performance evaluation techniques Douglas C. Bower Ryerson University, Toronto, Canada and Government of Ontario, Toronto, Canada, and Andrew D. Finegan Entrepreneurship, Commercialisation and Innovation Centre, University of Adelaide, Adelaide, Australia Abstract Purpose The purpose of this paper is to describe a Doctor of Project Management research study including summary of the literature review, the application of the combination of case study, survey and theory building research methodologies, key research ndings and potential areas for future research. Design/methodology/approach The research investigates the reasons for the limited adoption of earned value management (EVM) as a project performance evaluation technique. It proposes new extensions to this technique that will be benecial to project management practitioners. The multifaceted research approach incorporates the following elements: a review of previous and current literature on EVM; a survey of project management practitioners on their practices and attitudes towards EVM; analysis of the known challenges of the EVMtechnique; development of techniques to address andresolve the EVMchallenges; consolidationof those techniques intoa single frameworkandimplementationmodel; and validation of that framework and model through multiple methods. Findings The research conrms that EVM can be greatly enhanced and simplied though three key initiatives: include the cost assurance (i.e. risk transfer) provided by procurement contracts; measure project achievement and progress on the completion of each phase, rather than monthly; and combine the above into a simplied, single model. Originality/value This paper provides practitioners with an insight into how EVM can be enhanced and applied inproject management organisations. Inparticular, the integrated PAVAtechnique should be particularly useful to projects using the rolling wave approach, as its recognition of phases provides a framework for short- and long-term planning. Keywords Project management, Project evaluation Paper type Research paper Summary of the research thesis Many organisations have adopted the management of projects, programs and their portfolios as their preferred approach for a wide range of initiatives related to newproduct development, strategy implementation, and business transformation (Winter et al., 2006). The current issue and full text archive of this journal is available at www.emeraldinsight.com/1753-8378.htm Thesis URL link: http://adt.lib.rmit.edu.au/adt/uploads/approved/adt-VIT20080130.135140/public/ 02Whole.pdf Project performance evaluation 435 Received January 2009 Accepted March 2009 International Journal of Managing Projects in Business Vol. 2 No. 3, 2009 pp. 435-444 qEmerald Group Publishing Limited 1753-8378 DOI 10.1108/17538370910971072 Despite that progress within organisations, the conceptual base for project management theory has attracted criticism for its lack of relevance to practice (Morris, 1994, 2000) and also for its failure to contribute signicantly to improved performance of projects across various industries. A particular target for this criticism is the earned value management (EVM) technique. This is somewhat of a paradox. EVM is widely recognised as a core project management technique with obvious benets; however, its utilisation is not widespread beyond a few specic industries, notably the defence and aerospace organisations in the USA. Supporting this observation, Fleming and Koppelman (2004, p. 1) recount their conversationwiththe leadeditor of the HarvardBusiness Review(HBR), towhomtheyhad provided information on EVM in response to his inquiry about the topic. Apparently, he wanted assurances that EVMwas for real. After a fewweeks to reviewthat material, the HBR editor called back and asked this key question: If EVM is so good [. . .] why isnt it used on all projects? Fleming and Koppelman (2004) then cited three underlining reasons why, in their opinion, EVM has not been universally accepted on most projects. Their rst reason is that EVM advocates often speak in a foreign tongue, and refer to the obtuse acronyms (e.g. BCWS) that were originallyemployed. This has some validity, but it is not clear that the revisedversions (e.g. plannedvalue PV) will make a major difference. Their second reasonis because initiallythe DoD(Department of Defence) denedEVMto acquire major systems. In other words, EVM is meant for major defence projects not regular ones. There is a great deal of validity to this observation. Conventional EVM as codied in US standards may be overkill for most projects. Third, they content that sometimes management [. . .] doesnt really want to know the nal cost! This statement is borne out by the fact that EVM in the USA has been promoted mainly by nancial controllers, such as those at the US Federal Ofce of Management and Budget (Evans, 2005), rather than by practicing project managers or directors. In recent literature, some authors (Williams, 2005) have suggested