Decision Sciences
Volume 42 Number 1
February 2011
C 2011 The Authors
C 2011 Decision Sciences Institute
Decision Sciences Journal Understanding the Elusive Black Box
of Dynamic Capabilities
Paul A. Pavlou†
Fox School of Business, Temple University, Philadelphia, PA 19122,
e-mail: pavlou@temple.edu
Omar A. El Sawy
Marshall School of Business, University of Southern California, Los Angeles, CA 90089,
e-mail: elsawy@usc.edu
ABSTRACT
A major challenge for managers in turbulent environments is to make sound decisions
quickly. Dynamic capabilities have been proposed as a means for addressing turbulent
environments by helping managers extend, modify, and reconfigure existing operational
capabilities into new ones that better match the environment. However, because dynamic capabilities have been viewed as an elusive black box, it is difficult for managers
to make sound decisions in turbulent environments if they cannot effectively measure dynamic capabilities. Therefore, we first seek to propose a measurable model of
dynamic capabilities by conceptualizing, operationalizing, and measuring dynamic capabilities. Specifically, drawing upon the dynamic capabilities literature, we identify a
set of capabilities—sensing the environment, learning, coordinating, and integrating—
that help reconfigure existing operational capabilities into new ones that better match
the environment. Second, we propose a structural model where dynamic capabilities
influence performance by reconfiguring existing operational capabilities in the context
of new product development (NPD). Data from 180 NPD units support both the measurable model of dynamic capabilities and also the structural model by which dynamic
capabilities influence performance in NPD by reconfiguring operational capabilities,
particularly in higher levels of environmental turbulence. The study’s implications for
managerial decision making in turbulent environments by capturing the elusive black
box of dynamic capabilities are discussed.
Subject Areas: Decision Making in Turbulent Environments, Dynamic
Capabilities, Environmental Turbulence, New Product Development, and
Operational Capabilities.
INTRODUCTION
Decision making in turbulent environments is challenging because managers must
decide and act rapidly (Carlsson & El Sawy, 2008). In turbulent environments,
dynamic capabilities—“the ability to integrate, build, and reconfigure internal
and external competencies to address rapidly-changing environments”—(Teece,
† Corresponding
author.
239
240
The Elusive Black Box of Dynamic Capabilities
Pisano, & Shuen, 1997, p. 517) have been viewed as viable means for managing in
turbulent environments. However, dynamic capabilities have been described mostly
as abstract concepts or an elusive “black box.” The literature viewed dynamic capabilities as hidden or invisible (Itami, 1987), complex and tacit (Dierickx & Cool,
1989), difficult to observe (Simonin, 1999), and causally ambiguous (Williamson,
1999). Zahra, Sapienza, and Davidsson (2006) note several inconsistencies in the
usage of the dynamic capabilities concept. Dynamic capabilities have also been
criticized for their lack of precise definition, empirical grounding, and measurement (Williamson, 1999), and attempts to measure dynamic capabilities have used
distant proxies (e.g., Henderson & Cockburn, 1994). Arend and Bromiley (2009)
note a plethora of proxies, often unusual, for dynamic capabilities. Nerkar and
Roberts (2004) argue: “absent the ability to measure these (often intangible) assets
[dynamic capabilities] with any degree of precision, we assume that they develop
as a function of a firm’s accumulated experience” (p. 781). Zahra et al. (2006) suggest that dynamic capabilities are mostly empirically identified ex post. Galunic
and Eisenhardt (2001) further argue that the existence of dynamic capabilities is
often assumed without specifying their exact components. Given the difficulty in
articulating the characteristics and measurement of dynamic capabilities, Winter
(2003, p. 991) cites critics who even argued that dynamic capabilities are “born,
not made” and challenged whether they even exist in practice. Others have argued
that dynamic capabilities are tautologically linked to performance (Priem & Butler,
2001; Williamson, 1999). Finally, there is a notion that dynamic capabilities may
not even be amenable to managerial action (Grant, 1996), implying that managers
may not even recognize the existence of dynamic capabilities in practice. Concluding their review of the limitations of the dynamic capabilities view, Arend and
Bromiley (2009) argue that there is an urgent need for a coherent theory and model
for dynamic capabilities.
The poor understanding of dynamic capabilities and the lack of a measurable
model makes it difficult to study how dynamic capabilities can be used in actionable managerial decision making. To help managers make decisions in turbulent
environments with the aid of dynamic capabilities, our first objective it to open
the black box of dynamic capabilities to propose a measurable model of dynamic
capabilities by conceptualizing, operationalizing, and measuring dynamic capabilities. Such a measurable model would enable managers to rely upon dynamic
capabilities to reconfigure operational capabilities to enhance the quality of their
decision making in turbulent environments.
Second, to examine the performance effects of dynamic capabilities in turbulent environments, it is necessary to understand their links with operational
capabilities. By showing the positive effect of dynamic capabilities on reconfiguring operational capabilities to match turbulent environments, this study aims
to overcome the criticism that dynamic capabilities have a positive bias and are
tautologically linked to performance (e.g., Williamson, 1999; Priem & Butler,
2001; Zott, 2003). Moreover, to specify the boundaries of dynamic capabilities,
the study examines the effects of dynamic capabilities on operational capabilities
in different levels of environmental turbulence. Our second objective is thus to
propose a structural model by which dynamic capabilities shape performance by
reconfiguring existing operational capabilities in turbulent environments.
Pavlou and El Sawy
241
The proposed model is tested in the context of new product development
(NPD), which has been a prominent context for the study of dynamic capabilities.
Ettlie and Pavlou (2006) studied buyer–supplier NPD relationships to show that
interorganizational dynamic capabilities result in positive performance outcomes
in NPD. Pavlou and El Sawy (2006) showed dynamic capabilities to be developed
as a function of the firm’s leveraging competence of information technology (IT).
Agarwal and Selen (2009) showed dynamic capabilities to be built through collaboration among various stakeholders in service firms result in service innovations
and improved service offerings. Extending these studies of dynamic capabilities
in NPD, our work proposes a measurable model of dynamic capabilities and their
effect on performance by shaping operational capabilities in NPD.
CONCEPTUAL DEVELOPMENT
Conceptualization, Operationalization, and Measurement
of Dynamic Capabilities
Origins and scope of dynamic capabilities
The dynamic capabilities view originates in spirit from Schumpeter’s (1934)
innovation-based competition where competitive advantage is based on the creative destruction of existing resources and novel recombination into new operational capabilities. These ideas were further developed in the literature, such as
architectural innovation (Abernathy & Clark, 1986), configuration competence
(Henderson & Cockburn, 1994), and combinative capabilities (Kogut & Zander, 1992). Extending these studies, Teece et al. (1997) developed the notion of
dynamic capabilities, and their seminal paper is considered the most influential
source on dynamic capabilities, together with a recent framework of dynamic capabilities (Teece, 2007). Teece and colleagues (Teece et al., 1997; Teece, 2007)
see competitive advantage in turbulent environments as a function of dynamic
capabilities rather than competitive positioning or industry conflict. They used the
term “dynamic” to reflect “the capacity to renew competences so as to achieve
congruence with the changing environment” (p. 515). The dynamic capabilities
view follows the resource-based view (RBV) (Makadok, 2001), whereas RBV
emphasizes resource picking (selecting resource combinations), dynamic capabilities stress resource renewal (reconfiguring resources into new combinations
of operational capabilities). Because managers must regularly make decisions on
how to renew existing operational capabilities into new ones that better match the
changing environment, dynamic capabilities represent an important challenge for
managers in their quest to achieve a sustainable competitive advantage (Grewal &
Slotegraaf, 2007). Accordingly, to better understand dynamic capabilities, it is important to recognize their distinction from operational capabilities that they seek to
reconfigure.
Dynamic capabilities versus operational capabilities
It is imperative to distinguish between dynamic capabilities and operational capabilities. Capabilities are collections of routines (Winter, 2003), and they describe how effectively routines are executed relative to the competition (Nelson &
242
The Elusive Black Box of Dynamic Capabilities
Winter, 1982). According to Winter, a capability is described as a high-level routine
(or a collection of routines). A routine is a “behavior that is learned, highly patterned, repetitious, or quasi-repetitious, founded in part in tacit knowledge—and
the specificity of objectives” (p. 991). Although both capabilities are collections
of routines, dynamic capabilities describe the ability to reconfigure and change,
whereas operational capabilities denote the ability to “make a daily living” (Winter, 2003, p. 991). Operational capabilities were termed by Winter as ordinary or
“zero-order” capabilities whose goal is to “earn a living by producing and selling
the same product, on the same scale and the same customer population” (p. 992).
Operational capabilities are herein defined as the ability to execute day-to-day
activities. On the other hand, dynamic capabilities are termed “first-order” capabilities whose goal is to change the product, the production process, the scale,
or the markets (customers) served (Winter, 2003). Collis (1994) notes that dynamic capabilities govern the change of operational capabilities by reconfiguring
them to keep them relevant to the changing environment. Using March’s (1991)
exploration-exploitation view, dynamic capabilities would correspond to the exploration of new opportunities, whereas operational capabilities would correspond to
the efficient exploitation of existing resources. Following Winter (2003) and Teece
(2007), we define dynamic capabilities as those capabilities that help units extend,
modify, and reconfigure their existing operational capabilities into new ones that
better match the changing environment. Applied to NPD, whereas dynamic capabilities would focus on selecting the product to match the changing environment,
operational capabilities would focus on executing the day-to-day activities needed
to actually develop the product.
Identification of proposed set of dynamic capabilities
Extending the work of Teece and his colleagues (Teece et al., 1997; Teece, 2007)
from an abstract view of firm-level aggregate strategy to a more tangible view of
dynamic capabilities at the managerial decision-making level, we seek to identify,
conceptualize, operationalize, and measure a measurable model with a set of
identifiable and specific components of dynamic capabilities.
To identify the proposed set of dynamic capabilities, we relied on Eisenhardt
and Martin (2000) who note: “dynamic capabilities actually consist of identifiable
and specific routines that often have been the subject of extensive empirical research in their own right” (p. 1107). Extending this logic and drawing upon the
strategic management and decision sciences literatures, we aim at proposing an
identifiable and parsimonious set of dynamic capabilities. Our starting point is
Teece et al.’s (1997) (reconfiguring, learning, integrating, and coordinating) and
Teece’s (2007) (sensing the environment to seize opportunities and reconfigure
assets) distinct capabilities. Because different labels have been used in the literature to refer to similar capabilities, or similar labels for different capabilities, we
reconciled the various labels and meanings from the literature, and grouped them
under a parsimonious set to reflect Teece et al.’s (1997) and Teece’s (2007) conceptualization, our own interpretation of the literature, and relevance to NPD business
units. Specifically, dynamic capabilities are viewed as tools that enable the reconfiguration of existing operational capabilities (Galunic & Eisenhardt, 2001).
