Professional Documents
Culture Documents
Project Portfolio Control and Portfolio
Project Portfolio Control and Portfolio
ABSTRACT ■ INTRODUCTION ■
T
he project management research community has recently begun to
This article investigates the nature and relation-
pay increasing attention to the context in which projects are man-
ship of project portfolio control techniques and
portfolio management performance, and how aged (Engwall, 2003). The contextual setup has been examined, for
this relationship is moderated by situational example, in terms of:
idiosyncrasies of internal and external dynam- • how projects are linked to broader business interests and the strategy of the
ics, industries, governance types, and geo- firm (Artto & Dietrich, 2004; Morris & Jamieson, 2005);
graphic location. A worldwide questionnaire
• how projects relate to programs and portfolios (Turner & Müller, 2003);
with 242 responses was used, of which 136
high-performing responses were filtered out for • how projects are managed as part of strategic programs (Lycett, Rassau, &
quantitative analysis of best practices. Three Danson, 2004; Vereecke, Pandelaere, Deschoolmeester, & Stevens, 2003);
portfolio control factors were identified: portfo- • how projects are perceived in different industries (Blomquist & Wilson,
lio selection, portfolio reporting, and decision- 2007); and
making style. Two measures for portfolio man-
• how projects are managed as part of the portfolio of an organization’s proj-
agement performance were identified: achieve-
ment of desired portfolio results and achievement ects (Elonen & Artto, 2003; Engwall & Jerbrant, 2003; Payne & Turner, 1999;
of project and program purpose. The results Söderlund, 2004).
indicate that different portfolio control mecha-
nisms are associated with different perfor- All these tracks of literature acknowledge that single projects cannot
mance measures. A contingency model was
be treated as isolated entities, but they influence and are influenced by
developed, including moderating effects by
contextual variables. the complex and uncertain character of their context. This context is set by the
program or project portfolio of which a project is a part. Professional organ-
KEYWORDS: project portfolio manage- izations, such as the Project Management Institute (PMI; 2006) or the
ment; portfolio management performance; Association of Project Management (APM; 2004), have both addressed this
portfolio control through development of their own standards and guidelines for managing
programs and portfolios. While much has been written on the management
of programs, the management of portfolios is less often addressed in the lit-
erature. This is the focus of the present article.
A project portfolio is a group of projects that share and compete for the
same resources and are carried out under the sponsorship or management
of an organization (Archer & Ghasemzadeh, 1999a, 1999b). Turner and
Müller (2003, p. 7) defined a portfolio as “an organization (temporary or per-
manent) where projects are managed together to coordinate interfaces, pri-
oritize resources between projects, and thereby reduce uncertainty.” Project
portfolio management can therefore be considered a dynamic decision
process where a list of active projects is constantly updated and revised
Project Management Journal, Vol. 39, No. 3, 28–42 (Cooper, Edgett, & Kleinschmidt, 1997a). Related normative frameworks and
© 2008 by the Project Management Institute techniques for project evaluation, selection and prioritization, resource allo-
Published online in Wiley InterScience cation, and knowledge sharing between projects were often the focus of ear-
(www.interscience.wiley.com) lier project portfolio management research (Martinsuo & Lehtonen, 2007).
DOI: 10.1002/pmj.20053 The progress in understanding portfolio management and its applications
(Cooper et al., 1997a). They found some launched to achieve an organization’s performance. In summary, we see, how-
supporting evidence concerning the desired benefits or change, should ever, that the literature suggests a con-
link between portfolio-level results and accordingly include the project tingency view: fitting the portfolio
business-level performance, and prod- dimension, the portfolio dimension, selection approach to the surrounding
uct-level performance indicators. and the organization within which the organization’s characteristics and strat-
Nobeoka and Cusumano (1995, portfolio is executed. Our first egy (Englund & Graham, 1996; Stawicki
1997) focused on business-level perfor- hypothesis is therefore: & Müller, 2007).
mance indicators such as sales growth, A few empirical, qualitative studies
H1: Portfolio performance measures
and project performance. Martinsuo give partial support to the potential
in practice are multidimensional and
and Lehtonen (2007) identified percep- linkage between portfolio selection and
include the project, portfolio, and orga-
tions of portfolio management efficiency portfolio management performance.
