the statistical phenomenon identified by the factorial analysis. Group 3: Multi-Project Management (3.23) Some PMOs have mandates to man- age whole sets of projects in a coordi- nated fashion, which often involves program or portfolio management. These have become important aspects of project management, as signaled by PMI with the publication of the Organizational Project Management Maturity Model (OPM3®) (PMI, 2003) and the publication of standards on program and portfolio management (PMI, 2006a, 2006b). The coordina- tion of interdependences within pro- grams and portfolios is a central issue in multi-project management, as can be seen from the functions in this group: • Coordinate between projects • Identify, select, and prioritize new projects • Manage one or more portfolios • Manage one or more programs • Allocate resources between projects. MARCH 2007 PROJECT MANAGEMENT JOURNAL Table 3: PMO functions in decreasing order of importance MARCH 2007 PROJECT MANAGEMENT JOURNAL 83 Group 4: Strategic Management (3.06) There has been a tendency in recent years for project management in gener- al, and PMOs in particular, to become more involved with issues of strategic alignment and to become more closely tied to upper management. The factor analysis reveals that the following group of functions related to strategic management constitutes one of the underlying dimensions of PMO roles: • Provide advice to upper management • Participate in strategic planning • Benefits management • Network and provide environmental scanning. Involvement in these functions brings project management and the PMO closer to upper management. Networking and providing environ- mental scanning are used to keep abreast of current development so as to give up-to-date advice to upper management. The survey showed that these functions are more typical of central PMOs. Group 5: Organizational Learning (3.00) Organizational learning has been a very important topic in the management lit- erature and practice in recent years. Some PMOs are actively involved in organiza- tional learning through the following group of functions: • Monitor and control the perform- ance of the PMO • Manage archives of project docu- mentation • Conduct post-project reviews • Conduct project audits • Implement and manage a database of lessons learned • Implement and manage a risk database. The last four functions in this group are very directly related to orga- nizational learning. An examination of Table 3 shows them to be among the functions viewed as least important. From this it can be seen that, although organizational learning is of consider- able importance, it is often seen as less important than other functions more directly related to operational or strate- gic issues. The first two functions in this group are related to organizational learning, but can also be deployed in the pursuit of other objectives. Archiving project documentation has important operational aspects. The function to “monitor and control the performance of the PMO” can be seem as part of the learning feedback loop, and thus as closely related to the other organizational learning functions in this group. Recent inter- views with PMO personnel have revealed that some PMOs specifically use the evaluation of the perform- ance of their PMO in an organiza- tional learning perspective. It is, however, conceivable that the meas- urement of PMO performance may also be done in response to question- ing of the expenses generated by the PMO. The overall average importance of this group is influenced positively by the importance these first two functions may have for objectives not directly related to organizational learning. Thus, the average impor- tance of this group may overstate the overall importance of organizational learning for PMOs. Organizational learning is, however, important for a significant number of PMOs. Project management in general and PMOs in particular are participating in the general trend toward the increased importance of organizational learn- ing in the knowledge economy. Additional Functions Not Included in the Groups of Functions The factorial analysis produced the five groups of functions previously present- ed. Three functions not included in these groups complete the list of 27 functions identified in this study. These three functions are excluded from the groups previously listed, not because they are not important, but because their presence is neither statis- tically nor conceptually related to these groups. The remaining functions are presented here in decreasing order of importance. Execute Specialized Tasks for Project Managers (e.g., Prepare Schedules) (3.05) Many PMOs provide specialized servic- es to project managers and project teams. In order to execute these tasks, PMOs maintain specialized resources on their staff. The preparation of sched- ules is a common example, but such services can include many other tasks, such as contract and risk management. Manage Customer Interfaces (2.84) Some PMOs have the responsibility for managing customer interfaces. Responsibility for this activity depends to a great extent on the type of customer. Not all PMOs are in a position to fill