ArticlePDF Available

Abstract and Figures

textlessptextgreaterThis paper builds on the principles and insights from improvisational theater to unpack the nature of collective improvisation and to consider what it takes to do it well and to innovate. Furthermore, we discuss the role of training in enhancing the incidence and effectiveness of improvisation. We propose that two common misconceptions about improvisation have hindered managers' understanding of how to develop the improvisational skill. First, the spontaneous facet of improvisation tends to be overemphasized, and second, there is a general assumption that improvisation always leads to positive performance. Our goal is to clear up the conceptual confusion about improvisation by laying out the various aspects of preparation that are required for effective improvisation. In our theoretical model, we delineate how the improvisational theater principles of "practice," "collaboration," "agree, accept, and add," "be present in the moment," and "draw on reincorporation and ready-mades" can be used to understand what it takes to improvise well in work teams and to create a context favoring these efforts. Our findings support a contingent view of the impact of improvisation on innovative performance. Improvisation is not inherently good or bad; however, improvisation has a positive effect on team innovation when combined with team and contextual moderating factors. We also provide initial evidence suggesting that the improvisational skill can be learned by organizational members through training. Our results shed light on the opportunities provided by training in improvisation and on the challenges of creating behavioral change going beyond the individual to the team and, ultimately, to the organization.
Content may be subject to copyright.
Organization Science
Vol. 16, No. 3, May–June 2005, pp. 203–224
issn 1047-7039 eissn 1526-5455 05 1603 0203
doi 10.1287/orsc.1050.0126
© 2005 INFORMS
Improvisation and Innovative Performance in Teams
Dusya Vera
Department of Management, C.T. Bauer College of Business, University of Houston, 334 Melcher Hall,
Houston, Texas 77204,
Mary Crossan
Strategic Management Group, Richard Ivey School of Business, University of Western Ontario,
London, Ontario, Canada N6A 3K7,
This paper builds on the principles and insights from improvisational theater to unpack the nature of collective impro-
visation and to consider what it takes to do it well and to innovate. Furthermore, we discuss the role of training in
enhancing the incidence and effectiveness of improvisation. We propose that two common misconceptions about impro-
visation have hindered managers’ understanding of how to develop the improvisational skill. First, the spontaneous facet
of improvisation tends to be overemphasized, and second, there is a general assumption that improvisation always leads
to positive performance. Our goal is to clear up the conceptual confusion about improvisation by laying out the various
aspects of preparation that are required for effective improvisation. In our theoretical model, we delineate how the impro-
visational theater principles of “practice,” “collaboration,” “agree, accept, and add,” “be present in the moment,” and “draw
on reincorporation and ready-mades” can be used to understand what it takes to improvise well in work teams and to
create a context favoring these efforts. Our findings support a contingent view of the impact of improvisation on innovative
performance. Improvisation is not inherently good or bad; however, improvisation has a positive effect on team innovation
when combined with team and contextual moderating factors. We also provide initial evidence suggesting that the impro-
visational skill can be learned by organizational members through training. Our results shed light on the opportunities
provided by training in improvisation and on the challenges of creating behavioral change going beyond the individual to
the team and, ultimately, to the organization.
Key words: improvisation; creativity; spontaneity; performance; innovation; strategy; improvisational theater; teams
The ability to innovate is critical for organizational
survival (Amabile 1988). As firms strive for faster
cycle times and more innovative solutions, the sponta-
neous and creative facets of improvisation have been
proposed as a pathway to understand and begin act-
ing on what it takes to innovate (Crossan 1997a).
In fact, the role of improvisation in innovation processes
such as new product development has attracted growing
attention (e.g., Eisenhardt and Tabrizi 1995, Moorman
and Miner 1998b, Kamoche and Cunha 2001). Brown
and Eisenhardt (1998, p. 33) argue that improvisation
“enables managers to continuously and creatively adjust
to change and to consistently move products and services
out the door,” and Poolton and Ismail (2000) identify
improvisation as a key area of new development in the
innovation field.
In an effort to understand how individuals work
together in teams to innovate and adapt in real time, aca-
demics have turned to improvisational jazz and theater
(e.g., Crossan 1998, Hatch 1998) and asked: If musi-
cians and actors can learn to improvise and to be inno-
vative in real time, can these skills also be learned by
work teams in organizations? Despite the considerable
attention given to the need for teams to be more nim-
ble and to develop an improvisational capability, little
is known about how team members can learn this skill
and successfully apply it in organizations. Furthermore,
for training in improvisation to be successful, firms need
to create a safe context for improvisation to not only
happen, but to be effective (Crossan and Sorrenti 1997).
Training interventions have been designed for business
organizations based on exercises used by actors in the
world of improvisational theater (e.g., Crossan 1997b),
but limited theoretical work is available on what it takes
to develop this skill. Also, there is a lack of empirical
evidence supporting the success of any improvisational
training effort.
We believe that two common misconceptions about
improvisation have hindered managers’ understanding of
how to develop the improvisational skill in work teams.
First, the spontaneous facet of improvisation tends to
be overemphasized in the extant literature. When impro-
visation is restricted to the ability to “think on your
feet,” managers risk confusing improvisation with ran-
dom moments of brilliance and conclude that either you
have this ability or you do not. There is, however, much
preparation and study behind effective improvisation
(Weick 1998). Improvisation relies on rules and routines
that are preestablished and rehearsed. In improvisation,
it is possible to “prepare to be spontaneous” (Barrett
1998, p. 606) and to “rehearse spontaneity” (Mirvis
1998, p. 587). Second, there is a general assumption in
Vera and Crossan: Improvisation and Innovative Performance in Teams
204 Organization Science 16(3), pp. 203–224, © 2005 INFORMS
much of the literature that improvisation always leads to
positive outcomes and better performance. This assump-
tion is reinforced when improvisation is defined as “to
cope or ingeniously adapt to a set of circumstances”
(Preston 1991, p. 88) or as “devising resourceful solu-
tions to intractable problems” (Meyer 1998, p. 572).
When improvisation is positioned as a solution to all
organizational problems, managers risk underestimating
the need to create a context that supports improvisational
processes in work teams. Improvisation is not inherently
good or bad (Vera and Crossan 2004). Depending on the
skill of the improvisers, improvisation may be highly
innovative or chaotic; improvisation may solve a prob-
lem or worsen it. Efforts to train teams to improvise
need to be based on a realistic understanding of what
improvisation is, and what it is not.
The objective of this paper is to clear up the con-
ceptual confusion about improvisation by laying out in
detail the various aspects of preparation that are entailed
in effective improvisation. We seek to advance under-
standing of the performance implications of improvi-
sation by unpacking what collective improvisation is
(descriptive view) and by examining team and contextual
factors that help improvisational processes to be effec-
tive and positively impact innovation (prescriptive view).
Furthermore, we discuss and test the role of training in
enhancing the effectiveness of improvisation.
This paper focuses on collective improvisation occur-
ring in firms, that is, improvisation by work teams.
Research on the ability to learn the improvisational skill
will not only contribute to the body of knowledge on
improvisation, but also inform efforts to enhance training
for team innovation. Recent reviews highlight that most
innovation research has focused on individual employee
innovation, noting the need to understand how teams
within organizations can facilitate or inhibit innovation
(e.g., Drach-Zahavy and Somech 2001, Caldwell and
O’Reilly 2003, Shalley et al. 2004). West (1994) also
stresses the importance of training in innovation and
states that it is not sufficient to put a team together and
expect it to function effectively and to innovate. Under-
standing what team improvisation entails will shed light
on team innovation processes, in which the initiation and
implementation phases overlap in time (West 1990). We
begin by defining what collective improvisation is, and
what it takes to do it well and to innovate. In doing so,
we discuss the training effect on improvisation perfor-
mance. We then describe our methods, site, and results.
Finally, we summarize the contributions of this research.
The Skill and Context of Improvisation
Improvisation theory is largely based on insights ob-
tained from jazz and theater improvisation (e.g., Orga-
nization Science’s Special Issue on Jazz Improvisation,
1998). There have been detailed expositions about the
similarities and differences between improvisation in
jazz, theater, and other metaphors such as Indian music
(see Hatch 1997, Kamoche et al. 2003). Although we
build on theory arising from jazz improvisation, we rely
more heavily on insights from improvisational theater
for its value-added benefits of accessibility, transferabil-
ity, and universality. The theater metaphor is transparent
and accessible because the elements from which actors
improvise are the same ones used by teams in their
daily work. To understand jazz, we need specialized
musical knowledge (e.g., concepts such as head, chords,
melody, tones, and tempo), whereas theater improvisa-
tion is based on speech, gestures, and movement, which
are the materials of everyday interaction (Lawrence
2001). Lessons from theater improvisation are transfer-
able because the raw materials used in theater improvi-
sation are words, posture, facial expressions, and tone
of voice (in contrast to the musical notes of jazz impro-
visation). Therefore, anyone possesses a certain capabil-
ity to experience and learn from theater improvisation.
Furthermore, while jazz is rooted in specific cultural
traditions, dramatic expression is a universal and time-
less phenomenon. The form of theater may vary across
time and culture, but theater always interprets real life.