nizational level.
as subjective culmination of portfolio- Fricke and Shenhar (2000) reported pri-
strategy alignment, knowledge of oritization as a success factor in multi-
Portfolio Control
priorities, financial yield, realization project environments. Elonen and
None of these studies previously men-
of strategy, and efficiency of managing Artto (2003) reported resource alloca-
tioned covers portfolio control as a fac-
the project entity. tion issues and lack of portfolio-level
tor influencing portfolio management
Some studies have shown that in activities, including project overlaps
performance. However, they all include
multiproject settings, different types of and lack of prioritization, as problems
supporting evidence that points toward
performance should be considered with managing multiproject environ-
examining this phenomenon further.
jointly, instead of separating them into ments. Engwall and Jerbrant (2003,
For example, Fricke and Shenhar’s
those for projects, programs, and port- p. 408) stated that
(2000) concepts of clear goals, manage-
folios. A study of middle managers in
ment support, ownership, and prioriti- research on multi-project manage-
program and project portfolio manage- ment has to go beyond resource
zation; Nobeoka and Cusumano’s
ment acknowledged that portfolio per- allocation and start addressing
(1995) technology transfer; as well as all
formance, program performance, and incentive structures, accounting
of Martinsuo and Lehtonen’s (2007)
project performance should be studied systems and other deeply embed-
single-project variables could be con-
simultaneously (Blomquist & Müller, ded features of the organization. [. . . ]
sidered as means of control contribut-
2006a). This is supported by Martinsuo the whole system of managerial
ing to portfolio management. The procedures has to be reconceptual-
and Lehtonen (2007), who showed that
question is, what are an organization’s ized from the roots.
reaching of project goals mediates the
practices in executing portfolio-level
relationship between single-project
control and how are they used. Earlier Based on both the normative sug-
management success and portfolio
literature in this area can be catego- gestions on portfolio selection and the
management efficiency.
rized in three main areas: earlier qualitative studies, we hypothe-
However, portfolios are also not
• portfolio selection; size on the role of portfolio selection in
stand-alone entities. They are part of a
• portfolio reporting; and relation to performance.
wider organizational context. PMI
• portfolio decision making.
(2006) defined success of portfolio H2: Selection of projects for the port-
management as the portfolio’s contri- Portfolio Selection as a Form of folio based on the organization’s strate-
bution to the strategic measures of the Behavioral Control gy is positively associated with portfolio
organization. Portfolio decisions A variety of tools and techniques for the management performance.
(such as which projects to accept in optimal selection of projects and port-
Project Reporting for Portfolio Follow-
the portfolio, at what priority, and folios have been suggested (e.g., Archer
Up as Output Control
with which resources) must balance a & Ghasemzadeh, 1999b; Hall & Nauda,
multitude of conflicting goals of an 1990; Henriksen & Traynor, 1999; The APM (2004) guidelines on gover-
organization. These portfolio deci- Stawicki & Müller, 2007). Such tech- nance of project management defines
sions cannot be taken in a vacuum niques range from criterion lists, strate- reporting as one of the four dimensions
and must be balanced against other gy tables, scoring tables, visual graphs, for proper governance practice, thus as
organizational performance measures and force-field analyses to optimiza- a control structure. This is supported by
or pressures (Turner & Müller, 2003). tion models, holistic portfolio manage- research that showed the link between
Projects, portfolios, and their hosting ment frameworks, and the interaction information sharing and performance
organizations are interwoven. Control with a portfolio board. No estimation variables in multiproject settings
and performance measures in multi- exists about the contribution of individ- (Fricke & Shenhar, 2000; Nobeoka &
project settings (such as portfolios), ual techniques to portfolio management Cusumano, 1997).
