Abstract
Project portfolio management refers to a set of processes and practices to manage a group of projects in order to meet strategic business objectives. The main focus of project portfolio management until now has been on project selection and prioritization and on the strategic alignment of projects. Little mention is made of potential disturbances to project portfolios. This is the main topic of the research presented in this paper, which attempts to answer the research question: How is uncertainty affecting project portfolios managed in dynamic environments?
The conceptual framework draws from two main theoretical traditions. According to the concept of dynamic capabilities from strategy theory, resources and capabilities must be constantly re-allocated to adapt to changing environments. The dynamic capability framework, which includes the basic processes of sensing, seizing, and transforming is supplemented using Weick's interpretative sensemaking theory.
Two large multi-divisional corporations are the research sites. One organization is a high-technology multinational company with its headquarters in Europe. The other is a financial services company based in Canada. The methodology is inductive and exploratory. At the time of acceptance of this paper, the data collection phase was completed and the analysis begun, but only preliminary results are available. More results will be presented at the PMI Research and Education Conference in July 2010.
The results of this research indicate that the sources of change go beyond the two groups identified in The Standard for Portfolio Management (PMI, 2006), that is, (a) Portfolio Performance and (b) Business Strategy Changes. Even with the limited sample of four portfolios, a large number of types of changes were identified. Portfolio performance was shown to be an important cause of change. However, business strategy changes were not observed during the lifetime of the four portfolios.
The sensing mechanisms put in place by both companies primarily addressed uncertainty related to project scope. These included a separate role for content specification, various forms of product customization and prototypes, and, in the case of the software development company, a separate process to handle the analysis of requirements in a pre-study machine.
Although the results are preliminary, the findings contribute to an enriched view of the reality of portfolio management in dynamic environments. In many cases, the results expand upon and, in some cases, contradict the often implicit assumptions upon which much of the literature on the topic is based.
Keywords: Project Portfolio Management, uncertainty, dynamic capabilities, change, risk
Introduction
This paper addresses the question: How is uncertainty affecting project portfolios managed in dynamic environments? The project portfolio management (PPM) literature provides a framework for describing the overall PPM processes, but does not provide an adequate framework for the study of changes to the project portfolio between periodic review cycles, a phenomenon which is common in dynamic environments. Because the PPM literature does not deal extensively with the management of uncertainty or change, the literature on the management of uncertainty is reviewed. The review draws on different streams of the business strategy and organization theory literatures in which the issue of uncertainty management is addressed. Because of space limitation, the complete review is not presented herein. But from this review, a novel framework based on concepts derived from the dynamic capability and sensemaking literatures was developed and is summarized in this paper. Following the presentation of the theoretical framework, the research methodology is presented. The final sections of the paper are devoted to the presentation and discussion of preliminary results and a conclusion.
Project Portfolio Management
In the 1970's, research and development enterprises started to develop different quantitative decision models to support their project selection and allocation of resources between projects, much of which is based on mathematical programming and modeling (Henriksen & Traynor, 1999; Linton, Walsh, & Morabito, 2002). However, most empirical research shows the very limited use of these models in practice (Baker & Freeland, 1975; Cooper, Edgett, & Kleinschmidt, 2001; Henriksen & Traynor, 1999; Liberatore & Titus, 1983).
The most significant literature on PPM was developed in the study of new product development portfolios (Cooper, et al., 2001; Krishnan & Ulrich, 2001; McGrath, 2004) This empirically-based literature focuses on the project selection process and choices among many potential projects. The concepts of selection criteria, balancing and strategic alignment are central to this literature. The Standard for Project Portfolio Management—Second Edition (Project Management Institute, 2008a), is based on the same concepts and has much the same focus. The standard defines a project portfolio as: “a collection of projects or programs and other work that are grouped together to facilitate effective management of that work to meet strategic business objectives.” (Project Management Institute, 2008b, p. 138) This standard proposes a process that stresses the importance of the alignment of the project portfolio to the firm's strategy, as well as the identification and prioritization of the projects being fundamental to ensure that firms execute the most beneficial projects. This concept is analogous to financial portfolios but the primary focus of PPM is on how to select and prioritize projects to ensure that risks, complexity, potential returns, and resource allocations are balanced and aligned to the corporate strategy in order to provide optimal benefits to the enterprise.