Because the focus of this study is on the ability to learn
the improvisation skill, we take advantage of the acces-
sibility and transferability that theater provides to shed
new light on the factors that support improvisational
Prior management research has examined the occur-
rence of improvisation at different levels of analysis. For
example, Weick (1993) describes the individual actions
of a firefighter improvising to save his life in the Mann
Gulch disaster. Teams of individuals also improvise as
shown by Hutchins’s (1991) description of how the crew
of a ship whose navigational system had broken devel-
oped new routines in real time and made their way to
harbor. In the process, no crew member understood the
complete system they improvised, but their collective
actions enabled them to achieve their goal. Evidence of
team improvisation in the context of the arts, sports, and
organizational life has led researchers to conclude that
although collective improvisation builds on individual
improvisation, team improvisation is more than the sum
of individual improvisations because the joint activities
of individuals create a collective system of improvisa-
tional action (Hatch 1997, Moorman and Miner 1998a,
Weick 1998). While team improvisation is clearly a func-
tion of the improvisational capability of individuals in
the team, team improvisation is also influenced by factors
such as team characteristics (e.g., cohesiveness), team
dynamics (e.g., communication), and contextual influ-
ences stemming from the team and the organization (e.g.,
culture). Some theorists have characterized organization-
level improvisation as large-scale team improvisation
(Moorman and Miner 1998a, Cunha et al. 1999), while
Vera and Crossan: Improvisation and Innovative Performance in Teams
Organization Science 16(3), pp. 203–224, © 2005 INFORMS 205
others build on the view of organizations as cultural
entities (Cook and Yanow 1993) and relate organization
improvisation to culture and strategy (Crossan 1997a).
Our focus is on team-level improvisation as it presents
qualities that are distinct from individual improvisation.
However, the team level provides an essential building
block for organizational improvisation.
Finally, there are different degrees of improvisation
(Weick 1998, Zack 2000). Although much has been
written about improvisation in crisis situations, where
time is an obvious scarce resource and spontaneity is
at a premium, improvisation has also been associated
with everyday situations of discovery (Crossan et al.
2005). For example, Orlikowski (1996) describes how
the everyday improvisations and slippages of customer
support specialists adopting a new technology facilitated
the slow transformation of their organizational practices.
Teams improvise to incremental degrees when they make
an adjustment to a standard operating procedure, while
radical cases of improvisation have often been associ-
ated with crisis events. We examine improvisation as
it occurs across the full spectrum of the continuum
and do not differentiate between incremental and radical
Collective Improvisation as “Making Do” and
“Letting Go”
In the context of theater, Halpern et al. (1994, pp. 13–14)
argue that “true improvisation is getting on-stage and
performing without any preparation or planning,” and
that “improvisation is making it up as you go along.
Frost and Yarrow (1990, p. 2) describe the essence of
improvisation in drama:
Improvisation may be close to pure “creativity”—or per-
haps more accurately to creative organization, the way
in which we respond to and give shape to our world.
The process is the same whenever we make a new
arrangement of the information we have, and produce a
recipe, a theory, or a poem. The difference with doing it
à l’improviste,orall’improvviso, is that the attention is
focused on the precise moment when things take shape.
Seham (2001) effectively captures the many defi-
nitions of theater improvisation when she states that
improvisation is a mixture of “making do” and “letting
go.” In applying theater or jazz improvisation to orga-
nizations, theorists have suggested a variety of defini-
tions and dimensions (e.g., Preston 1991, Crossan and
Sorrenti 1997, Hatch 1997, Moorman and Miner 1998a,
Weick 1998, Cunha et al. 1999). For example, improvi-
sation has been defined as “intuition guiding action in a
spontaneous way” (Crossan and Sorrenti 1997, p. 156),
“the degree to which composition and execution con-
verge in time” (Moorman and Miner 1998a, p. 698),
and “the conception of action as it unfolds drawing
on available material, cognitive, affective, and social
resources” (Cunha et al. 1999, p. 302). Many definitions
tend to blend prescriptive and descriptive elements,
largely because management theorists have borrowed so
heavily from descriptions of improvisation in the arts
in which “effectiveness” and “quality of performance”
have been embedded in the phenomenon. We extract the
descriptive elements of spontaneity and creativity, defin-
ing improvisation occurring in teams as the creative and
spontaneous process of trying to achieve an objective
in a new way. As a spontaneous process, improvisation
is extemporaneous. In fact, Weick (1998, p. 552) sug-
gests that “to do things spontaneously is to become more
skilled at thinking on your feet.” The spontaneous—
“letting go”—dimension incorporates a time orientation
to the improvisation construct; individuals respond to
situations on the spur of the moment, reacting in the
moment rather than anticipating, or composing while
executing (Moorman and Miner 1998b). In addition,
the creative—“making do”—dimension incorporates the
search for novelty and usefulness in improvisational
actions, but acknowledges that a creative process does
not always lead to creative outcomes (Drazin et al. 1999,
Gilson and Shalley 2004). By defining improvisation as
a creative process, the focus is not on the creative out-
come that is novel and useful but on how teams “attempt
to orient themselves to, and take creative action in, sit-
uations or events that are complex, ambiguous, and ill
defined” (Drazin et al. 1999, p. 287). Finally, our defini-
tion also highlights improvisation as a conscious choice
people make rather than as random behavior. The deci-
sion to improvise may be made on the spot or may be an
option considered in advance, as when firms have formal
or informal norms enabling people to depart from rou-
tines at certain times to come up with something new.
Intentional improvisation can be observed, for example,
in a research and development (R&D) team improvising
to prototype a new product on time. In this case, team
members knowingly decide to engage in an extempora-
neous process and try to achieve an objective in a new
way—new, at least, to them.
By focusing on the creative process and not on the cre-
ative outcome, an advantage of our definition of impro-
visation is that it does not make any judgment about
the performance implications of improvisational pro-
cesses. Miner et al. (2001) differentiate improvisation
from creativity and innovation by arguing that creativ-
ity may involve absolutely no improvisation and that
innovation may be created through improvisation, but
also through planning. It is the spontaneity and real-
time nature of the action that differentiates creativity
and innovation from improvisation. Furthermore, inno-
vation is output oriented and defined as “the successful
implementation of creative ideas within an organiza-
tion” (Amabile 1996, p. 1). Given the process orienta-
tion of our improvisation construct, the basic premise
of our theory is that improvisation per se is not asso-
ciated with innovative outcomes. Consequently, we do
Vera and Crossan: Improvisation and Innovative Performance in Teams
206 Organization Science 16(3), pp. 203–224, © 2005 INFORMS
not hypothesize a main effect between improvisation
and innovation. Rather, there are several dimensions that
impact the effectiveness of improvisation and its link to
innovative outcomes.
Team Skills for Effective Improvisation
Experienced actors make improvisation look easy and
natural. However, many hours of practice and a reper-
toire of rules help improvisers to focus on the process
of creation without becoming overwhelmed by the pres-
sure of extemporaneous performance (Lawrence 2001).
In fact, much discipline and practice stand behind a suc-
cessful theater performance. As Halpern et al. (1994,
p. 34) argue, “Anyone can improvise, but like any game,
if the players don’t learn and obey the rules, no one
will play with them.” Improvisation is in essence unpre-
dictable, but that does not mean that it is without con-
siderable human infrastructure; practice, expertise, and
knowledge of the rules of collaboration enable team
members, both in theater and in firms, to influence the
quality of their improvisational processes. The following
sections present the key dimensions of effective impro-
visation that lead to innovation.
Expertise. In improvisational theater, actors know
in advance that when the time of the performance
arrives, they will improvise. Consequently, actors “plan
to improvise” and continuously work on improving
their improvisational ability. Weick (1998) worries that,
because of the emphasis on spontaneity, researchers may
overlook the major investment in practice and study
that precedes a stunning jazz performance. The observer
may be unaware that jazz musicians have many years of
experience learning the instrument, the standards, how
to play together, how to blend a sound, etc. Similarly,
improvisational actors learn exercises to develop the fun-
damental skills of listening and communication (Crossan
1998). They also need context-specific knowledge in
diverse areas such as politics, history, and music, so that
they can take on a variety of roles (Vera and Crossan
2004). Players do not know the suggestions that they will
receive from the audience and from fellow actors; thus,
the more expertise they develop in diverse topics, the
more options they will have when accepting a new role.
As in theater, expertise also plays a positive role
in improvisational processes in work teams. Exper-
tise encompasses the specialized skills and knowledge
that individuals bring to the team’s task (Faraj and
Sproull 2000) and is defined as domain-relevant and
task-related skills that depend on innate cognitive abili-
ties, innate perceptual skills, experience, and formal and
informal education (Amabile 1996). When discussing
creative processes, Amabile (1996, p. 95) explains,
“If the domain-relevant skills are already sufficiently
rich to afford an ample set of possible pathways to
explore during task engagement, the reactivation of this
already-stored set of information and algorithms may be
almost instantaneous, occupying little real time.” Exper-
tise has a positive impact on the quality of improvisa-
tional processes because with a larger and more diverse
set of skills in a work team come more alternatives
for developing new combinations of ideas (Vera and
Crossan 2004).