Pavlou and El Sawy
243
Figure 1: A framework for representing the proposed measurable model of
dynamic capabilities.
Spot, interpret, and
pursue opportunities
Revamp existing
operational capabilities
with new knowledge
Sensing
Capability
Need to revamp
existing
operational
External & Internal
capabilities
Stimuli
Existing
Operational
Capabilities
Embed new knowledge into
operational capabilities with
collective sense-making
Deploy tasks, resources, and
activities in reconfigured
operational capabilities
Integrating
Capability
Learning
Capability
Need to combine the
new knowledge in
operational capabilities
Coordinating
Capability
Need to
synchronize
tasks, resources,
and activities
Reconfigured
Operational
Capabilities
Reconfiguring operational capabilities and deploying new ones to address
turbulent environments is the ultimate goal of dynamic capabilities that seek to
achieve evolutionary fitness and prevent rigidities (Teece, 2007). Reconfiguration
refers to the appropriateness (Galunic & Rodan, 1998), timeliness (Zott, 2003),
and efficiency (Kogut & Zander, 1996) by which operational capabilities are reconfigured to fit the environment. Reconfiguration is markedly relevant in NPD where
new products are the outcome of reconfigured operational capabilities (Henderson
& Clark, 1990). The proposed dynamic capabilities that are proposed as tools for
reconfiguring existing operational capabilities are: (i) sensing; (ii) learning, (iii)
integration, and (iv) coordination capabilities. Those are graphically presented in
Figure 1, which briefly describes each capability, and it also explains the logic
by which the proposed dynamic capabilities help managers reconfigure the operational capabilities of their NPD units into new, more relevant ones that better
match the environment. The proposed dynamic capabilities are neither exhaustive,
nor sufficient for reconfiguration to occur, but they are posited as important enablers of the ability to reconfigure operational capabilities, as described in detail
below.
For simplicity of illustration, the proposed framework (Figure 1) presents
the four dynamic capabilities as interacting in a sequential logic to reconfigure
existing operational capabilities. However, there are reciprocal relationships among
these capabilities, which are theorized in the description of the proposed dynamic
capabilities.
Sensing capability
Reconfiguration requires a surveillance of market trends and new technologies to
sense and seize opportunities. Teece et al. (1997, p. 521) note: “The ability to
calibrate the requirements for change and to effectuate the necessary adjustments
would appear to depend on the ability to scan the environment, to evaluate markets
and competitors, and to quickly accomplish reconfiguration ahead of competition.” Sensing capability is defined as the ability to spot, interpret, and pursue
244
The Elusive Black Box of Dynamic Capabilities
opportunities in the environment. In NPD, business units must sense the environment to gather market intelligence on market needs, competitor moves, and new
technologies in order for managers to identify new product opportunities, and
decide to engage in exploratory early-stage research activities to pursue these
opportunities with new product prototypes.
The three basic routines of the sensing capability are: (i) generating market intelligence (Galunic & Rodan, 1998), (ii) disseminating market intelligence
(Kogut & Zander, 1996), and (iii) responding to market intelligence (Teece, 2007).
These routines are related to kindred routines in the dynamic capabilities literature.
Generating market intelligence relates to identifying customer needs (Teece, 2007),
being responsive to market trends (Amit & Schoemaker, 1993), identifying market
opportunities (Day, 1994), recognizing rigidities (Sinkula, 1994), and detecting
resource combinations (Galunic & Rodan, 1998). Disseminating market intelligence relates to interpreting market intelligence (Kogut & Zander, 1996), making
sense of events and developments, and exploring new opportunities (Teece, 2007).
Responding to market intelligence also relates to initiating plans to capitalize on
market intelligence (D’Aveni, 1994), and pursuing specific market segments with
plans to seize the new market opportunities (Teece, 2007).
The sensing capability of NPD units is proposed to enable the reconfiguration
of their existing operational capabilities. First, generating market intelligence raises
the NPD unit’s potential to identify new market opportunities for reconfiguration
(Zahra & George, 2002). Second, disseminating market intelligence helps NPD
units to achieve responsiveness to customer needs (Day, 1994). Third, responding
to market intelligence promotes product innovation and enables NPD units to
explore emergent opportunities for new products that better meet customer needs
(Jaworski & Kohli, 1993).
Learning capability
Once a market opportunity is identified, it must be addressed with new products,
which require a decision to revamp existing operational capabilities with learning,
and new knowledge and skills (Teece, 2007). For NPD units to take advantage
of market opportunities in a changing environment, they must engage in learning
to find new solutions, create new knowledge, and reconfigure existing operational
NPD capabilities to develop new products. There is a reciprocal two-way relationship between sensing and learning capabilities because learning enhances
the ability of NPD units to detect new opportunities (Cohen & Levinthal, 1990;
Zahra & George, 2002). Sensing and learning capabilities are distinct capabilities because sensing focuses on gathering new market intelligence, and learning
focuses on using market intelligence to create new knowledge (Hurley & Hult,
1998).
Learning capability is defined as the ability to revamp existing operational
capabilities with new knowledge. According to Zahra and George (2002) who
developed absorptive capacity (learning) as a dynamic capability, the four underlying routines of the proposed learning capability are acquiring, assimilating,
transforming, and exploiting knowledge. These routines relate to kindred terms in
the dynamic capabilities literature. First, acquiring knowledge relates to obtaining
Pavlou and El Sawy
245
new knowledge (Cohen & Levinthal, 1990). Second, assimilating knowledge relates to knowledge articulation (Zander & Kogut, 1995) and knowledge brokering
(Eisenhardt & Martin, 2000). Third, transforming knowledge relates to innovative
problem-solving (Iansiti & Clark, 1994), brainstorming (Pisano, 1994), and creative new thinking (Henderson & Cockburn, 1994). Finally, exploiting knowledge
relates to pursuing new initiatives (Van den Bosch, Volberda, & De Boer, 1999),
seizing opportunities with learning (Teece, 2007), and revamping operational capabilities (Grant, 1996).
Cohen and Levinthal (1990, p. 131) suggest that learning helps groups become more proactive by enhancing their “creative capacity.” Van den Bosch et
al. (1999) further argue that learning facilitates reconfiguration and innovation.
Therefore, learning is proposed as an enabler of reconfiguration by helping revamp existing operational capabilities (Zollo & Winter, 2002).
Integrating capability
Reconfiguration relies on integrating new resources, and assets (Galunic & Eisenhardt, 2001). This is because the reconfiguration of existing operational capabilities
requires a collective logic and shared interaction patterns (Okhuysen & Eisenhardt,
2002). Because new knowledge created by learning is mostly owned by individuals,
it must be integrated to a collective level (Teece, 1982). In the NPD context, because operational capabilities are supra-individual and do not reside in any specific
individual, NPD units must integrate their individual knowledge and patterns of
interaction into a collective system to deploy the new configurations of operational
capabilities.
Integrating capability is defined as the ability to combine individual knowledge into the unit’s new operational capabilities. Its routines—contribution, representation, and interrelation of individual input to the collective business unit—are
closely related to the dynamic capabilities literature. Specifically, contribution
relates to disseminating individual input within the business unit (Okhuysen &
Eisenhardt, 2002). Representation relates to visualizing how people fit in, how
other people act, and how the unit’s activities fit together (Crowston & Kammerer, 1998). Interrelation relates to integrating individual inputs within a unit to
hone the reconfigured operational capabilities by executing a collective activity
(Helfat & Peteraf, 2003).
Integrating capability is proposed to facilitate reconfiguration through its
three basic routines. First, contribution to the unit helps collect and combine individual inputs. Second, representation builds a shared understanding, creates a
common ground, and develops new perceptual schema (Weick & Roberts, 1993).
Third, because reconfiguration requires a new logic of collective interaction, interrelation helps the routinization of the reconfigured operational capabilities (Okhuysen & Eisenhardt, 2002). Weick and Roberts (1993, p. 377) argue that groups with
more integrated capabilities can better react in novel situations, whereas Zollo and
Winter (2002, p. 340) view dynamic capability as a collective activity by arguing
that reconfiguring in a disjointed way does not even exercise a dynamic capability. Finally, Teece (2007) views the integration of knowledge as a foundation of
dynamic capabilities.
246
The Elusive Black Box of Dynamic Capabilities
Coordinating capability
Because the new configurations of operational capabilities require effective coordination of tasks and resources and synchronization of activities (Iansiti &
Clark, 1994; Helfat & Peteraf, 2003), coordinating capability enables reconfiguration by administering tasks, activities, and resources to deploy the reconfigured operational capabilities. Coordinating capability is defined as the ability
to orchestrate and deploy tasks, resources, and activities in the new operational
capabilities. The basic routines of coordinating capability also draw upon the
dynamic capabilities literature, namely assigning resources to tasks (Helfat &
Peteraf, 2003), appointing the right person to the right task (Eisenhardt & Brown,
1999), identifying complementarities and synergies among tasks and resources
(Eisenhardt & Galunic, 2000), and orchestrating collective activities (Henderson,
1994).
Although the integrating capability is positively associated with the coordinating capability because coordination is enhanced by a shared language
(Galunic & Eisenhardt, 2001), the integrating and coordinating capabilities are
theoretically and empirically distinct (Kogut & Zander, 1996). Although coordination focuses on orchestrating individual tasks and activities, integration focuses
on building an overall collective sense-making and understanding (Crowston &
Kammerer, 1998).
Coordinating capability is proposed to facilitate the reconfiguration
of operational capabilities. First, it enables NPD units to recognize, assemble, and allocate resources (Collis, 1994) by facilitating the dissemination of market intelligence across the unit (Vorhies & Harker, 2000).
Second, coordinating capability helps NPD units assign the right person to the
right task (Eisenhardt & Brown, 1999). Third, coordinating capability helps
NPD units better synchronize their tasks and activities (Helfat & Peteraf,
2003).
Teece et al. (1997) argue “[dynamic] capability is embedded in distinct
ways of coordinating” (p. 519). Teece (2007, p. 1338) also argued: “In short,
both innovation and reconfiguration may necessitate cospecialized assets being
combined by management in order for (systemic) innovation to occur.” Okhuysen and Eisenhardt (2002, p. 382) argue that effective allocation of resources
enhances the flexibility of NPD units by helping assign the right person to
the right task, an essential element of successful reconfiguration (Eisenhardt &
Brown, 1999). Also, Quinn and Dutton (2005) noted: “coordination is the process people use to create, adapt, and re-create organizations” (p. 36). Thus, coordinating capability helps implement and deploy the reconfigured operational
capabilities.
Table 1 summarizes the definitions of the proposed dynamic capabilities, and
it also shows that many of the routines discussed in the dynamic capabilities literature can be categorized under each capability. This suggests that our proposed
set of dynamic capabilities is closely linked to the dynamic capabilities literature through their underlying routines. This is consistent with Brown and Eisenhardt (1997) who view dynamic capabilities as complex combinations of simpler
routines.