Up till now, the PPM literature has made little mention of potential disturbances to the portfolio typically found in dynamic environments. There are only two types of changes addressed by the process described in the PMI standard: (a) portfolio performance, and (b) business strategy changes. In the first case, the PMI standard proposes to periodically review the portfolio performance “to ensure that the portfolio contains only components that support achievement of the strategic goals. To achieve this, components must be added, reprioritized, or excluded based on their performance and ongoing alignment with the defined strategy in order to ensure effective management of the portfolio” (Project Management Institute, 2008b, p.77) The second type relates to significant changes in the business environment resulting in a new strategic direction: “as environments inside and outside the organization change, criteria for determining the composition and direction of the portfolio may also change… When the need for new criteria becomes evident, the portfolio management team needs to examine the current criteria in the strategic plan and move ahead with appropriate changes, usually focusing first on categorization. If strategic change is not occurring, the efforts should focus on portfolio balancing.” (Project Management Institute, 2008b, p.84)
It is not argued that the current processes and governance framework are incorrect, but just incomplete. It is therefore suggested that the existing processes be supplemented with additional empirical information. The assumption for this research is that portfolio managers might not only monitor changes but might also implement processes to manage and control change. In addition, adjustments to the ongoing portfolio might be made without going through a complete review cycle, an activity briefly mentioned in the PMI standard in the section on Communicate Portfolio Adjustment. The Association for Project Management (APM) mentions this type of change in terms of “adjustments of the portfolio with regard to the constraints, risks, and returns anticipated, and in the light of developing circumstances around the portfolio.” (Association for Project Management, 2006, p. 8)
Risks, Changes, Deviations, Unexpected Events, and Uncertainty
Risks and Risk Management
Risk management has been one of the core knowledge areas in project management for many decades. Literature abounds in this field (Chapman & Ward, 1997; Jaafari, 2001; Kendrick, 2009; Persson, Mathiassen, Boeg, Madsen, & Steinson, 2009; Raz, Shenhar, & Dvir, 2002; Wideman, 1992) and most general books on project management include at least a section on risk management (Andersen, 2008; Dinsmore & Cabanis-Brewin, 2006; Gray & Larson, 2008; Kerzner, 2006; Nicholas, 2004). A number of techniques have been developed to assess the probability of occurrence and the potential impacts to projects. A typical classification of risks is based on the level of knowledge about the possibility to foresee the risk (known or unknown) and the level of knowledge about the occurrence and impact (known or unknown) (Cleden, 2009). Risk management includes the different techniques to either reduce the probability of occurrence of an event or reduce its impact on the project (or inversely for positive risks.) It includes activities to identify, to assess, to plan a response, and implement a response.
Portfolio risk management is not mentioned in the first publication of The Standard for Portfolio Management (Project Management Institute, 2006) but the following subprocesses were included in the second edition (Project Management Institute, 2008b): (a) Identify Portfolio Risks, (b) Analyze Portfolio Risks, (c) Develop Portfolio Risk Responses, and (d) Monitor and Control Portfolio Risks. The standard proposes a definition for portfolio risk that is very similar to that for project risk:
An uncertain event, set of events or conditions that, if they occur, have one or more effects, either positive or negative, on at least one strategic business objective of the portfolio. (Project Management Institute, 2008b, p.139)
The techniques proposed to analyze and develop risk responses at project portfolio level are similar to the techniques identified in A Guide to the Project Management Body of Knowledge (PMBOK® Guide)— Fourth Edition (Project Management Institute, 2008a) for single projects, that is, avoidance, mitigation, transfer, and acceptance where the risks are defined as uncertain events which might have impacts on scope, schedule, resources, and cost.
Changes, Deviations and Unexpected Events
Instead of studying risks, some authors studied the different types of changes and deviations affecting projects and the techniques used to handle them once they occur. For example Hällgren and Maaninen-Olsson (2005) distinguish between risks, deviations, and changes based on the type of management action (i.e., proactive or reactive). While a risk is a known, yet unrealized situation, changes refer to “realized situations with a significant divergence to the project plan. In contrast to risks, changes are not addressed in advance, meaning that changes are managed when a situation has materialized, being reactive in nature.” (Hällgren & Maaninen-Olsson, 2005, p. 18) A deviation is defined as “a situation, regardless of consequence-positive or negative, large or small- that deviates from any plan in the project.” While changes focus on major project plans, deviations could be related to any level and any portion of the plans including operational day-to-day plans. (Hällgren & Maaninen-Olsson, 2005, p. 18)
Söderholm (2008) prefers to use the term unexpected events. He identifies three categories of unexpected events appearing in projects: re-openings caused by stakeholders redefining some of the project parameters, revisions to plan to improve its accuracy and adapt to events, and finally daily fine-tuning (i.e., adapting the day-to-day work to changing environments). These three categories are based on level of impact and how they are dealt with.
Uncertainty Management
The term risk refers specifically to events rather than being associated to more general sources of uncertainty. In projects undertaken in rapidly changing environments where uncertainty may be unavoidable, managers need to go beyond traditional risk management, adopting roles and techniques oriented less toward planning and risk management and more toward flexibility and learning (De Meyer, Loch, & Pich, 2002; Platje & Seidel, 1993).
The concept of uncertainty facing organizations is not recent and has frequently been studied in organization theory, psychology and economics. The term environmental uncertainty has been used both as a descriptor of the state of organizational environments and as a descriptor of the state of a person who perceives himself/herself to be lacking critical information about the environment. The former implies that it is possible to characterize environments in terms of how objectively uncertain they are; the latter implies that environmental uncertainty ought to be studied as a perceptual phenomenon. (Milliken, 1987) Some authors have advocated starting using the broader concept of uncertainty management instead of risk management, which is too focused on threats and events (Cleden, 2009; Perminova, Gustafsson, & Wikström, 2008; Ward & Chapman, 2003). “Uncertainty management is not just about managing perceived threats, opportunities, and their implications. … It implies exploring and understanding the origins of project uncertainty before seeking to manage it, with no preconceptions about what is desirable or undesirable.” (Ward & Chapman, 2003, p. 98-99).