The variable structure supported program purpose). Balancing priorities Portfolio Control
Hypothesis 1: portfolio performance is about fulfilling broader business A Varimax rotated factor analysis was
measures in practice are multidimen- interests in a balanced manner and done on the portfolio control practices
sional and include the project, portfo- includes three items (resource retention covered in the questionnaire. Three
lio, and organizational level. [inverse of resource turnover]; timely factors on portfolio control were extract-
Through the factor analysis we accomplishments of programs; and ed, explaining 67% of the sample vari-
identified three portfolio management stakeholder satisfaction). ance (KMO ⫽ 0.772, Bartlett’s p ⫽ 0.000).
performance measures: achieving For each factor (achieving results, Table 2 shows the factor structure and
results, achieving purpose, and balanc- achieving purpose, and balancing pri- loadings. The item “decisions are made in
ing priorities. Achieving results resem- orities), a summated scales variable the best interest” was common to portfo-
bles the most common measures of was created as the average of those lio selection and portfolio decision mak-
success in projects and includes four items loading higher than 0.5 on a fac- ing and it was included in both variables.
items (customer satisfaction; combined tor. Reliability tests led to the exclusion The results supported the three
time, cost, and quality results; financial of the balancing priorities summated mechanisms of portfolio control sug-
results; and user requirements). Achiev- scale as the only variable of unaccept- gested in the literature: portfolio selec-
ing purpose deals with achieving both ably low Cronbach’s alpha value (0.4). tion, portfolio reporting, and portfolio
project- and program-level purpose The other two variables were used as decision making.
and includes two items (achievement replacements of the original variables Portfolio selection is about aligning
of project purpose and achievement of in the further analyses. projects with strategy and prioritizing
management performance variables as The analyses supported the overall Here the relationship between portfolio
the dependent variables. The results are model shown in Table 5, by reinforcing selection and achieving purpose is
summarized in Table 6 for those correla- the relationships between portfolio strengthened in contexts of high external
tions that were found significant. The reporting and achieving results, as well dynamics, low internal dynamics, and
table shows the structural context vari- as portfolio selection and achieving pur- internal projects. The relationship
ables in the rows, and the base model, as pose, in different contexts. However, it between portfolio selection and achiev-
shown in Table 5, in the two left also revealed a latent relationship ing results appears to be strengthened in
columns. The right column shows those between portfolio selection and achiev- contexts of low internal dynamics, North
correlations identified solely during the ing results in some specific contexts. To American location, hybrid governance
regression analyses by different context that end, the relationship between structures, and internal and external
variables. The body of the table shows portfolio reporting and achieving results projects. Internal projects and circum-
the significant regression correlations, appears to be stable across all contexts. stances of low internal dynamics appear
their adjusted R-square and the signifi- The relationship between portfolio to be the context factors reinforcing the
cance (indicated by *). Results in paren- selection and the two portfolio manage- impact of portfolio selection on both per-
theses are based on small sample sizes ment performance measures, however, formance measures achieving purpose
of fewer than 30 responses. appears to be moderated by context. and achieving results.
The results support Hypothesis 5: decision making, using qualitative and trol, not previously covered empirically
the relationship between portfolio quantitative factors, being superior to at the multiproject level.
control and portfolio performance is individual decision making (Stawicki & The results further showed that
moderated by contextual factors. Müller, 2007). organizations with different gover-
nance styles differ in their use of differ-
Discussion and Conclusions Structure of Project Portfolio Control
ent portfolio control practices, whereas
We investigated the underlying struc- and Portfolio Management
other contextual factors did not appear
tures, relationships, and contextual con- Performance
as significant. Four types of governance
tingencies of portfolio control practices This study has examined project port-
structures were included in this study:
and portfolio management perform- folio control as a potential factor con-
multiproject organizations that did not
ance in successful organizations. As our tributing to portfolio management
have clear integration among projects,
primary research question, we asked: performance. The empirical results of
portfolio organizations, program
How are portfolios controlled, and how 133 organizations showed, first, that
organizations, and hybrid organiza-