that many of the problems inherent in project management practice (cost overruns, delays, etc.) reside in the prescriptive, functionalist, and quantitative tradition inherited by project management from the narrow perspective of operations management. For example, the contributors to Making Projects Critical (Cicmil and Hodgson, 2006a) further expanded research on project management, by considering it within a wider organizational and societal context. The practical experience of project managers conrms this perspective; it is not possible to successfully complete complex and dynamic projects through simply adhering to normative methodologies. Managers must also view their projects within a broad environment that includes the goals of the organisation and the needs of society. That broader perspective, however, does not relieve researchers of the opportunity or obligation to further improve current project management methodologies such as EVM. This thesis suggests that some researchers can develop and demonstrate more effective project management techniques in the area of project performance evaluation while others may advance knowledge and practices in less-prescriptive areas. Project management is a relatively immature area of management, one that can benet from re-examination of existing models as well as the development of new and varied ones. Other authors (Cicmil and Hodgson, 2006b) have asserted that the decision to study and investigate a management topic in a particular fashion involves a philosophical choice by the researcher on what is important. This is based on the IJMPB 2,3 436 premise that what one decides to study has methodological consequences (Holstein and Gubrium, 1995, p. 72). Cicmil (2006) further contends that the chosen theoretical approaches will determine the way in which the subject phenomenon will be represented. EVM has become the predominant theoretical tradition for the evaluation of project performance and progress over the past several decades. That key role is underlined not only by its prominence in most PM textbooks and bodies of knowledge, but also its codication within the US Government Requirements and Standards (NDIA, 1998). It provides a narrow perspective on the project status, as it evaluates performance and progress using a metrics approach in which just three measures are derived, all based on cost or other resource calculations. That narrow focus is a weakness; obviously the EVM indicators and forecasts are only as good as the quality of the information on which they are based. Further, EVM only considers that internal performance of the project that is, how well the project is performing in relation to its cost, time, and scope objectives. EVM does not address the performance of the results of the project, nor the impact of those results on society, the environment, etc. That said, the narrow focus if EVM may also be a strength, as it allows the underlying concepts of EVM to be readily understood, and limits the performance evaluation process to quantiable components. The question posed by the HBR editor If EVM is so good [. . .] why isnt it used on all projects? has been adopted as the central problem for this thesis, and has led to the formulation of the following research questions: RQ1. What is the basis and foundation of the earned value (EV) methodology? RQ2. To what extent is EVM utilised and accepted by project management professionals? RQ3. What are the strengths of EVM in the management of projects and programmes? RQ4. Does EVM have serious challenges, issues or aws that may be slowing its adoption? RQ5. What new concepts or approaches could address those challenges, and further enhance or alter EVM? RQ6. Could those enhancements to EVM be combined to form a valid new methodology? RQ7. Would project managers accept a new EV methodology? Yin (2003, p. 21) identies these ve components of research design as being particularly important: (1) a studys questions; (2) its propositions, if any; (3) its unit of analysis; (4) the logic linking the data to the propositions; and (5) the criteria for interpreting the ndings. Project performance evaluation 437 The units of analysis for this thesis are represented by the measures and indices within conventional EVM methodology as described in the previous section. Their validity and effectiveness have been analysed and conrmed by numerous studies identied in the thesis. Through the development of hypothesis and models, this thesis adopted the established EVMunits of analysis, and developed complementary units. In this thesis the new phase-assured value analysis (PAVA) methodology introduced the related measurement units of assured value and expected cost. Inthe phase-earned value analysis (PEVA) approach, the units of phase planned value, phase earned value, and phase expected cost are introduced. These PEVA values and calculations are illustrated in Tables I and II. In this thesis, the data or information is represented by the new algebraic