The notion of uncertainty is very important in the context of PPM. Daft and Armstrong (2009) and Duncan (1972) showed that, although a dynamic environment is not the only source of uncertainty, changes in the environment combined with high complexity always lead to increased uncertainty. An uncertainty management perspective draws attention to the need to understand and manage variability in organizational activities that have impacts on a number of projects. This perspective highlights the need to put in place different approaches to address some aspects of project-related uncertainty outside individual projects.
Collyer and Warren (2009) surveyed the literature to identify approaches that might be used to deal with projects in dynamic environments. The classification that they propose includes: (a) environment manipulation—making dynamic static, (b) planning approaches for dynamic environments, (c) scope control for dynamic environments, (d) controlled experimentation, (e) lifecycle strategies, (f) management controls: input, behavior and output, (g) diagnostic, belief, interactive, and boundary, (h) culture and communication for dynamic environments, (i) categorization, and (j) leadership style. It is not clear that all the approaches described above are applicable to portfolios or whether new approaches should be developed to manage uncertainty at that level.
The following section presents the conceptual framework. It is based on Teece's dynamic capabilities framework and Weick's sensemaking.
Conceptual Framework
The publications on dynamic capabilities theory argue that it is no longer sufficient to develop unique resources or capabilities, as initially proposed in the resource-based view, to gain a strategic advantage but that these resources and capabilities must be constantly re-allocated and re-optimized to adapt to changing environments. (Eisenhardt & Martin, 2000; Helfat, 2007; Zollo & Winter, 2002)
Killen, Hunt and Kleinschmidt (2008) linked PPM to dynamic capabilities by defining the project portfolio management capability as “a dynamic capability consisting of people, structures, and processes that are continually monitored and adjusted to meet the changing requirements of the dynamic environment.” (p.336). They focused primarily on the organizational learning mechanism aspects associated with the implementation and use of PPM. Their findings based on a research comparing service and manaufacturing environments suggest a positive relationship between structured PPM capabilities and improved outcomes. (Killen & Hunt, 2010)
Teece (2007, 2009) conceives the concept of dynamic capabilities as the ability of adapting, integrating, and reconfiguring clusters of resources to match the requirements of a changing environment and defines dynamic capabilities as:
The particular (nonimitability) capacity business enterprises possess to shape, reshape, configure, and reconfigure assets so as to respond to changing technologies and markets and escape the zero-profit condition. Dynamic capabilities relate to the enterprise's ability to sense, seize, and adapt in order to generate and exploit internal and external enterprise-specific competences, and to address the enterprise's changing environment (p. 87–88)
The dynamic capabilities framework identifies classes of relevant variables and their interrelationships. It is made of three main capabilities: (a) to sense and shape opportunities and threats (b) to seize opportunities, and (c) to maintain competitiveness through enhancing, combining, protecting, and when necessary, reconfiguring the business enterprise's intangible and tangible assets. (Teece, 2009)
The three classes of activities (sensing, seizing, and reconfiguring) defined in dynamic capabilities is reminiscent of Weick's decomposition of organizing into three separate subprocesses: (a) enactment (the process to create the information that the system adapts to), (b) selection (of sets of rules based on past experience), and (c) retention (i.e., integration of new items based on recent experiences) (Weick, 1969, p.91). These three levels are very similar to the grouping used by Teece (i.e. sensing, seizing, and transforming).
In this research Teece's framework was used to structure the micro-foundations of the dynamic capabilities used when managing project portfolios under high levels of uncertainty. However, due to the numerous analogies with the sensemaking model, some of the ideas proposed by Weick were integrated, primarily in the areas of sensing and seizing. The conceptual framework depicted in Figure 1 builds on these theories and identifies the concepts that were investigated in this research. It is composed of three main components: (a) organizational context, (b) dynamic capabilities, and (c) the micro-foundations to be investigated. A more extensive description and discussion of the development of the framework based on the work of Teece and Weick is available in another paper (Petit, 2009).
Figure 1: Conceptual framework to study the management of uncertainty in project portfolios.
Organizational context. The organizational context was analyzed to understand why the project portfolio is put in place and under which organizational constraints it must operate. It includes the following elements: (a) environment, (b) strategy, (c) organizational structure, (d) constraints, (e) governance, and (f) project portfolio characteristics.
Sensing. “Organizations must develop information processing mechanisms capable of detecting trends, events, competitors, markets, and technological developments relevant to their survival.” (Weick, 2001, p. 242-243) The type of information, the frequency of sensing, the filtering mechanisms will depend on the level of dynamism of the environment; basically in what areas and how often changes are expected.