does this control relate to portfolio successful organizations have an
tions that combined portfolio and
management performance in different organization-level practice of selecting
program governance. As hybrid organi-
contexts? and prioritizing projects in line with
zations represent an example of well-
Our study has developed the con- strategy. Second, successful organiza-
developed and mature project-based
cept of portfolio control and analyzed tions have a shared reporting approach
organizations, they tend to have higher
its association with portfolio manage- to channel information flows from
degrees of portfolio control in place, in
ment performance in different con- projects to the portfolio level. Third,
the form of portfolio selection, portfo-
texts. The final model for the relation- such organizations share responsibility
lio reporting, and decision-making
ship between portfolio control and for decisions at the portfolio level. The
practices. Also, firms in portfolio man-
portfolio management performance factor structure identified in portfolio
agement–driven governance structures
and the moderating effects of context control practices clearly adds to find-
were more advanced in decision-mak-
are shown in Figure 3. ings of earlier studies in general multi-
ing practices than less mature multi-
These results are supported by project management (Cooper et al.,
project organizations. These findings
other studies, such as those on the link 1997a, 1997b, 1999, 2000) and different
advance the current understanding
between strategy and organizational practices in multiproject contexts
that portfolio control distinguishes
success (Cooper, Edgett, & Kleinschmidt, (Fricke & Shenhar, 2000; Loch, 2000;
successful from less successful organi-
2004a, 2004b, 2004c; Morris & Jamieson, Martinsuo & Lehtonen, 2007) by sug-
zations. It shows that successful organ-
2005), those on the link between report- gesting contents for three portfolio
izations differ in their use of portfolio
ing and project success (Turner & control variables. Our study has given
control mechanisms, depending on the
Müller, 2004), as well as those on team shape to the concept of portfolio con-
type of organizational governance. Our
results suggest that at least governance
mechanisms in the portfolio’s organiza-
tional context should be taken into
Portfolio Portfolio account when managing the portfolio.
Control: Success: Prior research has suggested that
portfolio management performance
Portfolio reporting Achieving results
should not be one-dimensionally
examined at portfolio level only, but it
Portfolio selection Achieving purpose should also include achievements at
the single-project and organizational
level (Blomquist & Müller, 2006a;
Situation: Martinsuo & Lehtonen, 2007). Our find-
Governance type ings lend support to such studies and
Industry Hypothesis 1 on the multidimensionality
Geography of portfolio management performance.
Dynamics The results showed that successful
Project type organizations look into achieving desired
portfolio results, and achieving proj-
ect and program purpose for the
Figure 3: Final model.
overall portfolio. The factor structure
personal preferences of individual Blomquist, T., & Müller, R. (2006b). practices-III. Research Technology
managers. Middle managers in program and port- Management, 47(6), 43–55.
• Review portfolios periodically using folio management: Practice, roles and Elonen, S., & Artto, K. (2003).
comparable metrics. responsibilities. Newtown Square, PA: Problems in managing internal devel-
Project Management Institute. opment projects in multi-project envi-
The study provided insight into dif- Blomquist, T., & Wilson, T. L. (2007). ronments. International Journal of
ferent portfolio management practices. Project marketing in multi-project Project Management, 21, 395–402.
Organizations should use both the con- organizations: A comparison of IS/IT Englund, R., & Graham, R. (1996).
trol mechanisms and the success fac- and engineering firms. Industrial Creating an environment for success-
tors and assign operational values to Marketing Management, 36(2), ful projects—Steps and examples.
them. That allows tracking, setting, and 206–218. Presented at Project World Winter ‘96,
measurement of goals, and their Bresnen, M., Edelman, L., Newell, S., Hewlett-Packard Company, Santa
achievement, which is the aim of port- Scarbrough, H., & Swan, J. (2003). Clara, California.
folio management. The study has pro- Social practices and the management Engwall, M. (2003). No project is an
vided the metrics for it. ■ of knowledge in project environments. island: Linking projects to history
International Journal of Project and context. Research Policy, 32,
Management, 21, 157–166. 789–808.
References
Archer, N., & Ghasemzadeh, F. (1999a). Cooper, R., Edgett, S., & Kleinschmidt, Engwall, M., & Jerbrant, A. (2003). The
An integrated framework for project E. (1997a). Portfolio management in resource allocation syndrome: The
portfolio selection. International new product development: Lessons prime challenge of multi-project man-
Journal of Project Management, 17, from the leaders I. Research Technology agement? International Journal of
207–216. Management, 40(5), 16–28. Project Management, 21, 403–409.