Pavlou and El Sawy
247
Table 1: Definition of proposed capabilities and links to the dynamic capabilities
literature.
Capability
Definition
Sensing
capability
The ability to spot, interpret,
and pursue opportunities
in the environment.
Learning
capability
The ability to revamp
existing operational
capabilities with new
knowledge.
The ability to embed new
knowledge into the new
operational capabilities by
creating a shared
understanding and
collective sense-making.
Integrating
capability
Coordinating
capability
The ability to orchestrate and
deploy tasks, resources,
and activities in the new
operational capabilities.
Basic Routines
• Generating market intelligence
(Galunic & Rodan, 1998)
• Disseminating market intelligence
(Kogut & Zander, 1996)
• Responding to market intelligence
(Teece, 2007)
• Acquiring, assimilating,
transforming, and exploiting
knowledge (Zahra & George, 2002)
• Contributing individual knowledge to
the group (Okhuysen & Eisenhardt,
2002)
• Representation of individual & group
knowledge (Crowston & Kammerer,
1998)
• Interrelation of diverse knowledge
inputs to the collective system (Grant,
1996)
• Assigning resources to tasks (Helfat
& Peteraf, 2003)
• Appointing right persons to right
tasks (Eisenhardt & Brown, 1999)
• Identifying synergies among tasks,
activities, and resources (Eisenhardt
& Galunic, 2000)
• Orchestrating activities (Henderson,
1994)
The Effects of Dynamic Capabilities on Operational Capabilities
in Turbulence
Although dynamic capabilities have certain commonalities (Eisenhardt & Martin,
2000), they are likely to differ across business units because their complex nature
makes them difficult to imitate (Galunic & Eisenhardt, 2001). Due to their inherent complexity (Collis, 1994), path dependence (Teece et al., 1997), and causal
ambiguity (Lippman & Rumelt, 1982), dynamic capabilities are herein proposed
to differentially affect performance. However, the exact mechanisms by which dynamic capabilities influence performance are still not very well understood (Zott,
2003), and dynamic capabilities are often confounded with their effects. This
is why dynamic capabilities have been criticized as being tautologically linked
to performance (Williamson, 1999). Therefore, to distinguish dynamic capabilities from their effects on operational capabilities and performance, we propose a
248
The Elusive Black Box of Dynamic Capabilities
Figure 2: The structural model of the effects of dynamic capabilities in environmental turbulence.
Environmental Turbulence
H2
Dynamic
Capabilities
H1
Operational
Capabilities
Performance
in NPD
structural model where the effect of dynamic capabilities on performance in NPD
is mediated by operational capabilities and moderated by environmental turbulence
(Figure 2).
In the NPD literature, performance in NPD is composed of two dimensions of
product success (e.g., Clark & Fujimoto, 1991; Griffin, 1997; Kusunoki, Nonaka,
& Nagata, 1998): (i) product effectiveness (product quality, innovativeness) and
(ii) process efficiency (time to market at a low cost). Given the long-held trade-off
in NPD between product quality, timeliness, and cost (Cohen, Eliashberg, & Ho,
1996), we view NPD performance as the achievement of product effectiveness
and process efficiency. This is consistent with the marketing literature that suggests that customers seek quality products in a timely fashion and at a low price
(Vorhies & Harker, 2000). This is also consistent with the NPD literature that
empirically shows that successful new products offer a favorable combination of
high product quality, timeliness, and low cost (Atuahene-Gima & Li, 2004).
The primary operational capabilities in NPD needed for developing new
products are technical, customer, and managerial (Danneels, 2002). First, technical capability is the ability to physically develop new products by understanding
product technologies, evaluating the feasibility of product designs, testing prototypes, and assessing technical specifications (Pisano, 1994). Second, customer
capability is the ability to market the new products to customers through advertising, distributing, pricing, selling, and order entry (Day, 1994). Third, managerial
capability is the ability to administer activities at the operational level of the NPD
unit by monitoring and reporting progress, designing incentives, and managing
conflicts (Danneels, 2002). These three capabilities comprise the primary operational NPD capabilities of NPD business units.
New products are essentially the manifestation of an NPD unit’s operational
capabilities (Teece, 1982), and successful new products are directly dependent on
operational NPD capabilities (Clark & Fujimoto, 1991). Operational capabilities,
however, are not product specific, but they are the platform for developing any
new products (Griffin, 1997). Operational NPD capabilities offer the potential
to build technically sophisticated new products that better meet customer needs
with the aid of management (Song, Droge, Hanvanich, & Calantone, 2005). In
contrast, outdated operational NPD capabilities that do not match market needs
and do not follow technological developments—termed “rigidities” by LeonardBarton (1992)—result in poor process efficiency and low product effectiveness,
Pavlou and El Sawy
249
and inferior new products. Thus, NPD performance is directly dependent upon
operational capabilities.
The effect of dynamic capabilities on operational NPD capabilities
The value of dynamic capabilities is in the new configurations of operational
capabilities they help reconfigure (Eisenhardt & Martin, 2000). Although dynamic capabilities can reconfigure operational capabilities in various directions
(Helfat & Peteraf, 2003), we focus on the appropriateness (how well the configurations of operational capabilities match the environment), timeliness (how
quickly the operational capabilities are reconfigured into new configurations), and
efficiency (how economical and cost-effective the reconfiguration is) by which the
existing operational NPD capabilities are reconfigured into new ones that better
match the environment.
First, by having the dynamic capability to generate, disseminate, and respond
to market intelligence (sensing capability), NPD units are more likely to quickly introduce new products that better match customer needs (technical capability), thus
making it easier to sell such products (customer capability). Second, by quickly
generating new knowledge on technological breakthroughs (learning capability),
NPD business units are more likely to develop technologically sophisticated new
products (technical capability). Third, by being effective in integrating knowledge
and interaction patterns (integrating capability), and in orchestrating resources,
tasks, and activities (coordinating capability), NPD units are more likely to administer the new NPD activities by monitoring progress, designing incentives, and
managing internal conflicts (managerial capability). Thus, dynamic capabilities
that are responsible for selecting the new products are proposed to be positively
associated with operational NPD capabilities that are responsible for physically
building, managing, and selling new products. Empirical evidence suggests that
dynamic capabilities are associated with high product quality and fast cycle time
(Henderson & Clark, 1990; Iansiti & Clark, 1994). Integrating these arguments,
we hypothesize:
H1: An NPD unit’s dynamic capabilities are positively associated with its
operational capabilities.
H1 suggests that dynamic capabilities indirectly influence performance by reconfiguring existing operational capabilities into superior ones that better match the
changing environment. Although operational capabilities have a direct effect on
performance at any “snapshot” in time, interpreting H1 from a longitudinal standpoint, dynamic capabilities have a longitudinal effect on performance over time by
reconfiguring new operational capabilities that offer a series of new products that
better match the environment. This logic is consistent with Teece et al. (1997) who
argue that new products at any given point in time depend on operational capabilities in NPD, which depend on dynamic capabilities from a longitudinal standpoint.
Collis (1994) also explains that dynamic capabilities supersede operational capabilities over time. The proposed structural model (Figure 2) thus views dynamic
capabilities independent of performance, overcoming the tautological criticism of
the dynamic capabilities view (Williamson, 1999).
250
The Elusive Black Box of Dynamic Capabilities
Dynamic capabilities and operational capabilities in turbulent environments
The proposed effect of dynamic capabilities on operational capabilities is proposed
to be moderated by the level of environmental turbulence, which is defined in terms
of the frequency and amplitude of change in the environment and general conditions of uncertainty (Duncan, 1972). In NPD, environmental turbulence consists
of two primary sources: (i) market turbulence—uncertainty in market demands
and competitor moves; and (ii) technological turbulence—frequency of technical
breakthroughs (Jap, 2001).
Turbulent environments spawn new opportunities (Sull, 2009; Van den Bosch
et al., 1999), thus creating incentives to employ dynamic capabilities to reconfigure existing operational capabilities to pursue new opportunities. Because turbulent
environments create a discrepancy between existing and ideal operational capabilities (Fredrickson & Mitchell, 1984), the need for reconfiguration enhances the
value of dynamic capabilities. Teece et al. (1997) argue that there is great value
in the ability to reconfigure resources in turbulent environments, whereas Rindova
and Kotha (2001) argue that turbulent environments make it more likely to reconfigure operational capabilities. Also, rigidities act as alerts (Zahra & George,
2002), forcing NPD business units to take advantage of dynamic capabilities to
reconfigure their rigid and outdated operational capabilities.
The moderating role of environmental turbulence is also formally supported
by options theory. An option is acquired with a partial investment, and the holder
reserves the right to strike the option if the opportunity arises (Sambamurthy,
Bharadwaj, & Grover, 2003). Dynamic capabilities can be viewed as options
(Kogut & Zander, 1992; Moorman & Slotegraaf, 1999), offering the ability to
pursue new market opportunities that the environment creates. The higher the
degree of environmental turbulence, the more likely these options will become
valuable because more opportunities are likely to emerge. Even if options are
costly, NPD units that have the option to reconfigure their existing operational
NPD capabilities are more likely to end up with operational capabilities that
match the environment. On the other hand, less turbulent environments are less
likely to create opportunities for reconfiguring existing operational capabilities,
and investments in such options without the occasion to exercise the dynamic
capabilities may end up having little or even no value, while they are likely to
carry a cost burden (Winter, 2003, p. 993).
Applied to NPD, environmental turbulence enhances the value potential of
new products (Griffin, 1997). Dynamic capabilities thus become more valuable in
turbulent environments where NPD units need to reconfigure their existing operational NPD capabilities to build new products that better match the environment.
Environmental turbulence is expected to negatively moderate the positive effect
of operational capabilities on performance (Pavlou & El Sawy, 2006). This is
because operational capabilities help achieve process efficiencies in developing
a given product. However, turbulent environments erode the value potential of
existing products due to changes in market needs, technologies, and rival products (Danneels, 2002). Thus, operational capabilities may become rigidities due to
inertia and unwillingness to change (Leonard-Barton, 1992). Because operational
capabilities require a costly, time consuming, and often irreversible configuration of
existing resources, their frequent reconfiguration is likely to disrupt their efficiency
Pavlou and El Sawy
251
(Zammuto, 1988). Thus, the more turbulent the environment is, the more likely the
operational capabilities will become rigid. In contrast, less turbulent environments
favor more disciplined decision making (Brown & Eisenhardt, 1997), and they
reward the efficient exploitation of operational capabilities. Thus, environmental
turbulence is expected to reduce the value potential of existing operational NPD
capabilities on performance in NPD. Consequently, we hypothesize:
H2: The positive relationship between an NPD unit’s dynamic capabilities
and its operational capabilities is positively moderated (reinforced) by
environmental turbulence.
MEASUREMENT DEVELOPMENT
Wherever possible, measurement items were adapted from existing scales. For
new measures, standard scale development procedures were followed that helped
establish the face validity and content validity for the new measures. All measures
were adapted to the study’s unit of analysis (NPD business units), and they were
phrased relative to the competition (Appendix A).