According to Teece (2009) the micro-foundation of sensing might include (but is not limited to) the following processes: (a) to identify target market segment, changing customer needs; (b) to tap in exogenous technology; (c) to tap innovation (from suppliers and complementors); and (d) to direct internal R&D and select new technologies.
Similarly Weick treats the environment as information which must be processed in the organization. These informational inputs are typically ambiguous, uncertain, and equivocal. Equivocality refers the multiple interpretations that can be given to a situation. To study the sensing mechanisms, Weick (2001) proposes a framework based on two parameters: the level of organizational intrusiveness (passive or active) and the assumptions about the environment (analyzable or unanalyzable). The more passive approach of conditioned viewing refers to the interpretation of information which is easily analyzable, and which is based on routine collection of data. In undirected viewing, organizations will make little use of formal management information and data will be irregular and casual. In discovering, resources will be allocated to the data acquisition activities, and reports will regularly be produced to senior managers. When enacting, organizations will be more active and will influence and sometimes even create the environment itself through different mechanisms such as innovation and experiments.
Seizing. Teece (2009) defines seizing as the structures, procedures, designs, and incentives for identifying that changes are required once a new technological or market opportunity is sensed. The management of project portfolios involves a number of decision bodies and decision rules which are normally (but not always) defined in the project portfolio governance structure of the enterprise. The primary focus in this area is on how organizations seize opportunities and decide what to do in the face of changes and uncertainty. The main areas investigated are: (a) business model used, (b) governance rules, and (c) decision making protocols. In a context of PPM, selecting the decision-making protocols refers to the decision making by the managers and project portfolio committees, which are put in place to manage and decide on the different components of the project portfolio.
Transforming and Reconfiguring. The third and last capability in the framework is called transforming and reconfiguring. In the context of PPM, this was preferred to Teece's managing threats and transforming. In the face of changing environments, the enterprise might have to reconfigure and reassign existing capabilities and potentially develop new ones. In practice this involves changing the routines of the enterprise. “Routines help sustain continuity until there is a shift in the environment. Changing routines is costly, so change are not (and should not be) embraced instantaneously. Departure from routines will lead to heightened anxiety within the organization, unless the culture is shaped to accept high levels of internal change.” (Teece, 2009)
Methodology
The proposed research methodology has the following characteristics: (a) qualitative study, (b) descriptive rather than explanatory, (c) prestructured in its conceptual framework but evolving depending on early findings, (d) inductive, (e) focusing on processes, (f) in a limited number of well selected cases, and (g) using grounded theory.
Unit of Analysis
The unit of analysis is the processes involved in managing the changes in specific instances of portfolios within organizations dealing with dynamic environments. The focus is on the different aspects involved in sensing events, interpreting them and deciding the changes to be made in the project portfolio. This involves determining the actors involved, the type of decisions they make, their interactions, and the typical relationships between inputs and outcomes.
Cases Investigated
The research was limited to the in-depth study of a limited number of cases. This is one of the best approaches when very little is known about a topic. According to Yin (2003) case studies are the preferred strategy when “how” or “why” questions are being posed, when the investigator has little control over events, and when the focus is a contemporary phenomenon within some real-life context. The research question meets all three of these conditions. The selection was first based on what Patton (2002) calls criterion sampling where the following criteria where used to assess the cases:
- Firms must have mature processes for the management of individual projects,
- Firms must have dynamic environments with a high level of uncertainty and/or high volume of changes to their project portfolio, and
- Organizations must have some form of process to manage their project portfolio including some mechanisms to handle changes. The process must be in place for at least two years.
Two firms meeting the selection criteria have accepted to participate in the research. Despite similarities related to the selection criteria they come from two different industries and have different governance characteristics thus displaying an element of variation in their environment, which will allow comparison. The firms selected are not extreme cases. They manifest the phenomenon intensely (but not extremely). This corresponds to what Patton (2002) calls intensity sampling.
The firms to be investigated are very large and manage many project portfolios. The sampling was further specified within the organizations with respect to the characteristics of the portfolios according to the following criteria:
- The portfolio has been in existence long enough (i.e., more than two years) to have encountered different types of changes,
- The portfolio is sufficiently complex,
- The project and portfolio management practices are well established,
- There is access to documents and to people involved, and
- The history of the portfolio is well documented.
The data collection included approximately 45 interviews of people involved in the management of these fours portfolios plus the analysis of the supporting documentation (minutes of meeting of steering committees, list of changes, multi-project plans, project final reports, etc.). This covered the complete life of all four portfolios (ranging from 3 to 5 years) with the main focus on the most recent 24 months. At the time of writing this paper all interviews have been carried out and the data analysis is ongoing.
Company Soft
The first company referred to as Soft is a large multinational with over 60,000 employees out of which approx. 15,000 work in the R&D divisions where the investigation was carried out. The enterprise has a long history of managing software and hardware projects with well-documented practices and guidelines. They have experience in managing projects for more than 30 years. In the last 8 to 10 years, the company started to manage their projects as programs and portfolios.