Hypothesis 1. The greater the team’s expertise
(domain- and task-relevant), the more positive the rela-
tionship between collective improvisation and innova-
Teamwork Quality. In addition to practice and exper-
tise, the rule of collaboration between players is fre-
quently taken for granted when describing collective
improvisation. Team improvisation is not just a function
of having the “right” expertise on the team. Expertise
must be coordinated within the team; its interdepen-
dencies must be managed effectively (Faraj and Sproull
2000). The success of improvisational theater perfor-
mances depends on healthy team relationships and
dynamics because scenes evolve from the interdependent
work of the improvisers (Spolin 1963). This is consis-
tent with Frost and Yarrow’s (1990, p. 108) description
of the first cardinal sin of improvisation: “The actors
mustn’t be left stranded.” On the stage, every member
of the team is responsible for the other; actors look
after one another and take the pressure off of each other
rather than increase it (Frost and Yarrow 1990). The
emerging performance is a truly collaborative creation
that cannot be understood by simply analyzing the mem-
bers of the team individually (Sawyer 1999). Teamwork
skills associated with quality improvisation include trust
among players, a common goal, a shared responsibility,
a common vocabulary, and the ability both to lead and
to follow (Crossan 1998). Halpern et al. (1994, p. 92)
summarize these team relationships:
When a team of improvisers pays close attention to each
other, hearing and remembering everything, and respect-
ing all that they hear, a group mind forms. The goal of
this phenomenon is to connect the information created
out of group ideas—and it’s easily capable of brilliance.
The collaboration needed for innovative team impro-
visation is based on both cognitive and affective factors.
On the cognitive side, when improvisers share a col-
lective mind or mental model, this enables better coor-
dination when trying to come up with something new.
Weick and Roberts (1993) conceptualize a collective
mind as a pattern of heedful interrelations of actions in a
social system, and Cannon-Bowers et al. (1993) describe
a team mental model as including shared representa-
tions of tasks, equipment, working relationships, and sit-
uations. Shared mental models provide team members
with a set of organized expectations for team perfor-
mance. Under conditions of high workload or stress,
Vera and Crossan: Improvisation and Innovative Performance in Teams
Organization Science 16(3), pp. 203–224, © 2005 INFORMS 207
highly effective teams are able to adapt to and antic-
ipate other members’ information needs because of a
shared understanding of the situation, the environment,
and team interaction patterns (Cannon-Bowers and Salas
1998). Another concept that helps to explain the coordi-
nation among team members is that of transactive mem-
ory, defined by Wegner (1987) as the set of knowledge
possessed by team members, coupled with an aware-
ness of who knows what. In fact, transactive memory
is an implicit part of the improvisational technique of
rotational leadership, which means that actors let dif-
ferent people take the lead depending on the needs of
the situation. Knowing who has what knowledge or skill
in the team is instrumental when teams face new sit-
uations. As team members develop the ability to work
together smoothly, they face less need for planning,
greater cooperation, fewer misunderstandings, and less
confusion (Liang et al. 1995).
In addition to the cognitive aspects of collaboration,
effective improvisation builds on affective factors such
as trust, respect, and mutual support. Although teams
may improvise in the absence of trust and respect,
improvisation thrives in their presence because team
members know they can take risks and be supported by
others (Crossan 1998). Given the unpredictable nature
of improvisation, trust among team members reinforces
the belief that the collective improvisational process
will achieve its objective. Healthy and close team rela-
tionships are, however, not necessarily easy to develop
because competition, power, and status are often impor-
tant factors affecting team dynamics. Yet, the principle
of collaboration has important implications for cooper-
ation, teamwork quality, balance in member contribu-
tions in work teams, and the effectiveness of collective
Hypothesis 2. The greater the teamwork quality
(e.g., cooperation and trust), the more positive the rela-
tionship between collective improvisation and innova-
The Context for Effective Improvisation
Improvisational processes in teams require a context
that supports their creative and spontaneous nature.
Johnstone (1979, p. 118) refers to his experience in
teaching theatrical improvisation: “If I want people to
free-associate, then I have to create an environment in
which they aren’t going to be punished, or in any way
held responsible for the things their imagination gives
them.” To ensure this environment, actors rely on the
principle of “agree, accept, and add.” Improvisers also
learn to read the cues from their environment and to
“make do” with whatever they have at hand (Weick
1993). This is captured in two rules: “be present in
the moment,” and “draw on reincorporation and ready-
mades” (Johnstone 1979). In the next sections we build
on these theatrical principles to delineate the team con-
text and resources supporting an improvisational capa-
bility that leads to innovation.
Experimental Culture: “Agree, Accept, and Add.
Halpern et al. (1994, p. 35) state, “Anything can happen
in improv. The only rule that can never be broken is
the rule of agreement.” This rule is captured in the pop-
ular technique of “yes-anding.” To yes-and means that
actors accept the offer made to them and build on it.
This is consistent with Frost and Yarrow’s (1990, p. 110)
description of the second cardinal sin of improvisation:
“Blocking is a denial of the possibility of encounter.
The rule of agreement creates a context in which impro-
visers are required to accept, support, and enhance the
ideas expressed by other actors on stage without deny-
ing a player’s reality (Seham 2001). Blocking the ideas
of others is considered a form of aggression (Johnstone
1979) because answering “yes, but” or “no” erases any
scene being created (Halpern et al. 1994). Because of
the principle of agreement, actors know that their con-
text supports experimentation, that their actions are not
being judged by fellow players, and that nothing is seen
as a mistake. In this context, they can stretch a little
further than they have before (Crossan 1998).
Firms interested in promoting innovation need to
incorporate the rule of agreement as a norm of their
organizational and team cultures. Team cultural norms
are not isolated from the organizational culture, but
are also not dominated by it. Although some organiza-
tional culture theory claims that culture is created by
the firm’s founder and its top executives (e.g., Schein
1992), authors such as Kunda (1992) have found that
lower-level employees are not without agency and often
question the authenticity of the beliefs and emotions
promoted by the engineers of culture. The result is the
creation of subcultures and countercultures (e.g., Martin
and Siehl 1983). In addition, Gregory (1983) proposes
that organizations are most accurately viewed as “multi-
cultural,” not only including subcultures such as depart-
mental or project cultures, but also occupational cultures
that cross-cut several firms.
The norm of agreement is critical for the creation of
an experimental culture in teams, which is defined as
a culture that provides room for experimentation and
is tolerant of “competent” mistakes (Vera and Crossan
2004). Low levels of experimentation and low tolerance
for mistakes represent cultures that pursue efficiency
over effectiveness and exploitation over exploration
(Crossan and Hurst 2003). In contrast, high levels of
experimentation and tolerance for error are not associ-
ated with blind risk taking and lack of discipline, but
represent a culture that promotes action as opposed to
reflection as a way to understand and deal with reality
(Cunha et al. 1999) and where boundaries and mini-
mal constraints are defined so that experimentation can
Vera and Crossan: Improvisation and Innovative Performance in Teams
208 Organization Science 16(3), pp. 203–224, © 2005 INFORMS
An experimental culture not only promotes the use
of improvisation skills by motivating team members to
risk the “four Cs” (the desire to be competent, com-
fortable, consistent, and confident, i.e., Claxton 1984,
Crossan and Sorrenti 1997), but also provides them with
the resources (e.g., time, people, and money) that enable
improvisational efforts to be successful. This context is
critical for innovation because, as stated by Caldwell
and O’Reilly (2003, p. 500), “innovation is largely
unpredictable and requires flexibility, opportunism, and
adaptability. The fact that innovation demands creative,
nonroutine responses makes it difficult to design a pri-
ori programmed actions that will lead to innovation.
Caldwell and O’Reilly (2003) found that support for
risk taking and tolerance of mistakes were two cultural
norms that promoted behaviors associated with innova-
tion. When team members perceive their environment as
interpersonally nonthreatening and tolerant of, or even
supportive of, taking risks and trying new approaches,
higher levels of psychological safety and engagement in
innovative processes, such as improvisation, ensue (e.g.,
West 1990, Edmondson 1999, Gilson and Shalley 2004).
The Post-It Note discovery at 3M is an example of a
culture that offered people time to experiment and try
new things. A mistake by a 3M researcher was “yes-
anded” by another researcher, who considered the use of
the “failed” adhesive in keeping pieces of scrap paper
from falling out of his choir book (Fry 1987). Several
iterations of “yes-anding” among 3M researchers led to
the Post-It Notes we use today.
Hypothesis 3. The more experimental a team’s cul-
ture, the more positive the relationship between collec-
tive improvisation and innovation.
Real-Time Information and Communication: “Be Pres-
ent in the Moment. A basic rule of improvisational the-
ater requires players to be attentive to what is happening
around them, to be “present” and alert (Spolin 1963,
Johnstone 1979). When developing a common story in
the moment, a lack of attention and alertness to the infor-
mation coming from fellow players and the audience
leads to contradictory actions, conflict, and frustration
of both the actors and audience. If a player is planning
ahead and thinking about the direction he or she wants
the action to go, then the actor is not paying attention
to what is going on in the moment and will miss oppor-
tunities for discovery (Halpern et al. 1994). Frost and
Yarrow (1990, p. 100) summarize the principle of “pres-
ence” in improvising:
“Presence” means the performer is fully there, “present,
in the present tense, inside the moment. His attention is
wholly on the task and yet, most important, his aware-
ness extends beyond the immediate space around him to
include the audience’s space.
