RICS COBRA April 2018 RICS HQ, London, UK

Transcription

1 RICS COBRA April 2018 RICS HQ, London, UK In association with

2 RICS COBRA 2018 The Construction, Building and Real Estate Research Conference of the Royal Institution of Chartered Surveyors Held in London, UK in association with University College London April 2018 RICS HQ, London, UK RICS, 2018 ISBN: ISSN: Royal Institution of Chartered Surveyors Parliament Square London SW1P 3AD United Kingdom The papers in this proceeding are intended for knowledge sharing, stimulate debate, and research findings only. This publication does not necessarily represent the views of RICS or University College London.

3 COBRA 2018 CONSTRUCTION PROJECT PORTFOLIO MANAGEMENT TOOL (COPPMAN) Gozde Bilgin 1, Gorkem Eken 2, Beste Ozyurt 3, Irem Dikmen 4, M. Talat Birgonul 5 and Beliz Ozorhon Department of Civil Engineering, Middle East Technical University, Ankara, 06800, TURKEY 6 Department of Civil Engineering, Bogazici University, Istanbul, 34342, TURKEY ABSTRACT Most of the today s construction organizations are executing several projects concurrently; however, their management routines do not respond to the need of concurrent execution of projects. Portfolio management enables effective management of the right projects for the organization considering its resources, capabilities, strategic objectives, environmental factors, etc. Tools to support decision-making at the portfolio level are needed for successful project management, portfolio management and strategic management of companies; however, there have been very scarce studies responding to this need in the industry. This study reports fundamentals of a tool (alpha version) that has been designed as a decision support tool for portfolio management specific for construction companies. The tool (COPPMAN) will be improved by further testing its validity, reliability and usability within construction companies with real project portfolios. The final version of the tool is believed to be useful for adoption of portfolio management principles in construction companies by capturing project knowledge, enabling analysis of portfolios considering interdependencies between projects, enabling selection of the best portfolio considering strategic priorities of the company and facilitating decision-making by providing visual representations of alternative scenarios. Keywords: construction projects, decision support system, portfolio management, project dependencies, tool. INTRODUCTION Recent improvements and increasing competition in many of today s industries have forced the organizations to execute increasing number of projects simultaneously. As a result of this, traditional single project focused management routines may fall short to meet current requirements of organizations. Decisions based at the single project level may cause problems at the portfolio level (Martinsuo, et al. 2014). Portfolio level analysis is required for the selection and effective management of the right projects for the organizations (Meifort 2016). Thus, organizations need to hold a holistic view on their bunch of projects to achieve a sustainable success and competitive advantage (Kock, et al. 2016; Meifort 2016). As a project based industry, construction industry needs adoption of portfolio management solutions. However, current attempts in the industry have been very scarce and there is a need for an effective solution that would also address the issues in portfolio management. It has been appreciated that strategic alignment and balancing in terms of risk positively affect future profits (Martinsuo, et 1 gbilgin@metu.edu.tr 2 eken@metu.edu.tr 3 besteozyurt@gmail.com 4 idikmen@metu.edu.tr 5 birgonul@metu.edu.tr 6 beliz.ozorhon@boun.edu.tr

4 al. 2014), thus portfolio management functions should address this issue. Additionally, interdependencies between projects are appreciated as sources of uncertainties for portfolios (Martinsuo, et al. 2014), which have generally been underestimated in portfolio management frameworks (Meifort 2016). Besides, knowledge transfer and learning from/between projects are also evaluated as important considerations for project portfolios (Martinsuo, et al. 2014). Some of the provided frameworks also lack industry specific concerns that may affect the success of portfolio management (Meifort 2016). This paper presents a study, which is a granted research project, aimed at supporting portfolio management of construction projects as a response to this need and current issues in portfolio management. Within this context a tool has been generated (COPPMAN - COnstruction Project Portfolio MANagement) that would meet the requirements in this area. The tool is believed to help professionals with its advantages in better visualization and more realistic analysis of portfolios considering dependencies between projects related with resources, risks, learning potential and outputs. RESEARCH BACKGROUND Portfolio management The term portfolio refers to collection of single projects that are consuming the same resources and executed under management of an organization (Kock, et al. 2016). Thus, portfolio management is related with identification of shared demand between the projects and allocation of the available resources considering the projects at hand and the organizational capability (Project Management Institute 2013). Project Management Institute (2008) defines project portfolio management as, the centralized management of one or more portfolios, which includes identifying, prioritizing, authorizing, managing, and controlling projects, programs, and other related work to achieve specific strategic business objectives. The main goal of portfolio management is to maximize contribution of each project in the portfolio to organizational success. Thus, portfolio management is simply the simultaneous management of projects that demonstrate the investment strategy of an organization (Kock, et al. 2016), it establishes the link between formulation and implementation of the organizational strategy (Kopmann, et al. 2017; Meifort 2016). Effective portfolio management requires strategic concerns such as successful strategic alignment, adaptability to internal and external changes, and execution of projects with high value/benefit. Operational requirements are ensuring project visibility, transparency in portfolio decision making, and predictability of project performance (Patanakul 2015). Therefore, with successful utilization of portfolio management processes, decisionmaking about project investments becomes coordinated, risk and resources are balanced and the value of the project portfolio is maximized (Kopmann, et al. 2017). Portfolio management tools Portfolio management is a complicated process since it requires many parameters to be included in the decision-making process at the same time such as strategic objectives, financial returns, project performance, demand conditions, resources, capabilities, risks and so on (Levine 2005; Martinsuo, et al. 2014). The need for development of methods and tools to facilitate portfolio management is often emphasized in the literature (Cooper, et al. 2001; Levine 2005). Developments in