Soft is structured into a number of design units (DU) responsible for the financial success and the development of a portfolio of products. These design units are further structured into product design units (PDU). A number of design centers around the world are involved in the development of the components within the PDU. This includes centers in Europe, Asia, and North and South America. Portfolios are managed at the PDU level. The DU is composed of over 5000 employees who manage five portfolios that correspond to the five PDUs. Two portfolios, called Soft1 and Soft2, were studied.
The Portfolio Soft1 is composed of approximately 15 large projects/programs in one PDU and a total of approximately 50 subprojects/projects between 4 months and 18 months in duration. The planning horizon is between 18 months and 2 years. The portfolio was created approximately 5 years ago when the market for their product did not yet exist. It involves over 1000 people located in 7 sites on 3 continents.
The portfolio Soft2 is managed in another PDU and is composed of projects developing components and platforms re-used by the other PDUs and DUs, including the PDU managing Soft1. The main difference with the previous portfolio is that the products are not delivered directly to external customers but internally to other units. They have conflicting requirements coming from the different units which must be reconciled. Their products also include the integration of a large number of third-party products.
Company Fin
The second company, referred to as Fin, is a large Canadian financial services company. The projects are managed by project managers in the different business units and the relevant departments (e.g., IT). They have experience in managing projects for more than 15 years.
The portfolio Fin1 includes four programs covering a total of approximately 150 projects over a period of five years. The portfolio was put in place to comply with the Basel II Accord, which is an international agreement specifying the capital requirements for the underlying risks that financial institutions face (Bank for International Settlements, 2009). The portfolio was established in 2004 and is planned to continue for at least another year. The same portfolio manager is in place since 2005. The resources are mainly in Canada and are composed of a large proportion of consultants (approximately 50%). The history of this portfolio is well documented and includes a large number of changes.
A second portfolio in this enterprise; called Fin2, develops new procedures and tools to support upgrades to their accounting system. The portfolio was established in 2007 and is planned to be completed in 2011. It is composed of approximately 25 projects running in parallel. It includes short projects (less than 3 months) and longer projects (more than 1 year) with a high level of dependencies between projects.
Preliminary Findings
The elements covered in this paper are (a) rates and types of changes, and (b) sensing mechanisms. The seizing and transforming mechanisms will be presented in subsequent publications.
Rates and Types of Changes
Changes in the Project Portfolio in Company Soft
The environment of Soft is extremely dynamic and competitive for both portfolios (Soft1 and Soft2). Figure 2 displays the eight categories of changes described below according to their rates of change and their impact. The following paragraphs present the main drivers of change in approximately decreasing order of significance. In the case of Soft, the main source of uncertainty is related to the scope changes, which are all constantly evolving in a turbulent market. This is followed in importance by the project performance.
Figure 2: Rate of change and impact at Company Soft
New Product (Scope). The product was very new and there were a very large number of features that had to be put in place at the beginning to convince the customers that the product was viable. In the early phases of the product life cycle, it was not clear as to which features were of the highest priority to reach the market. Although Soft attempted to define and agree on the scope in the early phases of the projects using pre-studies, feasibility studies, and gate decisions, the scope had to be revisited continuously due to changing priorities.
The level of dependencies between the different projects of the portfolio was very high. Because the different projects are interrelated and must be developed towards a common release, most changes of scope affected more than one component of the software system. In Soft1 and Soft2, there were over 50 changes per year related to scope changes impacting more than one project or subproject. This was in addition to changes affecting individual nodes (which were in the hundreds). The high level of uncertainty related to the content was mentioned in almost all interviews with Soft.
Project performance. Due the high level of dependencies between projects, any major problem with project performance, such as delays or budget overruns in one of the projects, had ripple effects on other projects. The changes related to project performance were not so much to realign with the portfolio strategy, but to synchronize between projects and to ensure that the most critical functionality would still be delivered on time, that resource allocation was still balanced with availability, and that the portfolio would still be within budget. The causes of these problems with project performance were not specifically investigated. Because the projects share the same resource pools when a project had to keep resources longer than planned, the subsequent projects had to replan their resources and in most cases delay their project. Many interviewees mentioned that they would rather maintain the date by reducing the scope of the projects in order to avoid such cascading effects. There were around 15 to 20 changes per year to each portfolio due to project performance.
Changes in processes, Soft has project management and software development methodologies, which have been in place and improved over many decades. However, the requirement for increased flexibility due to the uncertain environment resulted in multiple attempts to modify the development process. This included the introduction of different variants of Agile processes, software release strategies, new integration, and testing approaches, in addition to the implementation of a new financial system across the enterprise. Both Soft1 and Soft2 introduced a resource planning process and tool by which line managers and project managers had to submit their resource demands and allocations on a monthly basis. All these changes had impacts on the project portfolio structure, such as attempts to reduce the duration and length of projects, attempts to reduce the dependencies between projects, separation of the pre-study of each project into a continuous activity outside the projects (see pre-study machine in a subsequent section on sensing). It is hard to estimate precisely the number of process changes during the year, but considering the number of comments regarding this issue, it could be estimated to approximately 5 to 10 per year.