expressions that have created to express the value relationships inherent in two new models: AVA and PEVA. Further data are contained within the combined model (PAVA) that provides the benets of both new techniques. The logic linking that data to the propositions is contained in the analysis and discussion that applies four types of validation to verify these new models: mathematical verication, practitioner evaluation, compliance with standards, and a retrospective case analysis. Context of the thesis This research study was the nal and examinable submission of work for the Doctor of Project Management (DPM) degree at RMIT University in Melbourne, Australia. The 1 Phase planned value (PV p ) is the total budget for a specic phase 2 Phase EV (EV p ) is equal to the phase PV, once that phase is completed 3 Phase AC (AC p ) is the total of all internal and external costs attributed to that phase 4 Phase cost variance (CV p ) is the phase EV less the phase AC: CV p EV p 2AC p 5 Phase cost performance index (CPI p ) is phase EV divided by phase AC: CPI p EV p =AC p 6 Cumulative cost variance (CCV) is cumulative EV less cumulative AC, for all completed phases: CCV CEV 2CAC 7 Cumulative CPI (CCPI) is the cumulative EV divided by cumulative AC, for all completed phases: CCPI CEV=CAC Source: Bower (2007, pp. 140-1) Table I. PEVA cost measures and indicators 1 Project start date (PSD) is the planned and actual date for the initiation of the project 2 Planned phase end date (PPED) is the calendar date on which the phase is planned to nish 3 Actual phase end date (APED) is the point when the phase deliverables are accepted as 100 percent complete 4 Phase schedule variance (SV p ) is the difference between planned and actual completion dates for that phase 5 Phase schedule performance index (SPI p ) is the current planned duration (in days), divided by the current actual duration, both counted from the project start to the end of the last completed phase 6 Forecast phase end date (FPED) is the calendar date on which an incomplete or unstarted phase is now forecast to nish, based on the planned duration multiplied by the SPI P for the last completed phase Source: Bower (2007, p. 141) Table II. PEVA time measures and indicators IJMPB 2,3 438 doctoral programis structuredto provide a distance-learningapproach(usinginformation communication technologies such as blackboard and e-mail) together with telephone and voice over the internet protocol facilities for communication. This doctorate is targeted at practising project managers who can take part in the program from anywhere in the world. DPM candidates enter the programwith a masters degree and a minimumof eight years of experience in PM environments. The candidate was attracted to the DPM program as it provided a exible but rigorous academic environment in which he could develop ideas, concepts and models that would address some of the questions related to project management such as the paradox of EVM being high in value, but low in usage. The program design facilitates not only the discipline that is necessary for performing doctoral level research, but also the collaboration that is possible within the community of practice that is formed by two dozen project management practitioners, some with decades of varied experience. The structure of the DPM program is shown in Figure 1. In this case, the PM procurement and ethics course, together with its reective learning course, initiated an interest in examining the relationship between EVM and procurement. An initial paper on the variation of EVM when it is applied in projects with different forms of contract and delivery scenarios focussed thinking in this area. In undertaking this work in research preparation 1, the candidate became curious to know why EVM did not take into account the added certainty or assurance that was provide by procurement techniques, particularly in the forecasting of the expected total project cost. Simply put, if a project manager had just signed one or more rm xed-price contracts for work to occur in the remained of the project, why did not EVM formulae include that information? That research resulted in the development of the assured value analysis (AVA) model, and the presentation of the resulting paper at two project management conferences at the end of 2004 (Bower, 2004a, b). Figure 1. DPM program outline Reflective learning (12 credit points) PM knowledge management (12 credit points) Research preparation 1 (24 credit points) Research thesis 1 (Draft) (48cp) Research thesis 2 (Submission) 36cp Reflective learning (12 credit points) Research preparation 2 (24 credit points) PM leadership (12 credit points) Reflective learning (12 credit points) Research preparation 3 (24 credit points) PM Ethics & procurement (12 credit points) Reflective learning (12 credit points) Research preparation 4 (24 credit points) PM practice 2 (12 credit points) Research thesis 3 (Clean-up) 12cp D P M
d i r e c t o r
Y 1 S u p e r v i s o r
f o r