this role. On the average, managing the cus- tomer interface is more important for PMOs with customers that are external to the organization. A PMO responsible for all the projects for a given customer may well have an important role to play in managing this customer interface. A PMO responsible for an outsourcing contract is an example of this. MARCH 2007 PROJECT MANAGEMENT JOURNAL Figure 6: “Hasthe relevance or even the existence ofthe PMO been seriouslyquestioned in recentyears?” MARCH 2007 PROJECT MANAGEMENT JOURNAL84 Recruit, Select, Evaluate, and Determine Salaries for Project Managers (2.35) This is the least important function for PMOs, but it remains important for 22% of PMOs. The human resource (HR) department in most organiza- tions carries out these HR activities, but the involvement of a number of PMOs in these activities is considered important in some contexts. The PMOs fit into very different organizational realities regarding HR management relative to project managers. Implications for Theory and Practice The existence of a statistical and con- ceptual link between two or more functions does not mean that they are, or should always be, implement- ed together. The statistical and con- ceptual links are too weak for this to be the case. Organizations must use considerable judgment when decid- ing which functions the PMO is to be mandated to fill. On average, the monitoring and controlling of project performance is the most important group and the les- sons learned group the least important. The rank ordering of the groups of func- tions may be misleading. All are impor- tant, and the differences among them are small. In any particular context, any one of them may be the most impor- tant. However, the number of func- tions in each group varies. In addition, the relative importance of the different functions in each group varies consid- erably, as can be seen from Table 3. This reinforces the need to adapt to the organizational and strategic context when deciding which functions to include within the mandate of a par- ticular PMO. The fact that the underlying high- level functions are statistically inde- pendent of each other is an indication that they identify a fundamental or deep structure. The identification of this underlying structure among PMO functions has profound consequences for both theory and practice. From both points of view, a few high-level functions are much more manageable than the long and unorganized list of possible functions. From the point of view of theory building, the identifica- tion of the structure that underlies the role of PMOs in organizations pro- vides a key to understanding the fun- damental roles of project management and of PMOs in the creation of value in organizational contexts. This ques- tion is at the heart of project manage- ment research at the present time. From the point of view of managers and practitioners, identifying the underlying structure greatly simplifies the task of analyzing and understand- ing existing PMOs and the task of designing or restructuring PMOs. Legitimacy and Performance of PMOs The data on the age of PMOs showed that PMOs are being shut down or rad- ically restructured almost as fast as they are being created. At the end of the survey instrument, after having described their PMO, respondents were asked, “Has the relevance or even the existence of the PMO been serious- ly questioned in recent years?” Forty- two percent of the respondents answered “yes.” This data is illustrated in Figure 6. The reality of PMOs in organizations is even darker than this result indicates. A survey of this type has a positive bias, particularly on eval- uative questions such as this. People who are interested enough to respond to the invitation to participate in the survey tend to have a positive attitude on the topic of the survey. Those unfa- vorable and strongly opposed tend not to respond. In this survey, there is an additional positive bias created by the fact that organizations that have shut down their PMO or have decided not to implement one have not responded to this questionnaire in which respon- dents are asked to describe an existing PMO. The extent of the bias is difficult to estimate, but it is not unreasonable to think that about half of organiza- tions are critical enough of PMOs to decide not to implement one or to seriously consider shutting theirs down if they already have one. This result clearly identifies a lack of consensus in the project manage- ment community. About half of PMOs are seen as legitimate within their organizational context. This level of strong support for PMOs combined with the large number of PMOs currently in existence under- scores the importance of PMOs in project management practice today. On the other hand, the very existence of the other half of PMOs is being questioned. Other questions in the survey confirmed that the issues of value for money and the contribution or lack of contribution to project and program performance are key to the perceived performance and ultimately to the legitimacy of the PMO. Poor- performing PMOs are seen as too costly and as contributing little to project and program performance, while