Need for customization. Soft attempted to develop flexibility in their products and the intent was to develop a standard product which could be customized using configuration parameters. This was not always possible and despite these attempts, it was found that most customers needed special customization for which they had to pay. A number of special high-priority customization projects were therefore put in place. Although these projects were fairly small in comparison with the ongoing development projects, they still competed for the same resource pool. This type of request occurred approximately once per month.
New customers and new market (scope change). Soft tries to enter new markets with their innovative products, in some cases targeting a completely different set of customers. They made the decision to invest in the products with the expectation that the customers would replace some of their existing software and hardware in order to reduce their total cost of ownership and develop new products and services of their own. Soft did not have any contracts with customers when they started the projects in the Soft1 portfolio in early 2005. At the end of 2006, they signed two contracts with major customers who decided to evaluate the products in their laboratory and eventually with a limited subset of their end customers. The advent of these contracts had a very significant impact on the content of the project portfolio, which became much more tailored to the needs of these two customers, at the expense of the development of a potentially more “standard” product. This created an enormous inflow of new requirements, most of which were considered “must-have functionality” to be developed very urgently. As a consequence, the capacity of the R&D organization was exceeded and had to be rebalanced by cutting out or postponing some content.
After these two initial contracts Soft won around 40 smaller contracts and tried to develop a more “standard” product using a portfolio of products that would suit a larger number of customers. During the spring of 2009, a letter of intent was signed with a very important and influential customer. The ongoing portfolio had to be completely reshuffled to suit the needs of this very important customer. Although such significant events are fairly rare (approximately once per year), they are at the heart of the main source of uncertainty for those managing project portfolios in Soft1.
Changes in agreements with third-party suppliers (scope change). Both products of Soft1 and Soft2 integrate a number of third-party products. Business decisions have to be made on a regular basis on whether to “make or buy.” There were a number of occurrences where the termination of an agreement with a supplier resulted in the creation of a replacement project and, inversely, a replacement project was terminated because it was decided to use a third-party product. These changes are somewhat similar to the changes in scope discussed above but are less frequent, approximately once or twice per year.
Structural reorganizations. Soft has a history of reorganizing their line organization every 12 to 18 months. Soft1 and Soft2 had to go through approximately 3–4 reorganizations during a 5-year period. This included the creation of the PDU concept, creation of new PDUs, merging of PDUs, transfer of PDUs, transfer of responsibilities between sites, and closure of some units. Even if these structural changes had major impact on the personnel, the impacts on the project portfolios were minimal but some replanning had to be done by the project managers and the project portfolio managers.
Technology. Surprisingly, changes in technology were not considered frequent nor as having significant impacts. The interviewees considered that technology could be planned at least 6 months to a year in advance and included in the project plans accordingly. There were only very rare cases of changes to the portfolio due to changes in technology. This was considered the most stable or at least the most predictable area.
Change in business strategy. Change in the business strategy is one of the two examples of changes described in The Standard for Portfolio Management (Project Management Institute, 2008a). However such changes were not observed during the 2-year period analyzed for the two portfolios. The types of changes described in this section include a number of changes related to how to achieve and implement the strategy without changing the strategy itself.
Changes to the Project Portfolio in Company Fin
In comparison to Soft the environment of Fin is less dynamic. The goals of both Fin1 and Fin2 were to implement new internal tools and procedures to comply with external norms. Figure 3 displays the six categories of uncertainty according to their rates of change and their impacts. In the case of Fin, the most significant source of uncertainty was scope change due to changes in the norms themselves but also in their interpretation by the employees of Fin.
Figure 3: Rate of change and impact at Company Fin.
New interpretations of the norms (scope change). The most frequently mentioned source of change during the interviews was not so much the changes in the international norms themselves, but their interpretation by employees of Fin and occasionally by external consultants involved in the projects. The project methodology included the approval of the project scope by the stakeholders at given gate decision points. However, there were numerous instances where the receiver (customer) of the tools and procedures changed their mind on the exact requirement to be implemented once the project was in execution. In addition, there were at least six mentions of changes of interpretation due to new personnel assigned to the project. There was an average of 50 change requests per year (of which 85% were due to scope changes) during the 4 years for which data was available. The scope changes were primarily due to the interpretation but, a few times per year, the norm itself changed.
Changes in the norms (scope change). The norms to be implemented in Fin1 and Fin2 were not all available at the outset. The international bodies responsible for the norms released documents defining the norms as the programs unfolded. In many cases, they were not actually approved when a project started. Because of the planned duration of the projects and the external deadlines imposed to comply with these norms, the project sponsor had to take some risks and approve the start of some key projects regardless of this uncertainty. Changes in the norms occurred between 1 and 6 times per year. There was one example in Fin2 where a planned modification of the norm was cancelled by the international bodies. Even though money had already been spent on feasibility and design, the corresponding project had to be cancelled.