The study’s unit of analysis is the NPD unit, which may be either intraor interfirm consistent with De Boer et al. (1999). Following the relational view
(Dyer & Singh, 1998) even if interfirm NPD units are formed through a partnership
among firms, they are distinct entities on their own right that assume their own
trajectory (Song et al., 2005). D’Adderio (2001) explains how resources are integrated across firm boundaries, and Ettlie and Pavlou (2006) and Lane and Lubatkin
(1998) show that interfirm partnerships have their own distinct capabilities.
Dynamic capabilities
Extending Ettlie and Pavlou (2006) and Pavlou and El Sawy (2006), the proposed
set of dynamic capabilities was captured with a formative model (Figure 3). The
relationship between first- and second-order constructs can be either reflective or
formative in nature (Edwards, 2000). Reflective ones assume that the second-order
factor “causes” the first-order factors. Formative ones assume that the second-order
factor is “caused” by each of the first-order factors where each factor represents
a unique input. The model posits formative indicators for the second-order latent
reconfiguration capability and reflective indicators for the first-order measurable
capabilities.
Because dynamic capabilities are abstract, intangible, and difficult to describe, they are modeled with a second-order model that is formed by the proposed
Figure 3: The proposed measurement model of dynamic capabilities.
Indicator 1
Dynamic Capabilities
Indicator 2
Sensing Capability
Learning Capability
Integrating Capability
Coordinating Capability
252
The Elusive Black Box of Dynamic Capabilities
four dynamic capabilities that are measurable. As each first-order dynamic capability is posited to “form” or enable reconfiguration, a formative second-order
model is more appropriate than a reflective model. A reflective model would imply
that reconfiguration “reflects” or causes the four first-order dynamic capabilities.
As with all models, the proposed simple formative model of dynamic capabilities
is an abstraction from reality and cannot capture the complexity of dynamic capabilities. Yet, it allows the direct measurement of dynamic capabilities through
the first-order measurable constructs. The proposed formative model is consistent with Diamantopoulos and Winklhofer’s (2001) guidelines. These principles
are: (i) specifying the content domain of the second- and first-order constructs,
(ii) proposing the effects of the first-order on the second-order construct, (iii)
specifying the relationships and distinctions among the first-order constructs, and
(iv) proposing the role of the second-order construct (dynamic capabilities) on the
study’s dependent variables (NPD performance).
Formative models must maintain theoretical distinction among their firstorder constructs so that each construct contributes a unique component to the
second-order construct. As explained earlier, each of the four dynamic capabilities
is distinct from each other with each of the four capabilities offering a unique
component to the overall ability to reconfigure existing operational capabilities into
new ones. The resulting formative model of dynamic capabilities is a measurable
model that can be readily used to capture dynamic capabilities through the four
underlying components.
Sensing capability captures the generation, dissemination, and responsiveness to market intelligence (Jaworski & Kohli, 1993). Learning capability captures
the acquisition, assimilation, transformation, and exploitation of knowledge (Cohen & Levinthal, 1990; Zahra & George, 2002). Coordinating capability captures
resource allocation, task assignment, and synchronization (Crowston, 1997). Integrating capability captures the contribution, representation, and interrelation of
individual input to the entire business unit (Weick & Roberts, 1993).
For identification purposes, dynamic capabilities were also measured with
two direct indicator items (Appendix A) that capture the potential for reconfiguration. The purpose of measuring second-order constructs with (direct) indicator
items is to test that the indirect measurement through the first-order dynamic capabilities is also consistent with their indicator items (Figure 3). All measurement
items of dynamic capabilities are shown in Appendix A.
Operational NPD capabilities
To operationalize operational capabilities in NPD, a formative second-order model
(Figure 4) is proposed. A formative model is deemed more appropriate than a reflective model to represent operational capabilities in NPD because the three proposed operational capabilities are complementary to each other (Nerkar & Roberts,
2004), and each capability contributes a unique component to the development of
new products (Song et al., 2005).
As shown in Appendix A, customer and technical capabilities were measured
by following Song and Parry (1997), whereas managerial capability was measured
following Sethi, Smith, and Park (2001). Two reflective indicator items were used
Pavlou and El Sawy
253
Figure 4: The proposed measurement model of operational new product development capabilities.
Operational NPD Capabilities
Indicator 1
Indicator 2
Customer Capability
Technical Capability
Managerial Capability
Figure 5: The proposed measurement model of new product development performance.
NPD Performance
Indicator 1
Indicator 2
Process Efficiency
Product Effectiveness
to measure aggregate operational NPD capability (Vorhies & Harker, 2000). These
indicator items tested whether the proposed customer, technical, and managerial
capabilities jointly span the domain of operational NPD capability.
NPD performance
Overall NPD performance is operationalized with a second-order formative model
formed by process efficiency and product effectiveness (Figure 5). This view places
different weights on the two dimensions, consistent with the logic that firms often
focus on one dimension over the other. Still, firms cannot place excessive emphasis
on any one dimension due to the risk of being left behind by competitors that focus
on both (Sethi, 2000). The correlation between product effectiveness and process
efficiency was r = .16 (p = .049), which is not particularly high, supporting the
trade-off (Clark & Fujimoto, 1991). Process efficiency and product effectiveness
were measured following Kusunoki, Nonaka, and Nagata (1998) and AtuaheneGima and Li (2004). Because archival performance sources are unavailable for
NPD business units, two reflective indicator items for overall NPD performance
were measured (Jap, 2001).
Despite the use of perceptual scales for NPD performance given the lack
of archival data, subjective scales have their own merits because they permit
meaningful comparisons across firms (Song et al., 2005), whereas objective scales
do not have a high level of specificity in terms of industry, time horizon, and
economic conditions (Song & Parry, 1997).
Environmental turbulence in NPD
Environmental turbulence was also modeled as a formative second-order model
(Figure 6) with market and technological turbulence as formative constructs.
Market and technological turbulence were measured following Jaworski and
Kohli (1993) to capture the pace of changes in customer needs and technological
254
The Elusive Black Box of Dynamic Capabilities
Figure 6: The proposed measurement model of environmental turbulence.
Environmental Turbulence
Indicator 1
Indicator 2
Market Turbulence
Technological Turbulence
Table 2: Control variables.
New product development (NPD) unit cross-functional integration: This construct
describes the quality of interaction in the NPD unit’s functional areas, which was
shown in the literature to influence NPD success (Song & Parry, 1997). It was measured
with three items following Song and Parry (1997), and it was controlled for its effect on
performance, dynamic capabilities, and operational capabilities in NPD.
NPD unit experience: Experience in NPD measures how many years the unit has been
working together (Song & Parry, 1997), and it is controlled for its effect on
performance, dynamic capabilities, and operational capabilities in NPD.
Business unit innovation type: NPD projects differ in terms of their innovation type,
ranging from routine engineering of existing products, to building incremental new
products, to creating radically new products. The innovation type of the NPD unit is
controlled for its impact on dynamic capabilities and performance.
Business unit size: The group’s number of NPD workers was also controlled for.
Business unit functional areas: The number of functional areas (e.g., marketing, research
and development, engineering) of the NPD unit was controlled for its effect on NPD
performance.
Business unit manager: Whether the group’s manager was a senior or mid-level manager
was also controlled on performance, dynamic capabilities, and operational capabilities
in NPD.
Firm size: The firm’s total number of employees and annual revenues were controlled for
their effects on NPD performance.
Industry concentration: Industry concentration is measured as the annual sales revenues
for the largest four firms in each 4-digit standard industrial classification code divided
by the sales for all firms in the industry.
Industry segment: Industries can be classified as either high tech or low tech industries,
following Francis and Schipper’s (1999) classification scheme that is based on the
3-digit SIC industry code. Industry segment case controlled for potential effects on
NPD performance.
breakthroughs. Overall environmental turbulence was also measured with two new
direct items for validation purposes.
Finally, a number of business unit, firm, and industry variables were controlled for (Table 2).
DATA COLLECTION, ANALYSIS, AND RESULTS
Survey Administration
Two distinct studies were conducted using the same survey instrument and
data collection methods. Key informant methodology was employed with NPD
Pavlou and El Sawy
255
managers as the key respondents (Sethi et al., 2001). This is because NPD managers are typically aware of the major competing NPD units in competitor firms that
develop similar products. The survey instrument was first pretested with personal
interviews with six academics with focus on NPD, and with 12 NPD managers
who offered comments on shaping the measurement items to the NPD context.
It was then tested in a small-scale study with 33 NPD managers to examine the
statistical properties of the study’s measures. For study I, NPD managers were
selected from the 554 participants of the Product Development and Management
Association conference (www.pdma.org). For study II, the sample was drawn from
the 161 participants of the roundtable management conference (www.CoDev.org).
The list was refined by contacting the participants, and inquiring if they had been
acting as NPD managers. The final list contained 386 (study I) and 121 (study II)
NPD managers. The NPD managers were asked to self-select an NPD business
unit that they have been managing. To avoid social desirability bias, they were
asked to select a NPD business unit that they are mostly familiar with, and not a
typical, successful, or failed one.
To address social desirability bias (Sethi, 2000), the performance of all NPD
work units was analyzed. The mean of the success outcomes was 3.44 on a 5-point
scale (STD = .78), which was roughly in the middle of the scale. Therefore, social
desirability bias is not a serious concern in this study.
Invitation e-mails were sent to the selected respondents, explaining the
study’s purpose and assuring that their responses would remain strictly confidential, and all results would only be reported in aggregate. The respondents were
asked to click on a link in the e-mail message, which directed them to our online
survey instrument (Appendix A). The respondents were offered as incentive a
customized report that summarized the results of the study and benchmarked their
NPD business unit. Over 90% of the respondents requested this customized report
that was sent within a month.
To assure a collective response, the instructions asked the NPD managers
to obtain input from other members of their NPD units. Ex-post communication
verified that the NPD managers did consult with other unit members. To focus
on the NPD unit, the instructions explicitly asked the respondents to focus on
their NPD unit’s attributes, not their firm’s attributes. A formal pretest and expost communication verified that the respondents focused on their NPD unit’s
attributes. Another formal check assessed the respondents’ familiarity with their
NPD unit. On a familiarity item, all respondents were deemed familiar (mean = 4.3,
STD = .8, minimum = 4.0), and all responses were thus retained. To collect a
roughly equal number of intra- and interfirm NPD units, the respondents were
asked to favor interfirm NPD units, which helped collect 56% the responses from
interfirm NPD units.