highly valued PMOs are seen as making significant contributions to performance. The ability to show con- tribution to performance at a reason- able cost is critical. The survey results show that PMOs are more legitimate in organi- zations with higher levels of organiza- tional project management maturity. The existence of a correlation between these two variables, organizational maturity and PMO legitimacy, does not reveal the nature of the relation- ship between the two variables. It may well be that the PMO is highly con- sidered in an organization that is mature in project management because project management is val- ued in this organization. On the other hand, it may be that a high-perform- ing PMO has raised the level of proj- ect management maturity in the organization. The relation is likely to be circular and self-reinforcing, with the high-performing PMO contribut- ing to the level of project manage- ment maturity and to the organizational context in which proj- ect management and the PMO are val- ued. The survey also showed that an organizational culture that is support- ive of the implementation of the PMO is associated with the legitimacy of the PMO. This is indicative of this circular relationship. Conclusion The literature promoting PMOs pres- ents them as a best practice with obvi- ous positive effects on project, program, and organizational perform- MARCH 2007 PROJECT MANAGEMENT JOURNAL MARCH 2007 PROJECT MANAGEMENT JOURNAL 85 ance. The reality is quite different. Many PMOs are struggling to show value for money and some are failing, causing a very high mortality rate among PMOs. Practitioners and organ- izations would be well advised not to implement a PMO under naive assumptions of value for money or because PMOs are popular. The results of the survey have shown the hypothesis that “the struc- tures, roles, and legitimacy of PMOs vary significantly from one organiza- tion to the next” to be true. The orga- nizational reality surrounding PMOs is complex and varied. Organizations establish a great variety of different PMOs to deal with their reality. Organizations may decide to include some or all of their project managers within the PMO or they may place them elsewhere in their structures. The PMO’s mandate may cover all the organization’s projects or only a select few. Organizations choose from among a number of possible roles or functions when deciding upon the mandate to give to a PMO. They also choose between a PMO in a support role with little or no authority and a PMO with considerable decision-mak- ing power. These organizational design choices create PMOs of varied form and function. Many different properties can be used to differentiate “types” of PMOs. The results presented here show that PMOs do in fact vary considerably one from another and that the variation is not limited to a small group of proper- ties or characteristics. The population of PMOs shows considerable variation of not just a few, but of many charac- teristics, thus creating a myriad of pos- sible forms that PMOs can and do take on. This creates a population that is difficult to reduce to a small number of models. The Ongoing Program of Research The survey that forms the basis of this paper is the first phase of a three-year research program to investigate PMOs and the dynamics through which they contribute to organizational perform- ance. In collaboration with their col- league, Dr. Denis Thuillier, the authors responded to PMI’s 2005 annual RFP for research proposals and were award- ed a research grant, titled “Modeling Organizational Project Management and PMO Performance.” The second phase of this research program involves four in-depth case studies of PMOs in their organiza- tional context. These commenced in late 2005, and was completed in 2006. The objective of the case studies is to uncover the organizational dynamics that lead to value creation through the use of project manage- ment. The PMO is seen as the point of entry into the organization to study the dynamics of value creation in con- text. The case studies draw upon the results of the survey to establish a rich and reliable representation of the reality of PMOs prior to the detailed investigations. The analysis of the case studies is intended to produce a model of value creation. The third phase of the research program will involve the validation of the findings from the in-depth case studies using a survey of a larger sam- ple of organizations and PMOs. This will be carried out in 2007. The objec- tive is to produce a conceptually rich and empirically grounded model. References Aubry, M., Hobbs, B., & Thuillier, D. (in press). A new framework for under- standing organisational project manage- ment through PMO. International Journal of Project Management. Callon, M., & Law, J. (1989). La proto-histoire d’un laboratoire ou le diffi- cile mariage de la science et de l’économie. Paris: Presses Universitaires de France. Cameron, K. S., & Quinn, R. E. (1999). Diagnosing and changing organi- zational culture: Based on the competing values framework. Reading, MA: Addison- Wesley. Crawford, K. J. (2002). The strate- gic project office. New York: Marcel