y e a r s
2
a n d
3 48 credit points = part time year 96 credit points = full time year Project performance evaluation 439 Similarly, while proceeding with research preparation 2, the candidate returned to his interest in project phases to question why the EVM methodology disregarded phases in any of its concepts and calculations. That line of inquiry led to the development of the PEVA concept, which was presented at several major conferences in 2005 and 2006 (Bower, 2005, 2006a, b). This progression of research and inquiry is shown in Figure 2, which represents the progression that a DPM candidate may experience in the program. The candidate began in the lower left quadrant (Q1) studying the as is situation regarding project procurement, phases and performance evaluation. During the program he moved to Q2 in proposing new concepts and techniques that could or should be utilised. From there, he continued to Q4 in developing new approaches to evaluating the progress and performance of a project, specically the new models for AVA and PEVA. This enabled several reective learning papers to be written including Bower and Walker (2007) that explored the concepts of rollout projects that linked to this research. Discussion and conclusions Project management is a relatively new area of business management studies, one that is still struggling to establish its legitimacy and importance within academic institutions and within the business community. That struggle has been reinforced by the efforts to establish standards by professional associations such as Project Management Institute (PMI) and the International Project Management Association. It is further supported by the increasing level of project management research that has been carried out by practitioners and educators. Project management standards and research are mutually dependent and interrelated. Research results can be widely recognised and endorsed by professional standards, and those standards may be legitimized by the results of rigorous research. In a new area such as project management, this need may be more obvious that in an established area of science. However, research also has the potential to contradict Figure 2. DPM program contribution N e w
c o n c e p t / T o o l / P e r f o r m a n c e
m e a s u r e m e n t
a g e n t s New application/Geographical area Low High High Auditing as is Situation analysis Mapping/plotting Reflection Modifying, studying new context influences and refining, action learning Developing new systems, new approaches, new tools Q1 Q2 Proposing could or should be Modifying + trial & error, action learning Q3 Q4 Knowledge management PM leadership PM procurement + Ethics PM practice 2 DPM program contribution IJMPB 2,3 440 standards, and rigid enforcement of standards can wrongly imply that further research is unnecessary. EVM has been developed and promoted within the project management community as the singular methodology for project performance measurement, particularly for cost control. In the USA, a major organisation has invested signicant effort in creating a national standard for EVM (NDIA, 1998), and the dominant PMI (2005) has created an EVM Practice Standard. Both of these documents tend to reinforce the notion that there is only one acceptable way to measure project performance, and that way is EVM. The only signicant exception is the recent recognition given to earned schedule as an interesting variant for improving on project progress measurement. In this thesis, PAVA is developed as an alternative to conventional EVM techniques, and in doing so have challenged the idea that EVM is the only suitable way to assess project cost performance. This research has opened the door not onlyto PAVA, but also to the possibility that other improvements may be made by other researchers in this area. The PAVA model takes note of the earned schedule concept and recent research on that initiative; however, it does not incorporate the earned schedule approach into PAVA as this study found it to be procedurally dubious and unnecessarily complex. Rather, by simplymeasuring the progress of each project phase with the PAVAtechnique, managers can readily identify the schedule variance expressed in days, and also arrive at a reliable schedule performance indicator. As PAVA is an intuitively simple and effective technique, it may also be adopted due to its value as a communication tool. To this end, PAVA facilitates phase end reviews, at which the status of the project may be assessed. Therefore, PAVA assessment can be a key component of the reporting that is delivered to key stakeholders, such as project executives and client groups. The results produced by PAVA lend themselves to graphic portrayal not only as trend lines to accompany the PMB S-curve, but also as horizontal bar charts that permit the viewer to quickly visualise the project status. When applied to a portfolio of projects, these PAVA diagrams encourage side-by-side comparison of a range of current