Because dyadic data from interfirm NPD work units were desirable, if the
participants selected an interfirm NPD unit, they were asked to provide the contact
information of the partner’s NPD manager. These NPD managers were then contacted with the same procedure. From the 99 interfirm business units, 47 names
were received, and 28 matched pairs were obtained (60% response rate). Dyadic
data were separately analyzed for similarities (Jaworski & Kohli, 1993). The average absolute difference for all constructs were less than 5%, the average correlation
256
The Elusive Black Box of Dynamic Capabilities
between the two NPD managers was .63 (range = .17–.87), and the interrater reliability was .71 (p < .01). These results indicate no systematic bias between the
two informants, and their responses were averaged to derive a single score for
each unit. In study I, 32% were suppliers, 19% customers, and 4% were alliance
partners; 45% were internal business units. In study II, 27% were suppliers, 21%
were customers, and 10% were alliance partners; 42% were internal business units.
In study I, of the 386 respondents, 44 could not be contacted, 12 respondents
indicated that firm policy forbids their participation, and 15 of the invitees indicated
that they were not qualified to participate in the study. Following two e-mail
reminders, 121 responses were received (39% response rate). In study II, of the 161
participants, 25 were unreachable, and four were unable to respond. Following two
e-mail reminders, a total of 59 responses were obtained (43% response rate). The
response rate is relatively high because: (i) personal communication was sought
with the participants, (ii) the study was endorsed by the conference organizers,
(iii) we participated in both conferences and established contacts, and (iv) several
completed responses through paper questionnaires were collected during the two
conferences. In sum, a total of 180 usable responses were obtained for 180 NPD
units from 180 distinct firms.
Nonresponse bias was assessed by verifying that early and late respondents
did not differ in their responses. The early respondents were identified by selecting
those that responded in the first two weeks. All possible t-test comparisons between
the means of the two groups in the two empirical studies showed insignificant
differences (p < .1 level).
Responses were collected from the high-tech (14%), manufacturing (12%),
medical devices (11%), consumer goods (8%), and telecom (7%) industries. Less
than 5% were collected from the chemical, electronics, auto, and food industries.
80% of the respondents were NPD managers and 10% were NPD executives.
The NPD purpose was mostly applied product development (68%), basic research
(23%), and routine engineering (9%). Demographics and descriptive statistics were
similar between the intra- and interfirm NPD work units. Using Chow’s (1960)
test statistic and Wilk’s lambda, the results of both the intra- and interfirm samples
were statistically similar. The data from both the intra- and interfirm NPD units
were thus pooled together for a single data analysis.
The Measurement Model
Convergent validity, discriminant validity, and unidimensionality were first assessed through an exploratory factor analysis, which showed an ideal loading pattern (Appendix B). The measures were also assessed with LISREL’s
confirmatory factor analysis (CFA). The CFA fit indices were: χ 2 741 = 1067,
χ 2 /df = 1.44, goodness of fit index (GFI) = .91, adjusted goodness of fit index (AGFI) = .90, normed fit index (NFI) = .95, comparative fit index (CFI)
= .97, root mean residual (RMR) = .04. A χ 2 /df ratio below 3 implies a
good fit; GFI, AGFI, NFI, and CFI above 0.90 are considered acceptable;
RMR below .05 is desirable. Good model fit indices give evidence of convergent validity and unidimensionality. Convergent validity was supported by the
large and significant standardized loadings for both the first- and second-order
Pavlou and El Sawy
257
Table 3: Correlation matrix and composite factor reliability scores for principal
constructs.
Construct
1. Dynamic capabilities
2. Sensing
3. Learning
4. Integrating
5. Coordinating
6. New product development
(NPD) performance
7. Environmental turbulence
8. Operational capabilities
1
2
3
4
5
6
.91
.50
.55
.52
.47
.40
.78
.60
.48
.45
.31
.88
.52
.50
.35
.84
.61
.33
.85
.31
.85
.25
.40
.29
.22
.24
.36
.19
.39
.15
.42
−.14
.54
7
8
.74
−.17
.82
Items on diagonal (in bold) represent reliability scores. Correlations above .15 are significant
(p < .05); above .20 (p < .01).
constructs (p < .001), and the t-values that exceeded the 2.0 threshold. Convergent validity was also supported by calculating the ratio of factor loadings to their
respective standard errors, which exceeded |2.0| (p < .01). Discriminant validity
was tested by showing that the measurement model had a significantly better model
fit to a competing model with a single latent construct, and with all other competing models in which pairs of latent constructs were joined. The χ 2 difference
between the competing models was significantly larger than that of the original
model, as also suggested by the factor loadings, modification indices, and residuals
(Marsh & Hocevar, 1985). In sum, these tests confirm convergent and discriminant
validity and unidimensionality.
Finally, reliability was assessed using both of the composite factor reliability scores. All measures exceeded .70 (Table 3), suggesting adequate reliability.
Reliability was also supported because the average variance extracted (Hair, Anderson, Tatham, & Black, 1995) exceeded .70 for all factors. Common method
bias was first tested with an exploratory factor analysis in which the study’s 38
measurement items factored in 11 constructs (eigenvalue > 1) that jointly explain
79% of the total variance (Appendix B). Besides, common method bias was also
tested with a CFA analysis in which a single-factor model was generated with all
38 measurement items joined into a single latent construct. This model had a poor
model fit (χ 2 741 = 3874, χ 2 /df = 5.23, GFI = .075, AGFI = .071, NFI = .70,
CFI = .77, RMR = .10). These tests did not suggest the presence of common
method bias (Podsakoff & Organ, 1986).
The Structural Model
The structural model was tested with LISREL by simultaneously estimating the
second-order models along with the structural relationships following Joreskog and
Goldberger’s (1975) “multiple indicators and multiple causes” model with both
formative and also reflective items (e.g., Diamantopoulos & Winklhofer, 2001).
The interaction effects of environmental turbulence were modeled following Ping
(1995), centering all variables to avoid identification problems (Cortina, Chen, &
Dunlap, 2001). Modeling interaction effects is still a debated issue in the literature;
258
The Elusive Black Box of Dynamic Capabilities
Figure 7: LISREL results on proposed structural model.
Market
Turbulence
.58**
H2:
.30**
.27**
Integrating
Capability
Coordinating
Capability
.46**
Environmental Turbulence
Sensing
Capability
Learning
Capability
Second-Order Constructs
** Significant at p<.01
* Significant at p<.05
Variance Explained shown in bold
Technological
Turbulence
Product
Effectiveness
-.21**
.56**
.36**
.32**
Dynamic
Capabilities
H1:
.36**
Operational Capabilities
.34**
.30**
Process
Efficiency
Marketing
Capability
.40**
Technical
Capability
.41**
.47
.45**
Managerial
Capability
.47**
NPD Performance .63
.15** .13* .14* .11*
CrossFunctional
Integration
Business
Unit
Size
Firm
Size
High
Tech
Industry
although Kenny and Judd (1984) suggest the use of all cross products as indicators
of latent variables (inflating the degrees of freedom), Ping (1995) proposed
a single
indicator for a latent variable X ∗ Y formed by X and Y: X ∗
Y= X
∗
Yi , with
i
the loadings of the interaction variable X ∗ Y to be: λX ,Y = λXi ∗ λYi . Overall,
an acceptable model fit was achieved (χ 2 990 = 1,346, χ 2 /df = 1.36, GFI = .93,
AGFI = .91, NFI = .97, CFI = .96, RMR = .034) (Figure 7).
First, as hypothesized, dynamic capabilities have a significant effect on operational capabilities (b = .36, p < .01), thus supporting H1.
As expected by the NPD literature, operational capabilities have a strong
effect on performance (b = .41, p < .01). Given this significant effect, to examine
the mediating role of operational capabilities on the performance effects of dynamic
capabilities, another model was tested where the path between dynamic capabilities
and operational capabilities was omitted, and only a direct path between dynamic
capabilities and performance was included. The competing model had inferior
fit indices (χ 2 990 = 1,545, χ 2 /df = 1.76, GFI = .88, AGFI = .87, NFI = .92,
CFI = .91, RMR = .061). Because the fit deterioration between the two models
was significant (χ 2 = 199, p < .01), the proposed impact of dynamic capabilities
on performance in NPD is shown to be mediated by operational capabilities.
Second, environmental turbulence positively moderated the effect of dynamic
capabilities on operational NPD capabilities (b = .30, p < .01), thus supporting H2.
As expected by the literature (Pavlou & El Sawy, 2006), environmental turbulence
had a negative moderating (attenuating) role on the effect of operational capabilities
on NPD performance (b = −.21, p < .01).
The two moderating effects of environmental turbulence significantly increased the variance explained in the two dependent variables (Carte & Russell,
2003), while a significant model fit deterioration (p < .01) was observed when the
two moderating effects were omitted from the model. Along with the study’s significant control variables—cross-functional integration, unit size, firm size, high-tech
industry—the proposed model explains 63% of the variance in NPD performance.
Dynamic capabilities, operational capabilities, NPD performance, and environmental turbulence were modeled as second-order constructs, and they were
Pavlou and El Sawy
259
modeled simultaneously with the rest of the structural model (Figure 7). However, to test the superiority of the second-order models, we compared each of the
second-order models to a competing first-order model by calculating the target
coefficient index (T) (T = χ 2 first-order model /χ 2 second-order model ) (Marsh & Hocevar,
1985).
For each of the second-order factors, T > 0.90 (upper bound = 1.0), implying
that each of the second-order factors explains almost all of the covariance in the
corresponding first-order factor, implying that each of the second-order factors is
adequate (despite being more parsimonious). We also modeled a reflective secondorder model for each of the four constructs, which resulted in significantly poorer
model fit indices, suggesting a considerable fit deterioration, thus rendering a
reflective model less likely. In sum, these tests validate the proposed second-order
formative models. These findings suggest that the proposed LISREL model with
the second-order factors is a superior representation.
IMPLICATIONS FOR DECISION-MAKING THEORY
AND PRACTICE
This study has two key findings. First, it identifies and articulates a set of dynamic
capabilities (Figure 1), and it proposes a measurable model to represent the nature
of dynamic capabilities. Second, it empirically supports a structural model in
which dynamic capabilities have an indirect positive effect on performance by
reconfiguring operational capabilities in NPD, an effect that is positively moderated
(reinforced) by environmental turbulence (Figure 2). These two key findings have
implications for (i) conceptualizing, operationalizing, and measuring dynamic
capabilities, and (ii) understanding the effects of dynamic capabilities in turbulent
environments.