Organizations and work teams that want to become
more improvisational need to learn to be attentive and
alert to what is happening in the now of the firm. This
requires an infrastructure that provides teams with rel-
evant real-time information from their context—within
their team and inside or outside of their firm. Real-
time information is defined as information about a firm’s
operations or environment for which there is little or no
time lag between occurrence and reporting (Eisenhardt
1989, Eisenhardt and Tabrizi 1995); it enables real-
time communication—that is, the interaction within and
between teams based on timely information (Brown and
Eisenhardt 1998). A low level in this variable means that
real-time information and communication is infrequent
and that teams primarily know about their specific step
in the process, but are not well aware of what is happen-
ing in the environment or in the rest of the firm. High
levels of real-time information and communication are
not to be understood as random or chaotic; they mean
that communication is fluid and flows are wide-ranging
and focused on operating information.
Research on team innovation has mentioned open
information sharing and communication as a critical
aspect to achieve high levels of participation in inno-
vation processes in teams (West 1990, Agrell and
Gustafson 1996). When teams feel they lack updated
information, the risks of engaging in the creative and
spontaneous process of improvising seem too high.
Being aware of what is happening facilitates improvisa-
tion because up-to-date information can replace the coor-
dinating role of a plan and provide teams with immediate
feedback about the consequences of their actions as they
improvise (Moorman and Miner 1998b, Cunha et al.
1999). This is consistent with Eisenhardt’s (1989) find-
ings concerning fast decision making; she argues that top
managers attending to real-time information are actu-
ally developing their intuition, which enables executive
teams to react quickly and accurately to changes in their
Hypothesis 4. The greater the team’s level of real-
time information and communication, the more positive
the relationship between collective improvisation and
Memory: “Draw on Reincorporation and Ready-
Mades.Johnstone (1979) emphasizes two narrative
skills improvisers need to develop: free association and
reincorporation. Frost and Yarrow (1990, p. 134) sum
them up: “Free association takes care of invention and
development; reincorporation takes care of structure.
That is, while free association reflects the basic nature
of improvisation as creative and spontaneous, the skill
of reincorporation reminds players that improvisation
does not mean “anything goes,” and that creating a
coherent scene requires them to remember and reincor-
porate what has already been introduced in the past.
This principle refers to both the troupe’s memory of
the current performance and the lessons from previous
Vera and Crossan: Improvisation and Innovative Performance in Teams
Organization Science 16(3), pp. 203–224, © 2005 INFORMS 209
performances. Indeed, all improvisers draw on ready-
mades (e.g., short motifs or clichés) as they create their
novel stories (Sawyer 2000). The improvisational troupe
develops memory about scenes created in the past that
actors can recombine in present improvisations (Vera
and Crossan 2004). For example, improvisation groups
such as Second City have planned sets that they know
work, around which they build their spontaneous sets
(Crossan 1998).
Work teams interested in improving their ability to
improvise need to store ready-mades or routines, make
them accessible, and learn to draw on them and reincor-
porate them while improvising. In firms, ready-mades
and the principle of reincorporation are translated into
the notion of organizational or team memory—“stored
information from an organization’s history that can be
brought to bear on present decisions” (Walsh and Rivera
1991, p. 61). Memory includes declarative and proce-
dural knowledge stored in the systems, structure, strat-
egy, culture, rules, and procedures (Crossan et al. 1999).
These knowledge depositories exist not only at the orga-
nizational level, but also at the team level. Transac-
tive memory, for example, acts as a storage device
in which teams, based on their shared experiences,
encode, store, and retrieve relevant information together
(Liang et al. 1995).
The role of memory in improvising is paradoxical.
Memory may impede the incidence of improvisation
when team members deal with novel situations by sim-
ply replicating past routines. However, when teams actu-
ally engage in improvisation, memory becomes a helpful
resource for them because improvisation is frequently
the result of the creative recombination of previously
successful routines of knowledge and action (Weick
1993, Moorman and Miner 1998a, Miner et al. 2001).
Access to diverse memory resources helps teams to
improvise more effective and innovative solutions than
they would with a lack of, or a limited pool of, insti-
tutionalized knowledge (Vera and Crossan 2004). In
the Honda case, for example, the management team
could quickly improvise a novel strategy to introduce
50cc bikes into the U.S. market by creatively recombin-
ing Honda’s repertoire of routines with previous expe-
riences in marketing and sales in different countries
(Pascale 1984).
Hypothesis 5. The greater the team’s level of mem-
ory (e.g., procedures and systems), the more positive
the relationship between collective improvisation and
Training. Building on the practice of theater, our last
hypothesis is that organizational members cannot only
learn from the principles of theater and translate them
to organizational life, but they can improve the effec-
tiveness of their improvisational process through training
that addresses the elements needed to improvise well
(creativity, spontaneity, expertise, and teamwork quality)
and the creation of a context that supports improvisa-
tion (an experimental culture, real-time information and
communication, and memory). Indeed, Spolin’s (1963,
p. 1) motto was “Everyone can act. Everyone can impro-
vise.” She exemplified this when teaching improvisa-
tional games not only to actors, but to children and to
illiterate immigrants in community work. Much research
supports the idea that individuals can learn to be more
spontaneous and creative (e.g., De Bono 1973, Gardner
1982, Amabile 1996, Sternberg 1999). However, this is
not an easy task because, as Johnstone (1979, p. 77)
argues, “Most children can operate in a creative way
until they’re 11 or 12, when suddenly they lose their
spontaneity, and produce imitations of ‘adult art’.” Nev-
ertheless, theater has shown that individual potentialities
to be spontaneous and creative can be rediscovered and
developed through exercises. Johnstone (1979, p. 10)
summarizes this thinking:
You are not imaginatively impotent until you are dead;
you are only frozen up. Switch off the no-saying intellect
and welcome the unconscious as a friend; it will lead you
to places you never dreamed of, and produce results more
“original” than anything you could achieve by aiming at
In organizational settings, training in improvisation
needs to start by developing an understanding of what
improvisation is and positioning it as a legitimate and
even recommended choice when facing circumstances
of urgency, ambiguity, and uncertainty (Crossan et al.
2005). Crossan and Sorrenti (1997, p. 174) emphasize
this point when they mention that “without an awareness
of the need for improvisation, or an understanding of
what it entails, there will be little motivation to engage
it.” As discussed earlier, in addition to developing aware-
ness, training in improvisation includes exercises aimed
to develop process skills such as listening and communi-
cation, context-specific knowledge, a perspective and a
context enabling team members to go out of their com-
fort zone, and techniques to promote “yes-anding” and
develop shared responsibility in teams (Crossan 1998).
The goal of the training is to increase the ability of indi-
viduals and teams to improvise well and their motivation
to rely more on spontaneous and creative actions, when
required by the situation.
Hypothesis 6. Training will increase the incidence of
Research Setting and Data Collection
This study was conducted in a large municipal set-
ting, which we refer to as the “City.” Top manage-
ment was interested in an improvisation intervention,
which involved 25 work teams (175 employees) and
Vera and Crossan: Improvisation and Innovative Performance in Teams
210 Organization Science 16(3), pp. 203–224, © 2005 INFORMS
was provided with an external training consultant. We
obtained authorization from the City and the training
consultant to invite the teams in the training to partici-
pate in this study. The City also authorized an additional
25 work teams (173 individuals) to be invited to take
part in the research. Random assignment of the teams
to the training was pursued with a few exceptions given
the time restrictions of some of the teams (e.g., teams
going through particular periods of high workload). The
selection of all targeted teams was made in coordination
with a panel of three senior City executives familiar with
the work of the corporation. All targeted teams worked
in an environment where improvisation was likely to
occur—that is, in jobs with direct contact with external
customers or jobs in which teams dealt with one or more
of the following: unexpected or novel events, resource
scarcity, and urgency. Our panel confirmed that, in all
teams, members were interdependent in their tasks.
Pretest surveys (before the training) and follow-up sur-
veys (two months after the training) were distributed
through the City’s internal mailing system to the poten-
tial respondents, who were instructed to put the com-
pleted questionnaire in a return envelope addressed to
the researchers. The respondents were advised that the
surveys were number coded because we needed to match
the responses of the pretest and follow-up surveys,
but that their responses would be revealed only to the
researchers. The company received summary statistics
of aggregated data. We collected data from two sources:
the team members and their supervisors. Team members
filled out a survey that included items measuring the
independent variable (team improvisation), contextual
variables, and demographics. On a separate rating form,
each team’s supervisor rated the dependent variable—
team innovation.
Our unit of analysis was the work team. We sent
the surveys to 50 work teams (348 individuals). Sur-
veys from teams with at least 30% of their members
having completed the follow-up survey (including the
team supervisor and a minimum of two team mem-
bers), were incorporated in the final data sets. Meeting
this requirement were 38 teams (22 teams in the train-
ing group and 16 teams in the control group). Average
response rate per team was 68%; in 82% of the teams
the response rate was over 50%. The overall response
rate of the pretest survey was 67% (232 individuals,
145 in the training group and 87 in the control group),
and the response rate of the follow-up survey was 51%
(177 individuals, 100 in the training group and 77 in the
control group). Teams in the final sample represented
all types of jobs, including auditors, marketing coordi-
nators, and engineering inspectors, in all departments of
the City (i.e., community services, works and transporta-
tion, planning and building, legal services, finance, man-
agement services, and business development and public
relations). There was no significant difference between
teams in the training (T) and the control (C) groups in
terms of average team size (T =686 members; C =763
members), average City tenure (T =755 years; C =
828 years), average team tenure (T =480 years; C =
601 years), average age (T =4220 years; C =4306
years), and gender composition (T =62% males; C =
60% males).