Dekker. Dai, C. X. Y., & Wells, W. G. (2004). An exploration of project man- agement office features and their rela- tionship to project performance. International Journal of Project Management, 22(7), 523–532. Dinsmore, P. C. (1999). Winning in business with enterprise project man- agement. New York: AMACOM. Englund, R. L., Graham, R. J., & Dinsmore, P. C. (2003). Creating the project office: A manager’s guide to lead- ing organizational change. San Francisco: Jossey-Bass. Garfein, S. J. (2005). Strategic port- folio management: A smart, realistic and relatively fast way to gain sustainable com- petitive advantage. Paper presented at the PMI Global Congress North America, Toronto, Canada. Hagström, P. S. Ö., & Hedlund, G. (1999). A three-dimensional model of changing internal structure in the firm. In A. D. Chandler, Jr., P. S. Ö. Hagström, & O. Solvell (Eds.), The dynamic firm: The role of technology, strategy, organiza- tion, and regions (pp. 166–191). New York: Oxford University Press. Hobbs, B. (in press). The multi- project PMO: A global analysis of the cur- rent state of practice. White paper. Newtown Square, PA: Project Management Institute. Hughes, P. T. (1987). The evolu- tion of large technological systems. In W. E. Bijker, T. P. Hughes, & T. J. Pinch (Eds.), The social construction of techno- logical systems: New directions in the soci- ology and history of technology (pp. 51–81). Cambridge: MIT Press. Interthink Consulting. (2002). State of the PMO 2002. Retrieved February 1, 2004, from http://www.interthink.ca/ research/home.html Kendall, G. I., & Rollins, S. C. (2003). Advanced project portfolio man- agement and the PMO: Multiplying ROI at warp speed. Boca Raton, FL: J. Ross Publishing. Light, M. (1999). The enterprise project office: Beyond 2000, strategic analysis report. Stamford, CT: Gartner Group. Morin, E. M., Savoie, A., & Beaudin, G. (1994). L’efficacité de l’or- ganisation: Théories, représentations et mesures. Montréal, Québec: Gaëtan Morin éditeur. Pettigrew, A. M. (2003). Innovative forms of organizing: Progress, performance and process. In A. M. Pettigrew, R. Whittington, L. Melin, C. Sanchez-Runde, F. A. J. Van MARCH 2007 PROJECT MANAGEMENT JOURNAL MARCH 2007 PROJECT MANAGEMENT JOURNAL86 BRIAN HOBBS, PMP, holds an MBA and PhD in industrial engineering, and is a Project Management Professional (PMP) certified by the Project Management Institute (PMI). He has been a professor at the University of Quebec at Montreal (UQAM) in the Master’s Program in Project Management for more than 20 years. Dr. Hobbs is very active internationally in both the projectmanagementprofessionaland research communities. He served a three-year term on PMI’sStandards Members Advisory Group (MAG) ending in 2002 and joined the Research MAG in July 2006. He is a reviewer for both the ProjectManagementJournal and the International Journal of ProjectManagement.He haspresented manypapersatboth research and professional conferences worldwide. MONIQUE AUBRY, MPM, is a professor at the Business School of Université du Québec à Montréal. She is finishing her doctoral thesis working on PMOs and their relation to organizational performance. Her research project has been sponsored within the PMI’s research program for its theoretical contribution in the field of organizational project management. She hasmore than 20 yearsofexperience in the managementofmajor projectsin the financialsector. She is member of the board of the PMI Montreal Chapter. den Bosch, W. Ruigrok, & T. Numagami (Eds.), Innovative forms of organizing (pp. 331–351). London, UK: SAGE Publications. Project Management Institute. (2003). Organizational project manage- ment maturity model: OPM3 knowledge foundation. Newtown Square, PA: Project Management Institute. Project Management Institute. (2004). A guide to the project manage- ment body of knowledge (3rd ed.). Newtown Square, PA: PMI. Project Management Institute. (2006a). The standard for program man- agement. Newtown Square, PA: PMI. Project Management Institute. (2006b). The standard for portfolio man- agement. Newtown Square, PA: PMI. Van de Ven, A. H. (in press). Engaged scholarship: Creating knowledge for science and practice. Oxford: Oxford University Press. Van de Ven, A. H., & Garud, R. (1994). The coevolution of technical and institutional events in the develop- ment of an innovation. In J. A. C. Baum & J. V. Singh (Eds.), Evolutionary dynam- ics of organizations (pp. 425–443). New- York: Oxford University Press. Williams, T. (2005). Assessing and moving on from the dominant project management discourse in the light of project overruns. IEEE Transactions on Engineering Management, 52(4), 497–508. etitive advantage. Paper presented at the PMI Global Congress North America, Toronto, Canada. Hagström, P. S. Ö., & Hedlund, G. (1999). A three-dimensional model of changing internal structure in the firm. In A. D. Chandler, Jr., P. S. Ö. Hagström, & O. Solvell (Eds.), The dynamic firm: The role of technology, strategy, organiza- tion, and regions (pp. 166–191). New York: Oxford University Press. Hobbs, B. (in press). The multi- project PMO: A global analysis of the cur- rent state of practice. White paper. Newtown Square, PA: Pr