Implications for Conceptualizing, Operationalizing, and Measuring
Dynamic Capabilities
Dynamic capabilities have been viewed as hidden and intangible assets, an elusive
black box. Winter (2003) notably explains: “probably some of the mystery and
confusion surrounding the concept of dynamic capability arises from linking the
concept too tightly to notions of generalized effectiveness at dealing with change
and generic formulas for sustainable competitive advantage. The argument here
is that clarity is served by breaking this link” (p. 994). Such confusion has made
it difficult for managers to understand, measure, and thus act upon dynamic capabilities. Dynamic capabilities have also been criticized for a lack of empirical
grounding and measurement (Williamson, 1999), and attempts to measure dynamic
capabilities have merely used distant proxies (e.g., Henderson & Cockburn, 1994;
Nerkar & Roberts, 2004; Arend & Bromiley, 2009). Integrating the dynamic capabilities, decision sciences, and NPD literatures, this study clarifies the semantics
of dynamic capabilities, sheds light on their inner-workings, and offers a simple
model to conceptualize, operationalize, and measure dynamic capabilities. Although Teece (2007) developed a comprehensive framework to capture numerous
activities in the realm of dynamic capabilities, this study offers a parsimonious
260
The Elusive Black Box of Dynamic Capabilities
model with a limited set of specific, concrete, and measurable dynamic capabilities. The study also identifies the underlying components of each capability,
thus showing that the proposed capabilities closely correspond to the dynamic
capabilities literature. Our measurable model alleviates the criticism that dynamic
capabilities cannot be measured, and they are thus born, not made (Winter, 2003, p.
991), and it has implications for offering an actionable set of dynamic capabilities
that decision-makers can use to manage in turbulence.
Viewing dynamic capabilities as abstract capabilities has created a “doubt
that deliberate efforts to strengthen dynamic capabilities are a genuine option for
managers” (Winter, 2003, p. 991). The proposed model of dynamic capabilities
with specific actionable guidelines for managers suggests that dynamic capabilities are managerially-amenable practices that managers can readily act upon. Teece
et al. (1997, p. 521) argue: “the capacity to reconfigure and transform is itself a
learned organizational skill. The more frequently practiced, the easier accomplished.” This study helps managers understand what dynamic capabilities are,
how they can be measured, thus offering guidance to identify, measure, benchmark, and enhance dynamic capabilities to enhance decisions in turbulent environments. The proposed model also creates a common language between managers
and researchers to allow further empirical research on dynamic capabilities.
Besides dynamic capabilities, agility is another means proposed in the literature to manage in turbulent environments (Sambamurthy et al., 2003). Although
agility stresses the ability to sense and respond, the proposed model of dynamic
capabilities offers a more complete picture of how managers can address turbulent environments. Having sensed the opportunities in the environment, the
“respond” component of agility is captured with the ability to learn to revamp
existing operational capabilities with new knowledge, integrate the new knowledge in the reconfigured operational capabilities, and coordinate to synchronize
tasks, resources, and activities to deploy the reconfigured operational capabilities.
The proposed model of dynamic capabilities outlines the exact factors needed to
respond to opportunities by reconfiguring existing operational capabilities. It thus
offers more specific and actionable guidelines for manages to decide and act in
turbulence.
Although dynamic capabilities have been viewed as firm-level capabilities
relevant only to decision-makers at the top executive level, this study shows that
dynamic capabilities are also relevant for lower-level managers that make decisions
about intra- or interfirm business units. This has implications for the dynamic
capabilities view by extending the level of analysis beyond the overall firm level
to the business unit as the level of analysis. This also has implications for the
enhanced scope of dynamic capabilities for virtually all levels of decision making
throughout the firm and extending the scope of dynamic capabilities to business
unit managers.
Implications for Understanding the Effects of Dynamic Capabilities
Although the literature intuitively agrees that dynamic capabilities are valuable
assets, this study empirically validates this intuition. Also, by describing the process
by which dynamic capabilities impact performance by reconfiguring operational
Pavlou and El Sawy
261
capabilities, this study empirically overcomes Williamson’s (1999) tautological
criticism of the dynamic capabilities view. This implies that dynamic capabilities
do have an effect on performance, but they do so indirectly by reconfiguring
operational capabilities into new ones that better fit the environment. The proposed
“indirect” view extends Zott’s (2003) direct, straightforward link between dynamic
capabilities and performance.
By implying that dynamic capabilities are only valuable in turbulent environments, the literature has assumed that dynamic capabilities are worthless or even
harmful in more stable environments (Moorman & Miner, 1998). To overcome
this apparent misconception, this study shows the positive role of dynamic capabilities in the entire spectrum of environmental turbulence. Although we cannot
confirm that this study’s levels of environmental turbulence represents perfectly
stable environments, this finding may be explained by the fact that dynamic capabilities can earn even higher rents by effectively reconfiguring existing operational capabilities. Penrose (1959) viewed resources as pieces of a “jigsaw puzzle”
(p. 70) that can be destroyed and recombined to produce superior outputs given
new opportunities. This finding implies that dynamic capabilities are valuable
in virtually all levels of environmental turbulence, implying that managers must
continuously try to identify new opportunities and make decisions to reconfigure
their existing operational capabilities, irrespective of the level of environmental
turbulence.
Limitations and Suggestions for Future Research
First, assuming that dynamic capabilities correspond to the exploration and operational capabilities to the exploitation mode (March, 1991), firms face a trade-off
between focusing on dynamic versus operational capabilities. Although an emphasis on dynamic capabilities may create disruptions to the efficiency of existing
operational capabilities, a focus on operational capabilities may eventually result
in rigidities. Given the ability to directly measure dynamic capabilities with our
proposed measurable model, future research could identify an optimum solution
given the degree of environmental turbulence. This might help achieve the ambidextrous organization that optimizes the trade-off between dynamic capabilities
and operational capabilities.
Second, the measurable model of dynamic capabilities opens avenues for
empirical research. Having identified a measurable model of dynamic capabilities,
it is possible to examine how dynamic capabilities can be enhanced by identifying
their antecedents. Because dynamic capabilities are information-intensive, IT may
help facilitate dynamic capabilities (e.g., Ettlie & Pavlou, 2006; Pavlou & El
Sawy, 2006). This may help address the question how IT influences competitive
advantage in turbulent environments (Sambamurthy et al., 2003). Also, recent
work proposed various antecedents of dynamic capabilities at the individual, firm,
and network level (Rothaermel & Hess, 2007). It is also possible to integrate
dynamic capabilities with related reconfiguration capabilities. For instance, Winter
(2003) presents a trade-off between dynamic capabilities and improvisation. Future
research could examine the impact of dynamic capabilities and improvisation on
performance in different levels of environmental turbulence.
262
The Elusive Black Box of Dynamic Capabilities
Third, the study’s unit of analysis (business unit) implies that dynamic capabilities can be uniform across the firm, and they may differ in terms of the business
unit’s operational capabilities they seek to reconfigure. For example, firms may
be more effective in reconfiguring their operational capabilities in NPD versus in
manufacturing or in logistics. Some firms may have different emphasis in reconfiguring different types of operational capabilities, or they may have a different
approach for dealing with environmental turbulence for shaping operational capabilities in different business units. Future research could extend the study of
dynamic capabilities to specific business units, areas, and functions in the firm
beyond this study’s NPD context.
Fourth, despite our focus on NPD, because dynamic capabilities do not contain a “certain domain of knowledge or skill, but the ability to learn new domains”
(Danneels, 2002, p. 1112), we have no reason to believe that our model will not
generalize to other levels and units of analysis. However, the study’s generalization
to other contexts, levels, and units of analysis remains to be examined.
Finally, we hope that the proposed simple and measurable model of dynamic
capabilities is a helpful stepping stone that will allow researchers to do further
research by better conceptualizing, operationalizing, and measuring dynamic capabilities, and that it will also help give managers more specific and actionable
guidelines to make high-quality decisions in turbulent environments. [Received:
October 2008. Accepted: May 2010.]
REFERENCES
Abernathy, W. J., & Clark, K. B. (1986). Innovation: Mapping the winds of creative
destruction. Research Policy, 14(1), 3–22.
Agarwal, R., & Selen, W. (2009). Dynamic capability building in service value
networks for achieving service innovation. Decision Sciences, 40(3), 431–
475.
Amit, R., & Schoemaker, P. (1993). Strategic assets and organizational rent. Strategic Management Journal, 14(1), 33–46.
Arend, R. J., & Bromiley, P. (2009). Assessing the dynamic capabilities view:
Spare change, everyone? Strategic Organization, 7(1), 75–90.
Atuahene-Gima, K., & Li, H. (2004). Strategic decision comprehensiveness and
new product development outcomes in new technology ventures. Academy
of Management Journal, 47(4), 583–597.
Brown, S. L., & Eisenhardt, K. M. (1997). The art of continuous change: Linking
complexity theory and time-paced evolution in relentlessly shifting organizations. Administrative Science Quarterly, 42(1), 1–34.
Carlsson, S., & El Sawy, O. A. (2008). Managing the five tensions of IT-enabled
decision support in turbulent and high-velocity environments. Information
Systems & e-Business Management, 6(3), 225–237.
Carte, T. A., & Russell, C. J. (2003). In pursuit of moderation: Nine common errors
and their solutions. MIS Quarterly, 27(3), 479–501.
Pavlou and El Sawy
263
Chow, G. (1960). Tests of equality between sets of coefficients in two linear
regressions. Econometrica, 28(4), 591–605.
Clark, K., & Fujimoto, T. (1991). Product development performance. Cambridge,
MA: Harvard Business School Press.
Cohen, M. A., Eliashberg, J., & Ho, T.-H. (1996). New product development:
The performance and time-to-Market tradeoff. Management Science, 42(2),
173–186.
Cohen, M. A., & Levinthal, D. (1990). Absorptive capacity: A new perspective on
learning and innovation. Administrative Science Quarterly, 35(1), 128–152.
Collis, D. J. (1994). How valuable are organizational capabilities? Strategic Management Journal, 15(1), 143–152.
Cortina, J. M., Chen, G., & Dunlap, W. P. (2001). Testing interaction effects in
LISREL: Examination and illustration of available procedures. Organizational Research Methods, 4(4), 324–360.
Crowston, K. (1997). A coordination theory approach to organizational process
design. Organizational Science, 8(2), 157–175.
Crowston, K., & Kammerer, E. E. (1998). Coordination and collective mind in
software requirements development. IBM Systems Journal, 37(2), 227–245.
D’Adderio, L. (2001). Crafting the virtual prototype: How firms integrate knowledge and capabilities across organizational boundaries. Research Policy,
30(9), 1409–1424.
Danneels, E. (2002). The dynamic of product innovation and firm competences.
Strategic Management Journal, 23(9), 1095–1121.
D’Aveni, R. (1994). Hypercompetition: Managing the dynamics of strategic maneuvering. New York: Free Press.
Day, G. S. (1994). The capabilities of market-driven organization. Journal of
Marketing, 58(1), 37–52.
De Boer, M., Van de Bosch, F., & Volberda, H. W. (1999). Managing organizational knowledge integration in the emerging multimedia complex. Journal
of Management Studies, 36(3), 379–418.
Diamantopoulos, A., & Winklhofer, H. M. (2001). Index construction with formative indicators: An alternative to scale development. Journal of Marketing
Research, 38(2), 269–277.
Dierickx, I., & Cool, K. (1989). Asset stock accumulation and sustainability of
competitive advantage. Management Science, 35, 1504–1511.