Data collection also included 20 semistructured inter-
views with individuals coming from different work
teams in our sample. Interviewees were invited to
describe events in which their team had to “come up
with something really fast” or “think on their feet.
The objective was to understand the circumstances in
which City employees improvise and to examine the fac-
tors influencing the success of improvisational processes.
Table 1 provides examples of the events reported.
Finally, we collected additional information from indi-
viduals participating in the training workshops. Because
the training intervention was developed independently
of this research, we examined its content to assess if
it reflected the theory of improvisation developed in
this study. Although the term “improvisation” was rarely
used during the training workshops, the training empha-
sized the goal of being responsive and creative under
pressure and the roles that spontaneity, flexibility, intu-
ition, expertise, and teamwork play in this process. The
training also discussed the need to create a culture that
supports and rewards experimentation, and the need to
continuously look at the internal and external environ-
ment to remain nimble. (For a detailed description of the
training, see Rosenberg 1998.)
Each training workshop lasted two days and included
between 10 and 20 people. Members of the same work
team were not able to attend the training workshop
together because of the requirement of top management
that the training not disrupt the flow of the City’s oper-
ations. However, an effort was made to train the mem-
bers of the same team within a short time frame so
they could start using the new skills as a team. Nev-
ertheless, we expected that the short-term effects of
the training workshops (e.g., period of approximately
two months after the training) would be stronger in
influencing individual behavior than they would be in
influencing team behavior. This expectation was based
on research arguing that team training is more helpful
than individual training for tasks performed in teams,
but team training is not more beneficial than individual
training for tasks performed individually (Hollingshead
1998). In our case, we expected that the fact that team
members were not trained together would not hurt the
development of individual improvisational skills, but
would delay the creation of shared mental models about
the demands of collective improvisational processes.
Cannon-Bowers and Salas (1998) propose that training
strategies designed to foster development of shared men-
tal models must develop in trainees a shared knowledge
Vera and Crossan: Improvisation and Innovative Performance in Teams
Organization Science 16(3), pp. 203–224, © 2005 INFORMS 211
Table 1 Examples of Improvisational Events
Context Events
Fire services We practice cutting cars with old models. On the scene, cars are different—for example, in terms of air
bags. Car manufacturers do not tell firefighters about their new standards. Every time we have people
trapped in cars, we have to make a decision in the moment and find a way to get them out of the car.
In one of the fires, firefighters came up with an instrument to save the life of a heavy person that they were
having trouble carrying. They saved him, but two other kids died. They lost some time trying to carry
the heavy person   This instrument is now a standard tool in other stations too.
Auditing We did an audit of a facility center, and in the process of that audit we came across a huge control issue
related to the approval of part-time labor   We had a meeting with the commissioner, director, and
manager of that area about what we found as weaknesses of control and what we felt they could do. It
was a huge and real risk   We came up with a new solution, and in half an hour we had some controls
in place that satisfied our issue and did not impede their business process.
Facilities management There was a mechanical breakdown of an air-conditioning system during a wedding reception in the
summer. Everybody was sweating to death. It was so hot outside that the actual unit on the roof was
overheating. We came up with the idea to set up a sprinkle system to cool the unit down so that it would
actually run.
Strategic planning In strategic planning, you use a process map and a charter. Oftentimes when you present them, it is not
what someone else thinks they are going to get. Thinking on our feet, we were able to redefine the model
and create a new model with a totally different look, a totally different visualbasically saying the
same thing; it was just a different way of expressing it and explaining it.
Consulting services The IT audit and consulting service is a new one   When we were creating this service, we had to
define the relationship between IT and audit   There was a lot of trial and error, going back and
forth negotiating to make sure the end product filled the IT needs and the audit needs   The process
included plans and being flexible and trying things   We had to try this, try that, see what worked
and what did not, and develop the service.
of the task, and an understanding of the role of each
team member in relation to all others and the character-
istics of each of the other members of the team. This
is consistent with the Liang et al. (1995) study, which
showed that student teams trained as teams performed
better than teams whose members were trained individ-
ually because of the development of transactive memory
during the training. In our case, teams at the City would
be able to start practicing the improvisational skills as
a team and start developing shared mental models and
transactive memory associated with improvisation only
after all team members had been trained. To address this
issue, we included measures for individual improvisation
and innovation in the pretraining and follow-up surveys.
In addition, six weeks after the training, training partici-
pants received a “transfer-to-work form” with questions
about their ability to use the new knowledge and skills
in their work, ways in which the training had impacted
their levels of innovation, and barriers they had encoun-
tered in using the skills. We received 96 forms (61%
response rate).
Measures for the dependent, independent, and moder-
ating variables are outlined below; the appendix pro-
vides the scale items. Seven-point Likert scales with the
anchors “strongly disagree” and “strongly agree” were
used. Measures were refined through two pilot tests,
one of them performed in the actual research site. The
13 divisions of the City participating in the pilot test did
not participate in the main study.
The dependent variable, innovation, was measured
with a two-item scale adapted from Roth’s (1993) inno-
vation scale developed for a service context. The two
items were averaged for an overall score (correlation
between items =074). We developed a seven-item
scale to measure the independent variable, improvisa-
tion, including both its creative and spontaneous facets.
Four items were adapted from the Tierney et al. (1999)
employee-creativity scale, and three items were created
for this study building on Unger and Kernan’s (1983)
measure of spontaneity and Moorman and Miner’s
(1998b) measure of improvisation. We averaged the
seven items (alpha =091). We used confirmatory fac-
tor analysis (CFA) with maximum likelihood estima-
tion to evaluate the validity of the multi-item measures
(Gerbing and Anderson 1988). We conducted a CFA
of the improvisation and innovation variables. Results
indicated a better fit for a two-factor model (GFI =
096, Jöreskog and Sörbom 1986; CFI =099, Bentler
1990; TLI =098, Tucker and Lewis 1973; 2=3844
df =23p=002; and 2/df =167relative to a
one-factor model (GFI =089, CFI =092, TLI =088,
2=13666 df =24p=000), and 2/df =569).
The difference in Chi-squares between the two models
was significant (2=9822 df =1, p<0001).
The measurement scale for the teams’ expertise was
based on three items adapted from the knowledge mea-
sures included in the Strategic Learning Assessment
Map (Bontis et al. 2002); items were averaged to cre-
ate an overall score (alpha =085). The scale assessing
Vera and Crossan: Improvisation and Innovative Performance in Teams
212 Organization Science 16(3), pp. 203–224, © 2005 INFORMS
teamwork quality consists of five items adapted from
the teamwork scale included in the “Getting Back to
Business” questionnaire of the “Improvise to Innovate”
training program (Crossan 1997b); items were averaged
for an overall score (alpha =090). Four items adapted
from the culture scale included in the “Getting Back
to Business” questionnaire of the “Improvise to Inno-
vate” training program (Crossan 1997b) were averaged
to create a measure of experimental culture (alpha =
081). The four-item scale measuring real-time informa-
tion and communication was partly based on Hanlon
and Taylor’s (1991) measure of organizational commu-
nication behaviors. We averaged the four items (alpha =
080). Finally, the measurement scale for team memory
consisted of four items adapted from Moorman and
Miner’s (1997) organizational memory scale and the
knowledge measures developed by Bontis et al. (2002)
(alpha =080). In a CFA of the moderating variables
the five-factor model provided a better fit (GFI =090,
CFI =096, TLI =096, 2=22354 df =141p=
000, and 2/df =158) than the one-factor model
did (GFI =064, CFI =067, TLI =062, 2=89490
df =151p=000, and 2/df =592. The differ-
ence in Chi-squares between the two models was signif-
icant (2=67236 df =1, p<0001).
To isolate the impact of improvisation, a number of
variables that are considered drivers of team innovation
and effectiveness (Hackman and Morris 1975, Cohen
and Bailey 1997) were included as controls: job char-
acteristics, team size, team tenure, company tenure, and
participation in training. A five-item scale for job char-
acteristics was newly developed. To measure team size, a
listing of the City’s personnel was obtained. Team tenure
and City tenure were based on individual-level data pro-
vided by team members who completed the survey. Par-
ticipation in training was measured through a dummy
variable—“1” indicated that the team had participated in
the training, and “0” indicated that it had not.
We present our results in two sections. First, we discuss
the testing of Hypotheses 1 through 5, which predict
the factors that enhance a team’s ability to improvise
and, ultimately, innovate. Second, we discuss the testing
of Hypothesis 6, which evaluates the role of training in
influencing the incidence of improvisation.