Project Performance. The project performance (time, cost, and scope) was monitored and reported on a regular basis to the portfolio manager and to the sponsor. Any deviation had to be documented using a change request at the portfolio level. There were two types of change requests: (1) requests for additional funds (if scope remained unchanged but more money was required) and (2) scope change request (covered previously). Approximately 15% of the change requests were related to additional funds (i.e., in most cases incorrect planning). This allowed the portfolio manager to take the required course of action: to reallocate money from other projects, to use contingency money, or to postpone other projects. The focus was similar to company Soft, where the changes related to project performance were not so much a realignment with the portfolio strategy, but to ensure that the resource allocation was still balanced and that the portfolio would still be within budget. However in the case of Fin, the main focus was on budget control at the portfolio level.
Portfolio budget reduction. During 2009, both portfolios in company Fin had to reduce their budget because of the crisis that hit the financial sector. Two main strategies were deployed: (1) 300 consultants were replaced gradually by (less costly) employees on the projects, and (2) the least urgent projects and functionality were delayed to 2010. Although this event only happened once, it had a significant impact when it occurred.
Technology. Technology changes were not considered frequent nor as having significant impacts. This was similar to the observation at company Soft. The interviewees considered that technology could be planned for in advance and included in the project plans accordingly. This was considered the most stable or at least the most predictable area.
Organizational change. The portfolio Fin1 has been in place since 2004 and portfolio Fin2 since 2007. During the lifespan of Fin1 there were 3 to 4 major restructurings of the company. One major reorganization, affecting both portfolios was ongoing in 2009, during the time of the interviews. The observations at Fin were the same as those in Soft. The changes to the portfolios due to the reorganizations were not major. They were, for example, new people assigned to projects (with new learning curves to build up competence and comprehension of the project), previous requirement specifications and norms for interpretation challenges, and slow down in the project due to uncertainty of the personnel involved. However, the content and the structure of the project portfolios remained largely unchanged.
Strategy. Changes in strategy were not observed during the 2- year period analyzed for the two portfolios in company Fin. It can be argued that some of the changes described previously might be considered as changes in how the strategy should be implemented but not so much changes in the strategy itself.
Sensing
Considering the specific type of uncertainty and the different levels of changes that Soft and Fin were experiencing, they implemented slightly different sensing mechanisms to monitor the potential changes and translate them into new requirements for the projects in the portfolio. Because Soft faced a more turbulent environment, they put in place more mechanisms than Fin to pro-actively sense changes in the external environment. This section describes the sensing mechanisms at Soft and Fin.
Dedicated role for specifying project content (Soft and Fin). Specifications for the content of projects were not done by the sponsor nor by the portfolio manager. In all four portfolios, there were people dedicated to specifying the content of projects and to ensuring that the needs for change were identified and integrated into the relevant projects when required. In the case of Soft, this responsibility was assigned to product managers who were responsible for keeping in contact with customers through the sales and marketing departments. They were responsible for all product content and ultimately the product benefits but they were not assigned a role within the projects. They were considered as the people that placed orders for projects. Respondents frequently pointed to product managers as being the prime source of changing priorities and project scope. In the case of Soft1, there were many product managers (around 50 people) within a product management department. The manager of that department was responsible for the complete scope of the project portfolio and had delegated the responsibility of the different components of the product to the respective product managers. The responsibility for product scope is critical for Soft but created some ambiguity regarding the exact responsibility for the project portfolio between the product managers, the project sponsor, the portfolio manager, and the resource owners.
In the case of Fin, business analysts were assigned to projects to interface with the customers, that is, the units which were receiving the new tools and processes and their employees. Most projects had fulltime business analysts assigned to their projects who were responsible for organizing workshops to gather, specify, and validate the project requirements. These activities were very intense in the early phases of the projects (identification and design) but continued throughout the complete project life cycle. Business analysts monitored sources of changes and ensured that they were identified and captured in due time. Fin1 introduced the notion of an integration business analyst who was responsible to oversee the requirements at the portfolio level to minimize duplications and conflicts.
Pre-Study Machine (Soft1). The software development model at Soft historically included four phases: pre-study, feasibility, design, and test. This was then followed by release activities and deployment at customer sites. In the case of Soft1, the amount of scope change was extremely high and the management felt that there was a lot of waste on studying and designing features which were removed later. Three interviewees mentioned that there was a 50% hit rate on the requirements (i.e., only 50% of the requirements identified in the early phase of the projects reached the market). Because the inflow of new requirements was continuous during the year and the project scope could not be planned for the entire year, the pre-study activities were separated from the individual projects. This became a continuous activity managed by a requirement request board (chaired by the system group in collaboration with product management) in parallel with the feasibility and execution of ongoing projects as shown in Figure 4. The investigations of individual requirements were fairly quick and led to a decision to incorporate them in a specific work package for a specific project. Once the requirement request board (RRB) was introduced there was no longer a need for change control boards at the individual project level. The planning horizon for the content of the projects was also shortened to approximately 1 to 2 months instead of the 6 to 9 months using the standard development process. The roles of the RRB were threefold: (a) to translate customer need into product requirements, (b) to evaluate impact (i.e., size of development), and (c) to decide which project should include this new requirement. The first function was clearly a sensing mechanism but the two others were seizing functions.