Duncan, R. B. (1972). Characteristics of organizational environments and perceived environmental uncertainty. Administrative Science Quarterly, 17(3),
313–323.
Dyer, J. H., & Singh, H. (1998). The relational view: Cooperative strategy and
sources of interorganizational competitive advantage. Academy of Management Review, 20(2), 343–378.
Edwards, R. (2000). Automotive transformations—A challenge for IT. Business
Briefing: Global Automotive and Manufacturing Technology, 108–110.
264
The Elusive Black Box of Dynamic Capabilities
Eisenhardt, K. M., & Brown, S. (1999). Patching: Restitching business portfolios
in dynamic markets. Harvard Business Review, 77(1), 72–82.
Eisenhardt, K. M., & Galunic, D. C. (2000). Coevolving: At last, a way to make
synergies work. Harvard Business Review, 78, 91–101.
Eisenhardt, K. M., & Martin, J. (2000). Dynamic capabilities: What are they?
Strategic Management Journal, 21(8), 1105–1121.
Ettlie, J., & Pavlou, P. A. (2006). Technology-based new product development
partnerships. Decision Sciences, 37(2), 117–148.
Francis, J., & Schipper, K. (1999). Have financial statements lost their relevance?
Journal of Accounting Research, 37(2), 319–352.
Fredrickson, J. W., & Mitchell, T. R. (1984). Strategic decision processes: Comprehensive and performance within an industry with an unstable environment.
Academy of Management Journal, 27(2), 399–423.
Galunic, D. C., & Eisenhardt, K. M. (2001). Architectural innovation and modular
corporate forms. Academy of Management Journal, 44(6), 1229–1249.
Galunic, D. C., & Rodan, S. (1998). Resource recombinations in the firm: Knowledge structures and the potential for Schumpeterian innovation. Strategic
Management Journal, 19(10), 1193–1201.
Grant, R. (1996). Toward a knowledge based theory of the firm. Strategic Management Journal, 17(1), 109–122.
Grewal, R., & Slotegraaf, R. (2007). Embeddedness of organizational capabilities.
Decision Sciences, 38(3), 451–488.
Griffin, A. (1997). The effect of project and process characteristics on product
development cycle time. Journal of Marketing Research, 34(1), 24–35.
Hair, J. F., Anderson, R. E., Tatham, R. L., & Black, W. C. (1995). Multivariate
data analysis. New York: Prentice Hall.
Helfat, C., & Peteraf, M. (2003). The dynamic resource-based view: Capability
lifecycles. Strategic Management Journal, 24(10), 997–1010.
Henderson, R. (1994). The evolution of integrative capability: Innovation in cardiovascular drug discovery. Industrial and Corporate Change, 3(3), 607–
630.
Henderson, J. C., & Clark, H. (1990). Architectural innovation: The reconfiguration of existing product technologies and the failure of established firms.
Administrative Science Quarterly, 35(1), 9–30.
Henderson, R., & Cockburn, I. (1994). Measuring competence? Exploring firm
effects in pharmaceutical research. Strategic Management Journal, 42(1),
63–84.
Hurley, R. F., & Hult, T. M. (1998). Innovation, market orientation, and organizational learning: An integration and empirical examination. Journal of
Marketing, 62(3), 42–54.
Iansiti, M., & Clark, K. (1994). Integration and dynamic capability: Evidence from
product development in automobiles and mainframe computers. Industrial
and Corporate Change, 3(3), 557–605.
Pavlou and El Sawy
265
Itami, M. (1987). Mobilizing invisible assets. Cambridge, MA: Harvard University
Press.
Jap, S. D. (2001). Perspectives on joint competitive advantage in buyer-supplier
relationships. International Journal of Research in Marketing, 18(1), 19–35.
Jaworski, B. J., & Kohli, A. (1993). Market orientation: Antecedents and consequences. Journal of Marketing, 57(1), 53–70.
Joreskog, K. G., & Goldberger, A. S. (1975). Estimation of a model with multiple
indicators and multiple causes of a single latent variable. Journal of the
American Statistical Association, 10(4), 473–496.
Kenny, D., & Judd, C. (1984). Estimating the non-linear and interactive effects of
latent variables. Psychological Bulletin, 96, 201–210.
Kogut, B., & Zander, U. (1992). Knowledge of the firm, combinative capabilities, and the replication of technology. Organizational Science, 3(3), 383–
397.
Kogut, B., & Zander, U. (1996). What firms do? Coordination, identity, and learning. Organizational Science, 7(5), 502–518.
Kusunoki, K., Nonaka, I., & Nagata, A. (1998). Organizational capabilities in product development of Japanese firms: A conceptual framework and empirical
findings. Organizational Science, 9(6), 699–718.
Lane, P., & Lubatkin, M. (1998). Relative absorptive capacity and interorganizational learning. Strategic Management Journal, 19, 461–477.
Leonard-Barton, D. (1992). Core capabilities and core rigidities: A paradox in
managing new product development. Strategic Management Journal, 13(7),
111–125.
Lippman, S. A., & Rumelt. (1982). Uncertain imitability: An analysis of interfirm
differences in efficiency under competition. Bell Journal of Economics, 13,
418–438.
Makadok, R. (2001). Toward a synthesis of the resource-based and dynamiccapability views of rent creation. Strategic Management Journal, 22(5), 487–
505.
March, J. G. (1991). Exploration and exploitation in organizational learning. Organizational Science, 2(1), 71–87.
Marsh, H. W., & Hocevar, D. (1985). Applications of confirmatory factor analysis to the study of self concept: First and higher order factor models and their invariance across groups. Psychological Bulletin, 97, 562–
582.
Moorman, C., & Miner, A. S. (1998). The convergence of planning and execution:
Improvisation in new product development. Journal of Marketing, 62(3),
1–20.
Moorman, C., & Slotegraaf, R. J. (1999). The contingency value of complementary
capabilities in product development. Journal of Marketing Research, 36(2),
239–257.
266
The Elusive Black Box of Dynamic Capabilities
Nelson, R. R., & Winter, S. G. (1982). An evolutionary theory of economic change.
Cambridge, MA: Bellknap.
Nerkar, A., & Roberts, P. W. (2004). Technological and product-market experience
and the success of new product introductions in the pharmaceutical industry.
Strategic Management Journal, 25(7/8), 779–799.
Okhuysen, G. A., & Eisenhardt, K. M. (2002). Integrating knowledge in groups:
How formal interventions enable flexibility. Organizational Science, 13(4),
370–386.
Pavlou, P. A., & El Sawy, O. A. (2006). From IT competence to competitive
advantage in turbulent environments: The case of new product development.
Information Systems Research, 17(3), 198–227.
Penrose, E. (1959). The theory of the growth of the firm. London: Basil Blackwell.
Ping, R. A. (1995). A parsimonious estimating technique for interaction and
quadratic latent variables. Journal of Marketing Research, 32(3), 336–
347.
Pisano, G. (1994). Knowledge, integration, and the locus of learning: An empirical
analysis of process development. Strategic Management Journal, 15(1), 85–
100.
Podsakoff, P. M., & Organ, D. W. (1986). Self-reports in organizational research:
Problems and prospects. Journal of Management, 12(4), 531–544.
Priem, R., & Butler, J. (2001). Is the resource-based “view” a useful perspective
for strategic management research? Academy of Management Review, 26(1),
22–40.
Quinn, R. W., & Dutton, J. E. (2005). Coordination as energy-in conversation.
Academy of Management Review, 30(1), 36–57.
Rindova, P. V., & Kotha, S. (2001). Continuous “morphing”: Competing through
dynamic capabilities, form, and function. Academy of Management Journal,
44(5), 1263–1280.
Rothaermel, F. T., & Hess, A. M. (2007). Building dynamic capabilities: Innovation driven by individual-, firm-, and network-level effects. Organizational
Science, 18(6), 898–921.
Sambamurthy, V., Bharadwaj, A. S., & Grover, V. (2003). Shaping agility through
digital options: Reconceptualizing the role of IT in contemporary firms. MIS
Quarterly, 27(2), 237–263.
Schumpeter, J. A. (1934). The theory of economic development. Cambridge, MA:
Harvard University Press.
Sethi, R. (2000). New product quality and product development teams. Journal of
Marketing, 64(April), 1–14.
Sethi, R., Smith, D. C., & Park, W. (2001). Cross-functional product development
teams, creativity, and the innovativeness of new consumer products. Journal
of Marketing Research, 38(1), 73–85.
Simonin, B. (1999). Ambiguity and the process of knowledge transfer in strategic
alliances. Strategic Management Journal, 20(7), 595–623.
Pavlou and El Sawy
267
Sinkula, J. (1994). Market information processing and organizational learning.
Journal of Marketing, 58(1), 35–45.
Song, M. X., Droge, C., Hanvanich, S., & Calantone, R. (2005). Marketing and
technology resource complementarity: An analysis of their interaction effect
in two environmental contexts. Strategic Management Journal, 26(4), 259–
276.
Song, M. X., & Parry, M. E. (1997). A cross-national comparative study of new
product development processes: Japan and the United States. Journal of
Marketing, 61(1), 1–18.
Sull, D. (2009). How to thrive in turbulent markets. Harvard Business Review,
77(1), 78–88.
Teece, D. J. (1982). Towards an economic theory of the multi-product firm. Journal
of Economic Behavior and Organization, 3(1), 39–63.
Teece, D. J. (2007). Explicating dynamic capabilities: The nature and microfoundations of (sustainable) enterprise performance. Strategic Management
Journal, 28(8), 1319–1350.
Teece, D., Pisano, G., & Shuen, A. (1997). Dynamic capabilities and strategic
management. Strategic Management Journal, 18(7), 509–533.
Van den Bosch, F., Volberda, H., & De Boer, M. (1999). Co-evolution of firm
absorptive capacity and knowledge environment: Organizational forms and
combinative capabilities. Organizational Science, 10(4), 551–568.
Vorhies, D. W., & Harker, M. (2000). The capabilities and performance advantages
of market-driven firms: An empirical investigation. Australian Journal of
Management, 25(2), 145–171.
Weick, K. E., & Roberts, K. H. (1993). Collective mind in organizations: Heedful
interrelating on flight decks. Administrative Science Quarterly, 38(3), 357–
381.
Williamson, O. (1999). Strategy research: Governance and competence perspectives. Strategic Management Journal, 20(12), 1087–1110.
Winter, S. (2003). Understanding dynamic capabilities. Strategic Management
Journal, 24(10), 991–995.
Zahra, S. A., & George, G. (2002). Absorptive capacity: A review, reconceptualization, and extension. Academy of Management Review, 27(2), 185–203.
Zahra, S. A., Sapienza, H. J., & Davidsson, P. (2006). Entrepreneurship and dynamic capabilities: A review, model, and research agenda. Journal of Management Studies, 43(4), 917–955.
Zammuto, R. F. (1988). Organizational adaptation: Some implications of organizational ecology for strategic choice. Journal of Management Studies, 25(2),
105–120.