The Role of Team and Contextual Factors
To test Hypotheses 1 through 5 concerning the fac-
tors contributing to effective improvisation, we used a
data set including the team members’ responses to the
follow-up surveys aggregated into team scores (N=38
teams, including training and control groups). To jus-
tify the aggregation, a within-team correlation rwgwas
computed for the improvisation and moderating vari-
ables (James et al. 1984). Mean rwg values ranged from
0.71 to 0.77, which indicated good agreement. For the
innovation variable we used the scores reported by the
teams’ supervisors. Because of the training manipula-
tion performed, we expected more variance in improvi-
sation in the posttraining sample than in the pretraining
sample, so that the moderating effects of the factors
would be more pronounced in the posttraining sample.
Table 2 displays means, standard deviations, and correla-
tions among all variables using the posttraining sample.
We tested our hypotheses with hierarchical models
estimated using ordinary least squares. Independent vari-
ables were mean centered to reduce multicollinearity
among the interaction terms and their individual compo-
nents (Aiken and West 1991). Given the limited num-
ber of teams, we utilized separate regression models for
each of the moderating factors hypothesized to influ-
ence the innovative outcomes of improvisation. To main-
tain good power and validity of a multiple regression
analysis, Cohen et al. (2003) recommend that only the
central independent variables be included. Hence, the
control variables, none of which had a significant corre-
lation with the innovation variable, were removed from
the regression analyses. Also, when we ran stepwise
regression analyses that included the block of control
variables, none of the controls was significant, which
confirmed the decision to drop them. The results of
the final hierarchical regression analyses are shown in
Table 3 and illustrated in Figure 1. Variables were
included in the regression models in the following order:
(1) improvisation, (2) the moderating factor, and (3) the
interaction effect between improvisation and the mod-
erating factor. Collinearity statistics calculated for all
regression analyses did not indicate distortions of results
due to correlations among regressors. We considered
10%, 5%, and 1% significance levels.
In Model 1 the impact of improvisation on innovation
was not significant (=033, p>010), consistent with
the underlying premise of the theory about the equiv-
ocal relationship between improvisation and innovation
when moderating variables were not considered. Given
no clear positive or negative relationship between impro-
visation and innovation, it made sense to look at the
roles played by the moderating factors.
The first hypothesis suggested that expertise had a
positive moderating effect on the link between improvi-
sation and innovation. Model 2 shows support for this
hypothesis. In Step 3, the interaction effect between
improvisation and expertise was positive and significant
(=179, p<001; R2=018, p<001). This result,
depicted in Figure 1, is consistent with comments from
our interviewees. The role of experience and knowledge
was frequently mentioned as part of the description of
successful improvisational events, as shown in the fol-
lowing example in the auditing unit:
[Regarding] a new situation with an account that needs
to be reconciled, because of my experience, I can think
quickly and find ways of doing it. This ability comes
Vera and Crossan: Improvisation and Innovative Performance in Teams
Organization Science 16(3), pp. 203–224, © 2005 INFORMS 213
Table 2 Descriptive Statistics and Correlation Matrix
Constructs 12345678910111213141516
1. Innovation
2. Improvisation 016
3. Expertise 004 048∗∗
4. Teamwork quality 014 065∗∗ 059∗∗
5. Experimental 026 058∗∗ 052∗∗ 069∗∗
6. Real-time information 006 022 025 026 047∗∗
and comm.
7. Memory 043∗∗ 033017 024 021 047∗∗
8. Training (dummy) 015 010 010 003 004 001 007
9. Team size 000 008 002 010 013 003 007 011
10. City tenure 002 011 010 007 002 012 003 018 035
11. Team tenure 012 010 000 013 019 021 013 025 021 067∗∗
12. Job Ch.— 003 006 043∗∗ 015 018 017 018 020 002 015 007
unexpected events
13. Job Ch.—new ways 016 032012 023 023 015 001 011 013 032031+037
of doing things
14. Job Ch.—repetitive 005 002 031+014 022 050∗∗ 037∗∗ 007 002 016 026 024 029+
15. Job Ch.—ambiguous 015 007 011 001 013 013 021 017 002 030+029+022 021 007
16. Job Ch.—structured 010 028+020 018 001 008 053∗∗ 023 035001 001 011 012 006 046∗∗
Mean 444 484 544 502 444 400 477 058 718 786 531 563 488 437 510 467
Standard deviation 106 050 051 055 056 083 065 050 360 199 242 092 066 079 100 087
Notes.N=38 teams.
∗∗p<001, p<005, +p<010.
naturally, while when you have no experience, it is
harder   Gut feeling helps and is related to your expe-
rience. When I am doing audits, I can deal with situations
and eliminate things to look at or not to look at, because
I just know.
Table 3 Results of Regression Analyses
Effects of improvisation on innovation by moderating condition
Model 1 Model 2 Model 3 Model 4 Model 5 Model 6
A=Expertise A =Teamwork A =Exp. culture A =Real-time A =Memory
quality info. and comm.
Improvisation 033 033 033 033 033 033
R2003 003 003 003 003 003
Improvisation 049 025 002 032 003
Moderating factor A 032 012 048 004 070
R2002 002 004 000 016
Improvisation 054 012 005 031 002
Moderating factor A 023 034 062+002 084
Interaction effect: 179∗∗ 142126088042
(improvisation ×A) 063066053040051
R2018∗∗ 012013012002
Overall R2003 023 014 020 015 020
F092 331 191 284 193 288
p035 003 014 005 014 005
Notes.N=38 teams.
Nonstandardized coefficients are reported with standard errors below in parentheses.
∗∗p<001, p<005, +p<010.
The second hypothesis argued that the better the
teamwork quality in a team, the greater the likelihood
that improvisation leads to innovation. Model 3 shows
support for this hypothesis. In Step 3, the interaction
between improvisation and teamwork was positive and
Vera and Crossan: Improvisation and Innovative Performance in Teams
214 Organization Science 16(3), pp. 203–224, © 2005 INFORMS
Figure 1 Effects of Improvisation on Innovation by Moderating Condition
Improvisation ×Expertise
Low High
Low expertise High expertise
Improvisation Teamwork skills
Low High
Low teamwork skills
tea mwo rk
Improvisation Experimental culture
Low experimental culture High experimental culture
Improvisation Real-time information and
Low High
Low real-time information and communication
High real-time information and communication
Improvisation Memory
Low High
Low memory High memory
Note. N=38 teams.
significant (=142, p<005; R2=012, p<005).
This interaction is plotted in Figure 1. To further inform
this result, we examined our interview data. Participants
expressed diverse levels of satisfaction about the way
their teams worked together. Although the existence of
trust among team members was only one of the items
in the scale measuring teamwork skills, trust was fre-
quently emphasized in conversations about improvisa-
tional events and seemed to be a critical factor when
team members tried to work together to find a new
solution to a problem. For example, in one success-
ful incident an interviewee mentioned: “We know that
we rely on each other to get things done. They know
that they can trust me with information, and I know
I can trust them with information.” In contrast, as part
of an unsuccessful incident another participant com-
mented: “At times there isn’t a lot of trust between and
within teams. People don’t entrust each other with the
right information on time.” Overall, the comments about
the role of teamwork dynamics in improvising were
consistent with our hypothesis, as summarized by one
In this project we were all willing to achieve a common
goal. We knew what the goal was   We all worked
together and we all succeeded together. But when you
see that decisions are not made by the right people, that
success is individualized, and that only one person takes
the credit, or that somebody doesn’t want to cooperate for
Vera and Crossan: Improvisation and Innovative Performance in Teams
Organization Science 16(3), pp. 203–224, © 2005 INFORMS 215
any reason, then it is very difficult to think creatively and
create something new. It doesn’t become a team concept.
It becomes an individualistic environment. It becomes
In the third hypothesis we predicted that the more
experimental a team culture, the more positive the effect
of improvisation on innovation would be. Model 4 shows
support for this hypothesis. In Step 3, the interaction
effect between improvisation and experimental culture
was positive and significant (=126, p<005; R2=
013, p<005). This result, illustrated in Figure 1,
is consistent with the comments of our interviewees.
The role of culture was often mentioned when describ-
ing successful improvisational events, as shown in this
example (also in the auditing unit):
In this new project, these guys had the freedom to work
and to make their own mistakes so that they learn from
them without being so close-minded or difficult so that
they don’t want to come and ask for advice or help. It’s
a balance, a very tricky balance, giving that comfortable
feeling that it’s OK to mess up—we will fix it, and we
are going to learn from it.
The fourth hypothesis suggested that real-time infor-
mation and communication had a positive moderating
effect on the relationship between improvisation and
innovation. Model 5 shows support for this hypothe-
sis. In Step 3, the interaction effect between improvisa-
tion and real-time information and communication was
positive and significant (=088, p<005; R2=
012, p<005). This interaction is plotted in Fig-
ure 1. To further understand how the availability of
timely and diverse information supported improvisation
effectiveness, we examined our interview data. There
was great variance in the degree to which teams felt
informed about what was happening in their teams and
in other teams in the City. Consistent with our hypoth-
esis, although some interviewees were positive about
the timeliness and relevance of the information they
received, others recognized the challenges they expe-
rienced when dealing with unexpected events without
knowing what was going on in the team or the firm.