Archer, N., & Ghasemzadeh, F. (1999b). Cooper, R., Edgett, S., & Kleinschmidt, Fricke, S. E., & Shenhar, A. J. (2000).
Project portfolio selection techniques: E. (1997b). Portfolio management in Managing multiple engineering proj-
A review and a suggested integrated new product development: Lessons ects in a manufacturing support envi-
approach. In L. D. Dye & J. S. from the leaders II. Research ronment. IEEE Transactions on
Pennypacker (Eds.), Project portfolio Technology Management, 40(6), 43–52. Engineering Management, 47, 258–268.
management: Selecting and prioritiz- Cooper, R., Edgett, S., & Kleinschmidt, Hall, D., & Nauda, A. (1990). An inter-
ing projects for competitive advantage E. (1999). New product portfolio man- active approach for selecting IR&D
(pp. 207–238). West Chester, PA: Center agement: Practices and performance. projects. IEEE Transactions on
for Business Practices. Journal of Product Innovation Engineering Management, 37, 126–133.
Artto, K. A., & Dietrich, P. H. (2004). Management, 16, 333–351. Henriksen, A. D., & Traynor, A. J.
Strategic business management Cooper, R., Edgett, S., & Kleinschmidt, (1999). A practical R&D project-selec-
through multiple projects. In P. W. G. E. (2000). New problems, new solu- tion scoring tool. IEEE Transactions on
Morris & J. K. Pinto (Eds.), The Wiley tions: Making portfolio management Engineering Management, 46, 158–170.
guide to managing projects (pp. 144–176). more effective. Research Technology Kasvi, J. J. J., Vartiainen, M., &
London: John Wiley & Sons. Management, 43(2), 18–33. Hailikari, M. (2003). Managing knowl-
Association of Project Management Cooper, R., Edgett, S., & Kleinschmidt, edge and knowledge competences in
(APM). (2004). Directing change: A E. (2002). Optimizing the stage gate projects and project organisations.
guide to governance of project manage- process: What best practice companies International Journal of Project
ment. High Wycombe, UK: Association do-II. Research Technology Management, 21, 571–582.
for Project Management. Management, 45 (6), 43–49. Loch, C. (2000). Tailoring product
Blomquist, T., & Müller, R. (2004). Cooper, R., Edgett, S., & Kleinschmidt, development to strategy: Case of a
Needs and prerequisites for project E. (2004a). Benchmarking best NPD European technology manufacturer.
portfolio management. Presented at practices-I. Research Technology European Management Journal, 18,
the PMI Central Sweden Chapter Management, 47(1), 31–43. 246–258.
meeting, Stockholm, Sweden. Cooper, R., Edgett, S., & Kleinschmidt, Lycett, A., Rassau, A., & Danson, J.
Blomquist, T., & Müller, R. (2006a). E. (2004b). Benchmarking best NPD (2004). Programme management: A
Practices, roles and responsibilities of practices-II. Research Technology critical review. International Journal of
middle managers in program and Management, 47(3), 50–59. Project Management, 22, 289–299.
portfolio management. Project Cooper, R., Edgett, S., & Kleinschmidt, Martinsuo, M., & Lehtonen, P. (2007).
Management Journal, 37(1), 52–66. E. (2004c). Benchmarking best NPD Role of single-project management in
the areas of project portfolio management and Sweden. He is head of the management section support project management research activities
organization development. She studied at and a director for the Erasmus Mundus Masters within IRNOP and Project Sweden. Previous par-
Helsinki University of Technology and holds Program in Strategic Project Management, a ticipation experience includes product develop-
D.Sc. (Eng.), Lic.Tech., and M.Sc. (Eng.) degrees joint program with Heriot-Watt University and ment and renewal projects in both industry and
in industrial engineering and management. Politecnico di Milano. In recent years he has the public sector. This also includes activities to
worked in research concerning projectified firms access and improve project management and
and how they organize, manage, and control portfolio management systems in firms. He
their projects, as well as studies in product holds an MSc in engineering and a BA in busi-
Tomas Blomquist is associate professor at development and customer projects. He had ness administration.
Umeå School of Business at Umeå University in been involved in work to coordinate and actively