Zander, U., & Kogut, B. (1995). Knowledge and the speed of transfer and imitation
of organizational capabilities. Organizational Science, 6(1), 76–92.
Zollo, M., & Winter, S. (2002). Deliberate learning and the evolution of dynamic
capabilities. Organizational Science, 13(2), 339–351.
268
The Elusive Black Box of Dynamic Capabilities
Zott, C. (2003). Dynamic capabilities and the emergence of intra-industry differential firm performance: Insights from a simulation study. Strategic Management Journal, 24(1), 97–125.
APPENDIX A: MEASUREMENT ITEMS
Dynamic capabilities
Instructions: “Please rate the effectiveness by which your work unit reconfigures its
operational capabilities in the NPD process to address rapidly-changing
environments relative to your major competitors.”
Sensing capability
We frequently scan the environment to identify new business opportunities.
We periodically review the likely effect of changes in our business environment on
customers.
We often review our product development efforts to ensure they are in line with what the
customers want.
We devote a lot of time implementing ideas for new products and improving our existing
products.
Learning capability
We have effective routines to identify, value, and import new information and knowledge.
We have adequate routines to assimilate new information and knowledge.
We are effective in transforming existing information into new knowledge.
We are effective in utilizing knowledge into new products.
We are effective in developing new knowledge that has the potential to influence product
development.
Integrating capability
We are forthcoming in contributing our individual input to the group.
We have a global understanding of each other’s tasks and responsibilities.
We are fully aware who in the group has specialized skills and knowledge relevant to our
work.
We carefully interrelate our actions to each other to meet changing conditions.
Group members manage to successfully interconnect their activities.
Coordinating capability
We ensure that the output of our work is synchronized with the work of others.
We ensure an appropriate allocation of resources (e.g., information, time, reports) within
our group.
Group members are assigned to tasks commensurate with their task-relevant knowledge
and skills.
We ensure that there is compatibility between group members expertise and work
processes.
Overall, our group is well coordinated.
Pavlou and El Sawy
269
Reconfiguration capability (indicator items for second-order constructs)
We can successfully reconfigure our resources to come up with new productive assets.
We often engage in resource recombinations to better match our product-market areas and
our assets.
NPD performance
Please rate the performance of your NPD work unit relative to your major
competitors in the following aspects:
2a. Process Efficiency
Overall development costs
2b. Product Effectiveness
Improvements in product
quality and functionality
2c. Perceived Competitive
Advantage
We have gained strategic
advantages in the
marketplace over our
competitors
Overall efficiencies of NPD
process
Accelerated time-to-market
Major innovations in
products as a whole
Creation of new product
We have gained a
concepts
competitive advantage
Operational capabilities in NPD
3a. Technical capability
Evaluating the technical feasibility of developing new products with continuously
changing features.
Recurrently evaluating tests to determine basic performance against
shifting technical specifications.
Frequently executing prototypes or sample product testing.
3b. Customer capability
Frequently determining market characteristics and trends.
Regularly appraising competitors and their products—both existing and potential.
Executing several test-marketing programs in line with commercialization plans.
3c. Managerial capability
Management effectively monitors the progress of this NPD group.
Management is actively involved in activities at the working level.
Management effectively administers relevant tasks and functions.
Overall NPD capability (indicator items)
We do a remarkable job of developing new products.
Our product development group gives us an edge in the market.
Environmental turbulence
The technology in this product area is changing rapidly. [Technological turbulence]
Technological breakthroughs provide big opportunities in this product area.
[Technological turbulence]
In our kind of business, customers’ product preferences change a lot over time. [Market
turbulence]
Marketing practices in our product area are constantly changing. [Market turbulence]
New product introductions are very frequent in this market. [Market turbulence]
The environment in our product area is continuously changing. [Indicator item]
Environmental changes in our industry are very difficult to forecast. [Indicator item]
270
The Elusive Black Box of Dynamic Capabilities
Control variables
Cross-functional integration
There are frequent interactions between our cross-functional NPD group.
The NPD process is truly a cross functional effort.
Group NPD experience
How long has this NPD group been in place? _________ years.
NPD innovation type [Select one option]
Basic research that lays the basic foundations for future product development effort.
[Basic research]
Applied work to develop specific, clearly defined products to fulfill immediate goals and
strategic directions. [Applied NPD]
Routine engineering for continuous improvement of existing products and processes.
[Routine engineering]
Intra- Vs. interorganizational NPD business units
Our NPD partner is a (please circle one): (a) supplier, (b) customer, (c) internal unit,
(d) other (please specify): _____
Estimated total
Number of
% of sales spent on
Industry
annual revenues
employees
research and
segment/product
development
domain
Sensing capability 1
Sensing capability 2
Sensing capability 3
Sensing capability 4
Learning capability 1
Learning capability 2
Learning capability 3
Learning capability 4
Learning capability 5
Integrating capability 1
Integrating capability 2
Integrating capability 3
Integrating capability 4
Integrating capability 5
Coordinating capability 1
Coordinating capability 2
Coordinating capability 3
Coordinating capability 4
.781
.747
.723
.750
.211
.190
.161
.142
.115
.105
.132
.147
.203
.233
.246
.265
.253
.134
.291
.231
.236
.157
.686
.738
.741
.805
.774
.174
.192
.142
.151
.138
.113
.140
.213
.210
.157
.118
.134
.231
.188
.204
.200
.231
.250
.647
.656
.715
.721
.698
.134
.201
.167
.204
.167
.146
.130
.163
.153
.112
.205
.274
.204
.198
.105
.170
.179
.125
.791
.768
.753
.754
.230
.203
.129
.227
.125
.205
.069
.049
.092
.187
.108
.036
.115
.110
.290
.095
.172
.100
.179
.176
.113
.189
.184
.145
.126
.164
.181
.215
.218
.133
.145
.104
.205
.043
.055
.177
Sensing Learning Integrating Coordinating Quality Efficiency
APPENDIX B: PRINCIPAL COMPONENTS FACTOR ANALYSIS
.132
.122
.047
.005
.018
.034
.055
.119
.122
.088
.176
.173
.139
.212
.208
.220
.107
.153
.058
.099
.109
.045
.079
.064
.042
.141
.187
.083
.136
.118
.106
.189
.223
.206
.152
.105
.052
.029
.074
.126
.063
.105
.091
.023
.093
.130
.154
.118
.125
.130
.160
.212
.221
.231
168
.138
.085
.145
.200
.202
.104
.053
.110
.165
.190
.095
.182
.164
.036
.088
.032
.156
Continued
.083
.087
.069
.089
.137
.114
.094
.107
.121
.354
.043
.052
.124
.122
.134
.144
.217
.242
Market Tech
Turb Turb Customer Technical Managerial
Pavlou and El Sawy
271
Performance 1 (quality)
Performance 2 (quality)
Performance 3 (quality)
Performance 1 (efficiency)
Performance 2 (efficiency)
Performance 3 (efficiency)
Market turbulence 1
Market turbulence 2
Market turbulence 3
Technological turbulence 1
Technological turbulence 2
Customer capability 1
Customer capability 2
Customer capability 3
Technical capability 1
Technical capability 2
Technical capability 3
Managerial capability 1
Managerial capability 2
Managerial capability 3
Variance explained (79%)
.187
.187
.121
.202
.135
.132
.123
.176
.165
.121
.135
.157
.206
.215
.128
.129
.123
.119
.125
.127
11.1
.213
.210
.163
.159
.233
.224
.145
.125
.144
.151
.156
.103
.161
.200
.184
.204
.205
.199
.225
.221
6.9
.223
.215
.165
.091
.085
.136
.241
.235
.220
.221
.165
.253
.144
.151
.184
.147
.122
.156
.184
.182
8.4
.225
.125
.174
.179
.153
.188
.187
.160
.138
.231
.301
.267
.285
.267
.238
.185
.145
.134
.124
.137
5.0
.825
.856
.731
.227
.223
.046
.079
.153
.085
.134
.122
.093
.112
.105
.144
.142
.185
.244
.204
.202
7.2
.255
.274
.204
.792
.828
.749
.242
.231
.212
.135
.135
.181
.147
.123
.209
.201
.103
.116
.127
.145
8.5
Sensing Learning Integrating Coordinating Quality Efficiency
APPENDIX B: (Continued)
.223
.214
.121
.106
.081
.073
.753
.699
.741
.142
.106
.111
.153
.137
.131
.195
.118
.211
.215
.223
5.5
.183
.173
.177
.123
.105
.104
.112
.119
.158
.775
.733
.103
.126
.140
.110
.105
.114
.146
.088
.135
5.2
.076
.153
.094
.112
.195
.144
.304
.298
.083
.085
.068
.830
.763
.772
.122
.096
.132
.111
.043
.100
8.8
.069
.049
.116
.143
.241
.216
.087
.088
.077
.183
.255
.215
.122
155
.752
.819
.861
.191
.167
.206
4.7
.326
.364
.281
.101
.153
.190
.215
.169
.235
.169
.218
.219
.163
.175
.201
.210
.241
.821
.726
.744
7.7
Market Tech
Turb Turb Customer Technical Managerial
272
The Elusive Black Box of Dynamic Capabilities
Pavlou and El Sawy
273
Paul A. Pavlou is an associate professor of management information systems,
marketing, and management and a Stauffer senior research fellow at the Fox School
of Business at Temple University. He received his PhD from the University of
Southern California in 2004. His research focuses on e-commerce, online auctions,
IT strategy, information economics, research methods, and NeuroIS. His research
has appeared in MISQ, ISR, JMIS, JAIS, JAMS, and CACM, among others. His
work has been cited over 1,100 times by the Social Science Citation Index of the
Institute of Scientific Information and over 3,000 times by Google Scholar. He
won several best paper awards for his research, including the “ISR Best Paper”
award in 2007, the 2006 “IS Publication of the Year” award, the “Top 5 Papers”
award in Decision Sciences in 2006, among others. He also won several Reviewer
awards, including the 2009 Management Science Meritorious service award, the
“Best Reviewer” award of the 2005 Academy of Management Conference, and the
2003 MIS Quarterly “Reviewer of the Year” award.
Omar A. El Sawy is a professor of information systems at the Marshall School of
Business at the University of Southern California (USC). His interests are around
IT-enabled business strategy in turbulent environments and designing business
models for services offered through digital platforms. He served as Director of
Research for USC’s Institute for Communication Technologies Management from
2001 to 2007, where he led industry-sponsored research programs. He holds a
PhD from Stanford Business School, an MBA from the American University in
Cairo, and a BSEE from Cairo University. El Sawy is the author or coauthor of
over 100 papers, serves on several journal editorial boards, and is a six-time winner
of the Society for Information Management’s Paper Awards Competition. One of
his previous papers in Information Systems Research with Paul Pavlou won the
2007 best published paper award. He is a Fellow of the Association of Information
Systems.