For example, a manager in a team with high levels of
information and communication commented: “There is
open communication in the team; we have regular meet-
ings. The door is always open for staff  We can talk
about problems while they are happening.” In contrast,
the team in charge of the City’s switchboard and infor-
mation desk commented on some of the difficulties faced
when responding to unexpected requests:
We try to be updated about what is happening, for exam-
ple by reading the newspaper early in the morning 
Whatever happens in the environment with an [City-
related] issue, we take the impact through the calls we
get   Sometimes there are problems of communica-
tion with other departments. We don’t know about all
the meetings happening. People come and ask us where
the meeting is or there are changes of location in the
building, and we are not updated.
Finally, in the fifth hypothesis, we predicted that the
more memory available to a team, the more positive
the effect of improvisation on innovation would be.
As Model 6 shows, this hypothesis was not supported.
In Step 3, the interaction effect between improvisation
and memory was not significant (=−042, p>010;
R2=002, p>010). Figure 1 illustrates this result.
This lack of support is informed by comments from our
interviews, in which participants frequently stressed the
sometimes “constraining” role that procedures may play
when improvising in the City. For example, one inter-
viewee explained:
All departments have specific rules and guidelines 
You can’t not deal, not worry, about legal liability and
health and safety issues with the public unless you
go by policies and procedures   We have hundreds of
rules and procedures in order to ensure the delivery of
services and the safety of citizens.
These comments suggested that participants believed
that following rules, standards, and procedures was the
“right” and “safe” thing to do in the City. We hypothe-
sized that a great amount of memory would enable teams
to richly recombine routines when needed. However, if
teams saw institutionalized practices as restrictive, they
might be less likely to creatively work with routines
when dealing with novel situations. Still, several inter-
viewees mentioned the tension between working in the
framework of rules and at the same time being ready to
react to situations for which no procedure is available.
For example:
As firefighters, every situation we face is different. We
have tools that we can use in different situations, but you
never know what you are going to get  When we are
trained, we are told to “go by the book.” But, at the
same time, we are trained to react, and rules get you to
some point. After years of experience, you know that the
book is not that smart.
Interview data also suggested that it was easier for
some teams to recombine informal procedural memory
when dealing with new events than it was to recombine
formal declarative and procedural memory. One facility
manager provided an example:
There are tricks we have picked up through observation
of how people dealt with situations in the past, of how
your supervisors dealt with situations in a crisis   Itis
a matter of drawing on this wealth of knowledge that the
team has secured over the years to be able to come up
with a new solution.
In this example, the subject referred to anecdotal
knowledge obtained through observation and practice,
transferred through conversations, and not likely to be
Vera and Crossan: Improvisation and Innovative Performance in Teams
216 Organization Science 16(3), pp. 203–224, © 2005 INFORMS
stored in manuals but in the team culture. This sug-
gested that, in contexts where strong value is assigned to
respecting policies and procedures, informal procedural
memory and, in particular, informal practices and gen-
erally accepted norms may be the most beneficial ones
when improvising.
The Role of Training
To test Hypothesis 6 concerning the effects of impro-
visation training, we used a data set with individual-
level data from respondents who completed both the
pretest and follow-up surveys (113 individuals, 73 in
the training condition and 60 in the control condition).
The data set included perceptions of individual impro-
visation, individual innovation, and moderating factors
prior to and after the training.
We performed between-group comparisons of the pre-
training and posttraining responses for the training and
control subsamples to isolate the effects of the train-
ing. In the pretraining condition, the training and control
subsamples were homogeneous in their improvisation
levels (improvisationtraining =519, improvisationcontrol =
524, t=038, p>010). Prior to the training, inno-
vation was marginally higher in the control subsample
than in the training subsample (innovationtraining =404,
innovationcontrol =440, t=181, p<010). This char-
acteristic of our sample constituted a tougher test of the
theory. In the posttraining condition, the level of impro-
visation in the training group increased (as expected);
however, the difference in improvisation between the
training and control subsamples was not significant
(improvisationtraining =536, improvisationcontrol =519,
t=−135, p>010). After the training, there was no
difference in innovation between the training and control
subsamples (innovationtraining =424, innovationcontrol =
448, t=135, p>010).
We also performed within-group comparisons of all
variables in the pretraining and posttraining conditions.
In the training subsample the only variable that showed a
significant difference was improvisation. The posttrain-
ing level of improvisation (mean =536) was larger than
the pretraining level (mean =519). This difference was
statistically significant (t=260, p001). In contrast,
there was no significant difference in the pretraining
(mean =524) and posttraining levels of improvisa-
tion (mean =519) in the control subsample (t=063,
p>010). Overall, the results of the between-group and
within-group tests provide partial support for Hypothe-
sis 6, which predicted that the training would increase
the incidence of improvisation.
To get further insight into the impact of the train-
ing on improvisation, we performed a post-hoc anal-
ysis of correspondence between the team-level effects
(tested in Hypothesis 1 to 5) and individual perceptions.
In doing so, we explored how the training impacted
the ability of individuals to improvise well by run-
ning a series of regression analyses of the effects of
improvisation on innovation for each of the moderat-
ing factors and the pretraining and posttraining condi-
tions. Because the individuals included in our data set
worked in teams, prior to performing the regressions we
tested for violations of the interdependence assumption.
A residual plot showed a random pattern of residuals,
which implied independent errors; thus, the assumption
of independence was not violated. In addition, indepen-
dent variables were mean centered prior to the regression
analysis. Results of these follow-up tests for the training
subsample are shown in Table 4 while results for the
control subsample are shown in Table 5.
Our expectation in the post-hoc analysis was that dur-
ing the training workshops participants would learn to
make use of the factors that help them to improvise bet-
ter (expertise, teamwork quality, experimental culture,
real-time information and communication, and mem-
ory), so that in the training subsample the interaction
effect between improvisation and the moderating fac-
tors would be stronger in the posttraining condition than
in the pretraining condition. In contrast, in the case of
the control subsample, we expected no difference in
the interaction effects between improvisation and the
moderating factors in the pretraining and posttraining
Table 4 shows that for expertise, teamwork qual-
ity, and experimental culture, while the interaction
effects were not significant in the pretraining condi-
tion (improv ×expertise =021, p>010; improv×teamwork =
009, p>010; improv ×exp. culture =008, p>010),
they were positive and significant in the post-
training condition (improv ×expertise =034, p<001;
improv ×teamwork =035, p<001; improv×exp. culture =
037, p<001). In the case of real-time information and
communication, the interaction effect was marginally
significant prior to the training (=023, p<010)
and significant after the training (=040, p<001).
Finally, no difference was observed in the interaction
effect between improvisation and memory, which was
significant in the pretraining (=027, p<005) and
posttraining conditions (=029, p<005). This pat-
tern of results clearly differed from that found in the
control subsample. Table 5 shows that for experimental
culture, real-time information and communication, and
memory, no difference was observed in the interaction
effects, which were not significant in both the pretrain-
ing (improv ×exp. culture =002, p>010; improv ×RT info =
021, p>010; improv ×memory =013, p>010) and
posttraining condition (improv ×exp. culture =−001, p>
010; improv ×RT info =−004, p>010; improv ×memory =
014, p>010). In the case of teamwork quality, the
interaction effect was marginally significant prior to the
training (=026, p<010) and not significant after
Vera and Crossan: Improvisation and Innovative Performance in Teams
Organization Science 16(3), pp. 203–224, © 2005 INFORMS 217
Table 4 Post-Hoc Tests: Results of Regression Analyses in Training Group
Effects of improvisation on innovation by moderating condition and pretraining and posttraining conditions
A=Expertise A =Teamwork A =Exp. culture A =Real-time A =Memory
quality info. and comm.
Improvisation 036036036036036036
Improvisation 034+021 005 027 038
Moderating factor A 006 037∗∗ 039∗∗ 032∗∗ 016
R2000 013∗∗ 015∗∗ 013∗∗ 003
Improvisation 039019 007 032+032+
Moderating factor A 007 038∗∗ 041∗∗ 031∗∗ 009
Interaction effect: 021 009 008 023+027
(improvisation ×A) 015014010014012
R2003 005 001 003+006
Overall R2006 009 019 022 022 015
F456 223 547 633 646 414
p004 009 000 000 000 001
Improvisation 002 002 002 002 002 002
R2000 000 000 000 000 000
Improvisation 003 003 012 009 002
Moderating factor A 057∗∗ 049∗∗ 050∗∗ 043∗∗ 031∗∗
R2032∗∗ 032∗∗ 032∗∗ 031∗∗ 014∗∗
Improvisation 002 004 007 010 004
Moderating factor A 049∗∗ 043∗∗ 042∗∗ 039∗∗ 025∗∗
Interaction effect: 034∗∗ 035∗∗ 037∗∗ 040∗∗ 029
(improvisation ×A) 015011013012013
R2005008∗∗ 007∗∗ 010∗∗ 006
Overall R2000 037 040 040 041 019
F001 1354 1544 1519 1583 545
p091 000 000 000 000 000
Notes.N=73 individuals.
Nonstandardized coefficients are reported with standard errors below in parentheses.
∗∗p<001, p<005, +p<010.
the training (=017, p>010). Finally, the interaction
effect for expertise was marginally significant in the pre-
training condition (=032, p<010) and significant
in the posttraining condition (=039, p<001).