5 computer and software-based tools have diversified the possibilities of visualization and there has been a progressive improvement in information gathering and display options (Dansereau and Simpson 2009). New decision support systems are being developed with the provision of visual tools with flexible cognitive systems (Tergan and Keller 2005). It has been suggested that the use of various algorithms in conjunction with visual methods in decision-making phases of projects has improved the performance of tools (Rivera and Duran 2004). However, it is emphasized in the literature that computer-aided systems have not yet been adequately addressed in project portfolio management and there is a need of progress in this area (Marcus and Colerman 2007). When existing tools and methods developed for portfolio management are investigated, it can be observed that they are produced for different purposes. Cooper et al. (2001) divided the tools used in portfolio management into three categories: mathematical programming tools, classical portfolio tools, and mapping tools. Mathematical programming tools include mathematical models that optimize resources; classical portfolio tools represent tools used for scoring and classification purposes; mapping tools consist of tools that can graphically represent balance of portfolios. Oh et al. (2012) pointed out that there are more than one hundred methods for project portfolio management and categorized them into three main groups as prioritization approach, mathematical optimization approach and strategic management approach. The prioritization approach refers to the prioritization of the returns of the projects as a result of comparative financial analysis (scoring method, analytical hierarchy process, net present value method, etc.). Although this method is the most widely used method, it is insufficient to maintain the portfolio balance since it is based only on financial evaluation. The mathematical optimization approach includes methods in which various functions are optimized by limiting criteria such as resource, project logic and dynamics, technology and project related strategies. Finally, strategic management approach consists of methods that provide a balanced portfolio as a complement to the prioritization approach (bubble chart, portfolio map, etc.). METHODOLOGY The research has proceeded in three main steps which are needs analysis, development of the process model and the tool. The research has started with investigation of related literature and field observations for needs analysis. Following the needs analysis, the process model has been developed by a brainstorming process. A survey was conducted with the participation of 108 company professionals to finalise the algorithms and determine the figures/coefficients (such as similarity coefficients, risk factors, etc.) that will be embedded in the tool. A prototype of the model as a numerical example has been generated by using the current results of the survey and as a result of depiction of the model necessary revisions are made. The final model mainly integrates dependency analysis and knowledge management to project portfolio management, while providing assessment of risks at the portfolio level and strategic prioritization. The revised prototype has been presented to the company for evaluation, and a survey on modules and requirement specification based on the current model has also been held in the company. The model has been realized by generation of COPPMAN, where service procurement has been provided for its generation. It has been developed within a spiral model, which allows generation in

6 sprints where each sprint follows evaluation of the previous one, by joint contribution of the research team and the developer company on possibilities. During development process, several algorithms have been generated for operation of the tool. Within this context, calculation methods are provided for assessments of similarity and dependency between projects, learning potential of a project, predictions based on past project information, risk and strategic fit of a portfolio, project/portfolio and dependency map properties, and necessity of a warning. Details of these calculations are beyond the scope of this study; however, details of dependency assessment are provided in study of Bilgin et al. (2017) and further details of knowledge management module are provided in study of Eken et al. (2017). The alpha version of the tool representing the example of paper prototype is provided to an expert panel for evaluation of the model and the first release. Verification of the tool has been made with black-box testing and pilot testing of the alpha version is done in addition to validation by the expert panel. Usability testing and actual implementation with a construction company are the current future studies of the tool evaluation process (further studies are depicted in grey in Figure 1). Figure 1: Methodology COPPMAN The contribution of the tool lies mainly in its following properties: utilization of previous project knowledge to portfolio analysis, similarity and dependency analysis between projects, analysis of risks at the portfolio level, and incorporating strategic fit