Figure 4: Pre-Study Machine Overview
Prototypes (Soft and Fin). Prototyping techniques were used in all four portfolios although for different reasons. In Soft1, this was used to develop new ideas and show potential new products in the very early stages to customers either in their laboratory or in trade shows. This technique identified and validated the product requirements at low cost. At Soft2, prototyping was more in the form of early deliveries (not fully tested products) of their platforms and components to internal customers so as to allow them to develop their applications and provide feedback. At Fin, prototyping was used primarily to validate the interpretation of the norms by stakeholders in the early stage of the projects to reduce subsequent rework and reach agreement on the project scope.
Innovation through employee contributions (Soft1). Employees were encouraged by means of internal competitions to submit ideas for new services. The first year they received 300 ideas from employees primarily from the head office location. It became a challenge to manage all of these ideas. The following year, the firm decided to provide some themes to limit the number of contributions. The result was actually the reverse from the one expected with over 500 ideas coming not only from the head office but also from all the sites in the PDU. The winning employees were invited to help implement those ideas in a demonstration laboratory in order to show some of the potential new products to customers.
Product customization (Soft1). Some customers have long product evaluation periods where different suppliers must comply with a list of use cases. The qualification process is often based on the percentage of cases which are demonstrated successfully to the customers. A program was set-up to support such activity to ensure that test cases would be executed successfully in the customer laboratories. The team on-site was supported by designers, which can patch the system to make it work. These requirements are then fed back into the normal requirement evaluation process.
Focus on time-to-decision (Soft). Soft measures the time from when an idea is identified by customers until the time it is included in a project. Although the time to design and develop the product once the content is defined was equivalent to the competition, they realized that the front-end period to determine the content was much longer than its competitors. This focus on time to decision allowed them to identify the critical bottlenecks in the decision process preceding the launch of projects or the inclusion of features in projects.
Discussion
The monitoring subprocess in the Standard for Portfolio Management (PMI, 2008a) includes two types of changes, changes induced by: (a) Portfolio Performance and (b) Business Strategy Changes. The preliminary findings in this research indicate that the drivers of change go beyond the two groups identified in the PMI standard. Even with the limited sample of four portfolios, a large number of types of changes were identified. Reviews of risks and portfolio performances were observed with a main focus on the performance of individual projects within a portfolio context rather than a reassessment of their raison d'être or business objectives. However, changes in strategy were not observed during the lifetime of the four portfolios.
In all four portfolios, change in scope was the most important source of uncertainty. Sensing mechanisms to identify and interpret potential changes in scope were put in place in all four cases, which included the creation of specific roles to define project content, the use of prototypes, and the introduction of a pre-study machine in the case of Soft.
Surprisingly, technology changes were not considered frequent nor as having significant impacts. There were only very rare cases of changes to the portfolio due to changes to technology. This was considered the most stable or at least the most predictable area.
Conclusion
This paper presents findings on the management of uncertainty in project portfolios in dynamic environments. It is based on a study of four portfolios in two firms performed during the summer and autumn of 2009. Only preliminary findings are presented in the paper.
The dynamic capability framework based on: sensing, seizing and transforming provides a useful framework to study PPM when faced with a high level of uncertainty in dynamic environments. This paper presented some of the sensing mechanisms observed in order to organize for uncertainty. The description of managing changes to project portfolios in dynamic environments is richer than that provided in the literature; many additional types and sources of change have been identified. In addition, the most frequently mentioned source of change from the literature was not observed. Although these results are preliminary, they indicate that this research has the potential to significantly enrich the current understanding of the management of project portfolios.
Even though the concept of dynamic capabilities has been prevalent in the strategic management literature for at least ten years, only a few such capabilities have been investigated empirically and, unfortunately, there are very few descriptions of how firms can implement and maintain dynamic capabilities in practice. The present research has the potential to contribute to the development of the theory of dynamic capabilities and to the practical aspects of their management.
Bios:
Yvan Petit, MEng, MBA, PMP, is a PhD candidate in management at the University of Quebec at Montreal (UQAM), Canada. His doctoral research is on how uncertainty is managed in project portfolios in dynamic environments. He has over 25 years of experience in project management primarily in software development for the telecommunications industry. This includes assignments in England, Sweden, and Singapore. Yvan is also adjunct professor at UQAM and trainer for the PMP exam preparation for the PMI—Montreal Chapter.
Brian Hobbs, PMP, Project Management Research Chair www.pmchair.uqam.ca, holds a degree in Industrial Engineering, an MBA, and a PhD in Management. He has been a professor at the University of Quebec at Montreal (UQAM) in the Master's Program in Project Management for more than twenty years. This program, of which he is a past director, is accredited by PMI's Global Accreditation Center. He has served terms on both PMI's Standards and Research Members Advisory Groups. He is a reviewer for both the Project Management Journal and the International Journal of Project Management. He has presented many papers at both research and professional conferences worldwide.