projects. This serves to not only inform key stakeholders, but also to support decision-making within a program portfolio. Over time, the simplicity and features of PAVA should lead to increased understanding, acceptance and implementation of it in a range of industries and project types. It is suggested that adoption of PAVA is likely to occur in two major arenas: (1) PAVA may be utilised as a compliant EVM system by organisations that must adhere to the performance measurement standards that the USA, the UK and possibly other government agencies require. The research demonstrates that PAVA is fully compliant with NDIAs (1998) 32 criteria for an EVMS. Several implementation routes can be surmised. Some organisations may chose to implement a form of PAVA for all of its performance measurement needs, as it produces results inconformitywith accepted standards for EVM. Other rms may decide to use PAVA on a trial basis to quickly implement performance measurement, with the option of eventually easing into the full use of conventional EVM. Of course, such rms may well decide that PAVA addresses all of their needs. Finally, some rms may already be using conventional EVM, but realise that theycanachieve all of their performance measurement objectives with PAVA, and devise a process for converting to it. Project performance evaluation 441 (2) PAVA may also be adopted by organisations that do not need to comply with specic project management standards. The recognition of the real benets to be realised from improved control of cost and time will drive this adoption. Where there is apprehension about the formidable administrative burden of conventional EVM, PAVA might be more readily accepted if it is promoted within such organisations as a completely new approach, and not simply as a variant on standard EVM techniques. There are a few situations in which PAVA may not be appropriate. A project that cannot be readily divided into project phases would obviously be a poor candidate for PAVA; however, it is difcult to imagine that any signicant project would not have several phases. Similarly, it would be awkward to implement PAVA on a project with many phases that are scheduled to end at the same time. Such a situation would denitely be highly unusual, and might be addressed through re-planning of those phases so that they are divided into sequential sub-phases. PAVA would not be appropriate on a project for which no cost information is available. For example, in some organisations projects that are undertaken by internal staff do not have formally recognised budgets. Certainly, this would make PAVA implementation impossible, but the same would be true for EVM implementation or any other form of cost control. Since PAVA includes the value and cost of procurement contracts, it would appear more valuable in organisations with projects that include a signicant degree of procured services and deliverables. However, no complications result when PAVA is used on a project with no procurement that aspect of the model is simply not utilised. In the interest of simplicity, it would be advantageous to use PAVA consistently for all projects, so that procurement agreements would be captured in any projects that include them. As already noted, EVM has not been widely adopted on architecture, engineering and construction projects. Owing to the maturity of the construction industry, it has evolved alternate mechanisms for cost and schedule control. Those mechanisms may actually approximate the PAVA concept, particularly if they include the AC of vendor contracts, and their budgeted value, in the cost to complete calculation. The recognition of phases in PAVA facilitates its application to projects that utilise the rolling wave approach to project planning. While one phase is underway, detailed planning can be performed on the next phase, and future phases may remain planned at a high level of both budget and schedule. By the same token, PAVA combined with rolling wave is highly appropriate for projects that are not fully dened at the outset. Budgets allocations and target milestones can be identied for the long-term phases, then revisited as the project proceeds and its true nature becomes more clearly understood by the organisation, project team and client. Finally, this research into conventional EVM and the development of new and improved techniques for project performance measurement was undertaken to conformto the intentions of the DPM program. The thesis describes the creation, description and validation of PAVA, and makes a unique and valuable contribution to project management. Therefore, it is most gratifying that at the RMIT Universitys research awards in November 2008, Bower (2007) was awarded the University Research Prize in recognition of the achievement of excellence in this thesis. This is a tting tribute to the researchers inspiration and effort, and to the DPM program. IJMPB 2,3 442 Future research As the name implies, PAVA addresses three aspects