7 into analysis of portfolio value. Details of the tool are presented below starting with the conceptual framework and modules, and following with process model details. Conceptual framework Main objective of the conceptual framework of the model is conversion of projects to portfolios with minimum input by the user (see Figure 2). User is required only to check the project information and carry out analysis at the portfolio level. Therefore, the project knowledge captured in the database by the project team and managers constitutes some of the main components of the framework. Database serves for provision of the supportive information, which is valuable information for current projects retrieved from the previous projects, and also for automatic calculation of project dependencies. Thus, analyst should investigate the supportive information (as retrieved information of similar projects, lessons learned, predictions and learning potentials) and evaluate current projects prior to analysis in terms of their risks and strategic fits. Project dependencies are automatically included in the portfolio analysis process as a part of the portfolio risk, and the user gets portfolio level measures such as portfolio risk, portfolio strategic fit, and other portfolio properties in addition to the information of projects within the portfolios. Thus, at the end, the user obtains possible portfolios that can be formed with respect to the current projects in hand, and receives some useful advice about selection and management of these portfolios. Figure 2: Conceptual Framework Modules According to the conceptual framework, the tool is decided to be designed with five main modules as building blocks of the required main system (see Figure 3). The modules and their roles in the system are as presented below.

8 Figure 3: Modules System Management Module: is required to establish a user management system including defining specific roles, users and their authorization to support establishment of a system to successfully utilize the tool. Knowledge Management Module: encapsulates the requirements for managing both explicit and tacit knowledge in terms of post project information and lessons learned from projects. Various data retrieval options are to be provided as predictions for specific project (based on post project review information) and related lessons learned. Further information of similar projects and learning potential of each project are to be provided for investigation before assessment of projects. Risk Assessment Module: provides customizable risk evaluation forms to assess risk of each project, and these risk scores would be further utilized in portfolio risk. Risk evaluation histories of projects would also be kept for evaluation upon request and a system for keeping risk evaluations up-to-date at the time of analysis would also be included. Strategic Assessment Module: provides customizable strategic fit evaluation forms to assess strategic fit of each project, and these strategic fit scores would be further utilized in portfolio strategic fit as in risk assessment module. Strategic fit evaluation histories of projects would also be kept for evaluation upon request. Portfolio Analysis Module: enables automatic formation of portfolios. Following this, project and portfolio properties together with dependency maps of each portfolio would be presented through tables, bubble diagrams and bar charts, together with warnings on portfolios upon either selection or management of portfolios. Process Model COPPMAN identifies projects under three categories as completed, on-going, and potential projects. Since it needs completed project information to successfully represent the past and establish new portfolios accordingly, all the valuable information should be loaded to the system. Then the tool helps user to extract this information to evaluate the current projects as on-going and potential projects. Once evaluations are completed, user can proceed with portfolio analysis and obtain summaries and warnings on possible portfolio scenarios (see Figure 4). Complete process model of the tool may be expanded on three sections as data input for

9 establishing the database, data analysis for displaying supportive information and evaluating current projects, and data output for portfolio analysis results. Data Input: User should make the necessary adjustments first to use the tool effectively then enter the required information. Necessary process for establishing the database may be summarized in the following order: Project Inputs: COPPMAN provides ready-to-use drop-down lists that are editable for company specific use to ease the data entry process. Preferences: To provide flexibility, the tool lets the user change some numerical figures used in calculations and evaluation criteria used in the process. Setting the roles and authorization of users are also required at the beginning to manage users of the tool for establishing an effective utilization where every user accesses only to the necessary functions/operations of the tool. Project Entry: Once the inputs and the preferences are set, then projects are to be entered. There are standard forms for every project type that asks minimum information required for the projects. The forms are equipped with dropdown and recall lists to minimize the effort, whereas some free text areas are provided to leave the user free to describe the case. Only the completed project form includes post project appraisal section additionally to make user complete this information. Lesson Learned Entry: COPPMAN also forces the user for entering the lesson learned for completed projects; however, user is left optional to enter lessons learned for current projects in case of an intent of timely capture. A single form is designed for lesson entry to ensure simplicity while a tagging system is provided to make lesson entries applicable to all types of projects and experiences. A hierarchical tag tree of around 2000 concepts in construction is provided in an editable format to make the user freely classify and effectively retrieve the lessons. Lesson entry form itself provides some level of classification in terms of type, effect, and responsible party of a lesson. Then user is further asked to search related tags where they are highlighted in the tree with their location in the hierarchy. Thus, user enters a particular lesson according to the intended level of categorization by assigning tags (Eken, et al. 2017). Data Analysis: COPPMAN directs user to investigate available past project information before evaluation of current projects by several retrieval mechanisms. Supportive information section provides user to display the following:

10 Figure 4: Process Model Similar Projects: investigation of similar projects where all completed projects are automatically sorted according to calculated similarities to the current project at hand. This ability attracts user to the projects that need attention. User can investigate project card where the information is summarized or directly reach to the entered information. Lessons Learned: that may be most beneficial for the project in hand are retrieved by several mechanisms as filtering-based, similarity-based, and tagbased search. Thus, user is reminded with the considerations of past. Predictions: information entered in the post project appraisal section are presented to the user as predictions where mode or average of the entered information are provided based on selection of either similarity-based or filtering-based results. Predictions provides some level of insight according to the scope of the project. Learning Potential: attracts user by a score that indicates the project that may be more beneficial in terms of learning since it requires a comparatively new execution process for the company and may also be critical since it includes further risk. Following investigation of available information, user can proceed with risk and strategic fit evaluations for the current projects. Both these areas have an evaluation list and scoring in its default; however, user can set his own evaluation factors and criteria. In both cases, COPPMAN calculates a risk/strategic fit score over 100 and uses this value in calculation of risk/strategic fit scores for the portfolios. Average of the scores are taken for portfolio strategic fit, whereas COPPMAN integrates dependencies to average risk score to obtain portfolio risk. Additionally COPPMAN keeps and presents histories of these evaluations and warns user to update his/her risk evaluation prior to portfolio analysis. Data Output: COPPMAN establishes all the portfolio alternatives considering potential projects. Potential project combinations are added to the current set of on-going projects. Once the user selects the potential projects to be included in the analysis, all the portfolio options are obtained and user can investigate the portfolios and select one of them. Details are as provided below:

11 Portfolio Alternatives: are summarized in table of portfolio properties as potential projects in the portfolio, average risk and strategic fit scores of the projects, network density of dependency map of the projects, portfolio risk score, portfolio strategic fit score, portfolio success potential (value indicates minimum portfolio risk), portfolio value (integrated value of portfolio strategic fit and success potential), and portfolio profit (average adjusted profit of the projects where expected profit is adjusted according to past project financial data). Bubble diagram of alternatives is provided indicating risks, strategic fits, and adjusted profits of the portfolio alternatives where each bubble opens the details of the selected portfolio. Portfolio values are displayed in a bar chart representing the portions of portfolio strategic fit and success potential. Corresponding change in the portfolio profit according to change in the portfolio value is demonstrated through bubble diagram and bar chart, where unit change bar chart graph is also distinguishable. Finally, a warning on portfolio selection based on this profitvalue change equilibrium is provided to the user. After investigation of the portfolios at once, user can further investigate a portfolio in detail. Portfolio Details: open up with a table that reminds the properties of the selected portfolio and summarizes the project properties as project name, start and end dates, status (on-going or potential), risk score, strategic fit score, centrality value, and details button that directs user to the project card. As in bubble diagram of portfolio alternatives, risks, strategic fits, and adjusted profits of the projects are displayed on a bubble diagram where each bubble opens the related project card. The most important ability of the tool as dependency map is located in this section. The calculated dependencies are presented through arrows between project nodes in different colours each indicating a different dependency. Project nodes are drawn in different sizes comparatively according to the level of their centralities that further represents criticality of a project in the portfolio. Magnitudes of dependencies and centralities are provided by pop-up information boxes and matrix table of the dependencies also opens up on the graph. Finally, warnings on the portfolio at hand is provided based on dependency map network density (for critical attention to dependencies), centrality of projects (for most critical projects), completion percentage of projects (for consideration of projects that are close to completion), profitability of projects (for projects with low profit), risk score of projects (for projects with high risk), clients/partner companies as critical actors (for clients/companies that caused problem previously), financial dependency of projects on same currency/client (for major financial dependency), resource dependencies of projects (for consideration in work/procurement planning), learning dependency of projects (for establishment of information transfer) and outcome dependency of projects (for consideration of the dependent result of the projects) (Bilgin, et al. 2017). Portfolio Selection: is provided by sorting the alternatives according to the priorities of the user. User can sort the alternatives according to risks, strategic fits, portfolio values or adjusted profits to direct his/her focus on a smaller set of portfolio alternatives. Expected Benefits In brief, COPPMAN is designed to establish portfolios based on information and evaluations at project level, and to integrate past project information and project

12 dependencies to the analysis. Considering the presented functions, the tool has the ability to display portfolios and provide support for either revealing the current portfolio or selecting a new project/portfolio (examples of diagrams presented in analysis are provided in Figure 5). The tool can be deemed as easy-to-use since processes prior to analysis requires considerations only at project level. Thus, COPPMAN has a potential to assist construction companies in adapting portfolio management solutions. Successful utilization of the tool may provide companies to establish their corporate memories and use their knowledge to manage their portfolios in accordance with the strategies, which may further provide efficient construction and resource management. The tool design may also be adapted to other project-based sectors, since it requires an easy to follow project-based procedure and creates a difference by integrating knowledge and handling dependencies issue in portfolio management. CONCLUSION Figure 5: Portfolio Analysis Outputs A portfolio management tool for construction projects has been generated according to literature review and company observation to respond to the current need in the industry. It will be improved with further testing for its usability and a real application process in a construction company will enable evaluation of its actual benefits while revealing need of any further considerations for revision. Adoption of the tool by a construction company and its utilization in management process may reveal vital improvements required for maximizing its benefits. This version of the tool is believed to serve as a benchmark for development of any further studies in this area. ACKNOWLEDGEMENTS This research is funded by a grant from the Scientific and Technological Research Council of Turkey (TUBITAK), Grant No. 213M493. TUBITAK s support is gratefully acknowledged. REFERENCES Bilgin, G., Eken, G., Ozyurt, B., Dikmen, I., Birgonul, M. T., and Ozorhon, B. (2017) Handling project dependencies in portfolio management In: Proceedings of the Fifth ProjMAN - International Conference on Project MANagement, November , AIS Affiliated Conference, Barcelona, Spain.

13 Cooper, R. G., Edgett, S. J., and Kleinschmidt, E. J. (1999, July). New product portfolio management: Practices and performance. Journal of Product Innovation Management, 16[4], Cooper, R. G., Edgett, S. J., and Kleinschmidt, E. J. (2001). Portfolio Management for New Products. Perseus Pub: Cambridge, MA. Dansereau, D. F., and Simpson, D. D. (2009, February). A picture is worth a thousand words: The case for graphic representations. Professional Psychology: Research and Practice, 40[1], Eken, G., Bilgin, G., Dikmen, I., and Birgonul, M. T. (2017) Knowledge-based portfolio management: A taxonomy for lessons learned In: Pellicer, E., Adam, J. M., Yepes, V., Singh, A., Yazdani, S. (eds) Proceedings of ISEC-9: Resilient Structures and Sustainable Construction, July , ISEC Press, Valencia, Spain, C-13. Kock, A., Heising, W., and Gemünden, H. G. (2016, April/May). A contingency approach on the impact of front-end success on project. Project Management Journal, 47[2], Kopmann, J., Kock, A., Killen, C. P., Gemünden, H. G. (2017, May). The role of project portfolio management in fostering both deliberate and emergent strategy. International Journal of Project Management, 35[4], Marcus, A. M., and Coleman, D. 2007, Collaborative PPM Gets Software-as-a- Service, Project Management.com, [Online] Available at: Martinsuo, M., Korhonen, T., and Laine, T. (2014, July). Identifying, framing and managing uncertainties in project portfolios. International Journal of Project Management, 32[5], Meifort, A. (2016, June). Innovation portfolio management: A synthesis and research agenda. Creativity and Innovation Management, 25[2], Oh, J., Yang, J., and Lee, S. (2012, August). Managing uncertainty to improve decision-making in npd portfolio management with a fuzzy expert system. Expert Systems with Applications, 39[10], Levine, H. A. (2005). Project Portfolio Management: A Practical Guide to Selecting Projects, Managing Portfolios and Maximising Benefits. Jossey & Bass: San Francisco. Patanakul, P. (2015, July). Key attributes of effectiveness in managing project portfolio. International Journal of Project Management, 33[5], Project Management Institute. (2008). A Guide to the Project Management Body of Knowledge. Project Management Institute, Inc: Newton Square, PA. Project Management Institute. (2013). The Standard for Portfolio Management. Project Management Institute, Inc: Newton Square, PA. Rivera, F. A., and Duran, A, (2004, August). Critical clouds and critical sets in resource constrained projects. International Journal of Project Management, 22[6], Tergan, S.-O., and Keller, T. (2005). Knowledge and Information Visualisation. Springer-Verlag: Berlin.