of project management: project phases, assurance (reduced uncertainty) and EV. Of these three topics, project phasing is an area that deserves more research attention. Phases are unique to projects or similar endeavours, and may be used to frame all aspects of project planning and implementation. This is potentially a fertile eld for further research. Uncertainty is an underlying component of project management, one that is addressed through many established techniques: scheduling, budgeting, quality control, etc. In this thesis, procurement is selected as an element of PAVA, not simply because procurement practices address uncertainty (through risk transfer), but also because they can be expressed as values (i.e. budgets and costs), and therefore can readily assist in predicting a revised project nal cost (EAC). However, there are other risk management techniques that should be incorporated into project performance assessment. Some exploratory work has been undertaken (Hillson, 2004) on the integration of risk management predictive techniques (e.g. Monte-Carlo analysis) into conventional EVM. These techniques have not been incorporated into PAVA at this point, as the intention was to address the shortcomings of EVM one of which was its apparent administrative burden. Further research may be warranted in this area, to determine if risk management techniques should be used to predict a range of possible outcomes for the nal cost and completion date of the project phases, or of the project as a whole. References Bower, D.C. (2004a), Assured value analysis: earned value extended, paper presented at the PMI Global Congress North America 2004, Anaheim, CA, October. Bower, D.C. (2004b), Assured value analysis: EVM meets procurement, paper presented at the 16th Annual International Integrated Program Management Conference, Tysons Corner, VA, November 14-17. Bower, D.C. (2005), Phase earned value analysis, paper presented at the PMI Global Congress North America, Toronto, September. Bower, D.C. (2006a), Phase earned value analysis: a proposal for simplifying yet enhancing EVM, paper presented at the PMI Research Conference 2006, Montreal, July 16-19. Bower, D.C. (2006b), Phase earned value analysis: a proposal to simplify yet enhance EVM, paper presented at the Business Performance and Project Management Summit, London, April 27. Bower, D.C. (2007), New directions in project performance and progress evaluation, unpublished research, RMIT University, Melbourne. Bower, D.C. and Walker, D.H.T. (2007), Planning knowledge for phased rollout projects, Project Management Journal, Vol. 38 No. 3, pp. 45-60. Cicmil, S. (2006), Understanding project management practice through interpretive and critical research perspectives, Project Management Journal, June, pp. 27-37. Cicmil, S. and Hodgson, D. (2006a), Making Projects Critical, Palgrave, New York, NY. Cicmil, S. and Hodgson, D. (2006b), Making projects critical: an introduction, in Cicmil, S. and Hodgson, D. (Eds), Making Projects Critical, Palgrave, New York, NY, pp. 1-28. Project performance evaluation 443 Evans, K.S. (2005), Improving Information Technology (IT) Project Planning and Execution, Ofce of Management and Budget, Executive Ofce of the President, Washington, DC, Memorandum. Fleming, Q.W. and Koppelman, J.M. (2004), If EVM is so good . . . Why isnt it used on all projects?, The Measurable News, Spring. Hillson, D. (2004), Earned value management and risk management: a practical synergy, paper presented at the PMI Global Congress North America, Anaheim, CA, October. Holstein, J.A. and Gubrium, J.F. (1995), The Active Interview, Sage, Thousand Oaks, CA. Morris, P.W.G. (1994), The Management of Projects: A New Model, Thomas Telford, London. Morris, P.W.G. (2000), Researching the unanswered questions of project management, paper presented at the PMI Research Conference, London. NDIA (1998), ANSI/EIA-748 Guide: Earned Value Management System (Standard), American National Standard Institute/Electronic Industry Association, Washington, DC. PMI (2005), Practice Standard for Earned Value Management, Project Management Institute, Newtown Square, PA. Williams, T. (2005), Assessing and moving on from the dominant project management discourse in the light of project overruns, IEEE Transactions on Engineering Management, Vol. 52 No. 4, pp. 497-508. Winter, M., Smith, C., Morris, P.W.G. and Cicmil, S. (2006), Directions for future research in project management: the main ndings of a UK government-funded research network, International Journal of Project Management, Vol. 24, pp. 638-49. Yin, R.K. (2003), Case Study Research: Design and Methods, Sage, Thousand Oaks, CA. Corresponding author Douglas C. Bower can be contacted at: dbower@ryerson.ca IJMPB 2,3 444 To purchase reprints of this article please e-mail: reprints@emeraldinsight.com Or visit our web site for further details: www.emeraldinsight.com/reprints Reproducedwith permissionof thecopyright owner. Further reproductionprohibited without permission.