We also performed Chow tests (Chow 1960) to test
the difference between the pretraining and posttraining
regression lines in the training subsample. We obtained
significant differences for the models including expertise
(F=211, p<01) and teamwork quality (F=251,
p<005). As a whole, the pattern of results in our
post-hoc analysis shows correspondence between team-
level effects and individual perceptions. Furthermore,
the post-hoc analysis offers new insights into Hypothe-
sis 6 by suggesting that the training increased not only
the incidence, but also the quality of improvisation by
improving the individual ability to rely on expertise
and teamwork while improvising. Although the Chow
tests for the experimental culture and real-time infor-
mation and communication models were not significant,
the direction of the results was suggestive. With the
exception of the memory variable, the role of the influ-
encing factors tended to be more pronounced in the
posttraining sample, suggesting that after the training
individuals made better use of these factors while impro-
vising. In contrast, in the control subsample, the pattern
of results suggested nonsignificant effects of most of the
moderating factors in both the pretraining and posttrain-
ing conditions.
Finally, to further understand the effects of the train-
ing on the improvisational behavior of the participants,
Vera and Crossan: Improvisation and Innovative Performance in Teams
218 Organization Science 16(3), pp. 203–224, © 2005 INFORMS
Table 5 Post-Hoc Tests: Results of Regression Analyses in Control Group
Effects of improvisation on innovation by moderating condition and pretraining and posttraining conditions
A=Expertise A =Teamwork A =Exp. culture A =Real-time A =Memory
quality info. and comm.
Improvisation 012 012 012 012 012 012
R2001 001 001 001 001 001
Improvisation 003 006 001 005 004
Moderating factor A 029028028013 016
R2007009009002 002
Improvisation 003 003 000 006 007
Moderating factor A 027+027028012 014
Interaction effect: 032+026+002 021 013
(improvisation ×A) 016013015015018
R2006+006+000 003 001
Overall R2001 013 016 009 006 004
F041 289 356 190 114 076
p052 004 002 014 034 052
Improvisation 028 028 028 028 028 028
R2004 004 004 004 004 004
Improvisation 015 006 001 019 018
Moderating factor A 037∗∗ 044∗∗ 058∗∗ 034∗∗ 032
R2013∗∗ 019∗∗ 027∗∗ 017∗∗ 010
Improvisation 016 004 001 018 020
Moderating factor A 044∗∗ 045∗∗ 058∗∗ 034∗∗ 031
Interaction effect: 039∗∗ 017 001 004 014
(improvisation ×A) 013013013011014
R2012∗∗ 003 000 000 002
Overall R2004 030 025 031 021 016
F260 796 633 845 500 354
p011 000 000 000 000 002
Notes.N=60 individuals.
Nonstandardized coefficients are reported with standard errors below in parentheses.
∗∗p<001, p<005, +p<010.
we examined the information from the transfer-to-
work forms (N=9661% response rate). When asked
the open-ended question, “What changes have you
noticed in yourself?,” the most frequent responses were,
“I am more open-minded,” “I look for more innovative
options,” or “I look at things from different perspec-
tives” (35 people). Participants reported that after the
training they had been “eager” to change their behavior
and that they had been able to apply the skills in their
job “to some extent.” They were asked about barriers to
using the skills. The two barriers most frequently men-
tioned were, “It was not practical to my situation,” and
“Our culture does not support these behaviors.” When
asked about “other barriers” to using the skills, par-
ticipants mentioned factors such as “high workload,
“lack of time,” or “have not encountered the situation
yet.” The most frequent suggestions for improving the
training were, “All employees, management, and coun-
cil members should attend training” (15 people), and
“refresher courses with team members,” or “some type
of follow-up” (6 people).
The three main findings of this study are (1) the
improvisation-innovation main effect is equivocal,
(2) the magnitude of the improvisation-innovation main
effect is contingent on the positive influencing effects of
four of the five predicted moderators, and (3) training
in improvisation increases the incidence and the qual-
ity of improvisation. In other words, improvisation is on
Vera and Crossan: Improvisation and Innovative Performance in Teams
Organization Science 16(3), pp. 203–224, © 2005 INFORMS 219
average valueless but has a clear positive effect on inno-
vation when combined with team and contextual mod-
erating factors. These findings offer support to the view
that improvisational processes are not inherently good or
bad, and suggest that the same theater principles helping
improvisers to master this skill (e.g., “practice,” “col-
laboration,” “agreement,” and “presence”) are not only
applicable in business settings, but can be learned and
effectively applied by organizational members through
The only theater principle showing no effect in im-
proving the innovative outcomes of improvisation was
that of “draw on reincorporation and ready-mades,
which we translated into the notion of team memory. We
concluded that the lack of strong support for this effect
could be associated with the specific characteristics of
the research site, a municipality in which multiple guide-
lines, bylaws, and policies guide the work. The view that
standard operating procedures are the rule in govern-
ment organizations seemed to prevail in our sample—
this line of thinking follows the traditional Weberian
bureaucracy model. Nevertheless, we hypothesized that
norms and procedures would enable teams to richly
recombine routines when needed, which was consistent
with Lipsky’s (1980) proposition that street-level bureau-
crats (e.g., police officers, public lawyers, and other
public officials) frequently find themselves in circum-
stances in which they have to make sense of the rules
and procedures, interpret them, and use their discretion
to adapt laws and procedures to specific cases. From
this perspective, the work of street-level bureaucracies
is rule saturated but not rule bound (Maynard-Moody
and Musheno 2003). In this respect, our interview data
suggest that informal norms and practices, and rules
of thumb accepted by the team, were useful pieces of
memory when people improvised in the City. The idea
that informal procedural memory may be more helpful
than declarative memory while improvising differs from
previous conceptual work associating high declarative
memory with the quality of improvisation (Cunha et al.
1999) and the generation of coherent and novel impro-
visational actions (Moorman and Miner 1998a). It may
be that this relationship differs in the public and private
sectors and that, in the former context, employees find
that informal stories about past successes or failures are
easier to adapt and recombine when facing a new situ-
ation because they are not part of a rigid manual, but
rather knowledge transmitted through practice, observa-
tion, and conversation. Future research will be needed to
further explore the idea that when memory is preserved
in stories, it is more open to adaptation, innovation, and
When assessing the success of the training, we
observed a small increase in the incidence of improvisa-
tion in the training group. More apparent was the effect
of the training on the innovative outcomes of improvi-
sation. In our post-hoc analysis, comparisons between
pretraining and posttraining conditions showed the effect
of the training in teaching individuals to take advan-
tage of expertise and teamwork dynamics to improve
the effectiveness of improvisational processes. In con-
trast, the effect of the training was weak in influencing
the individual ability to leverage contextual variables
such as culture and real-time information. Our find-
ings are focused, however, on individual-level training
effects, not team-level ones. In the open-ended ques-
tions included in the transfer-to-work form, participants
reported that they observed a larger change in behavior
in themselves than they did in their teams. Their com-
ments suggested several reasons for this effect. First,
the objective to train all members of the team, includ-
ing its supervisor, could not be achieved for all of the
teams (10% of individuals in the training group can-
celled their attendance for personal or work-related rea-
sons). Furthermore, top management of the City received
only a three-hour version of the training. Creating a cul-
ture and a context that supports collective improvisa-
tion is difficult if not all team supervisors and members
are trained, and if top management’s commitment to
the training is not highly visible to employees. Sec-
ond, to avoid disrupting the City’s operations, teams
were not trained together. However, training a team
is more than individually training the members of a
team. It is easier for a team to develop and practice
skills together if those skills have been learned together.
As mentioned previously, team training develops shared
mental models and transactive memory, which enhance
team performance. Individuals trained together develop
complex beliefs about one another’s skills, specialize in
remembering different aspects of the task, coordinate
behaviors more effectively, and display greater trust in
each other’s expertise (Moreland and Myaskovsky 2000,
Mohammed and Dumville 2001). Third, a few partici-
pants reported that their job was independent from the
rest of their team. These individuals may not report
behavioral change at the team level because they do
not work in constant coordination with others. Finally,
we evaluated the training with information collected two
months after the training. This might have been enough
time for individual improvisation skills to become inter-
nalized, but not enough time for team improvisation
skills to develop. Furthermore, this time period may be
too short for individuals to act on the contextual vari-
ables (culture, real-time information, memory) that sup-
port the improvisational skill. In summary, our training
data shed initial light on the opportunities for firms pro-
vided by training in improvisation, but at the same time
it points out the challenges of developing skills and cre-
ating behavioral change in individuals and teams.
Vera and Crossan: Improvisation and Innovative Performance in Teams
220 Organization Science 16(3), pp. 203–224, © 2005 INFORMS
A limitation of this study impacting the statistical power
of our tests was our access to data from a limited pool
of teams in one firm. Lack of power increases the prob-
ability of making Type II errors, that is, reaching the
wrong conclusion by accepting the null hypothesis when
it should be rejected (Lindsay 1993). Post-hoc analysis
of power in regression showed that, with a sample size
of 38 teams, our test had an 84% chance of detecting
a large effect p =005, a 45% chance of detecting a
medium effect, and a 10% chance of detecting a small
effect. In addition, because the study was performed in
only one firm (a public-sector organization), the exter-
nal validity of the results was affected. Nevertheless,
we suggest that our findings can also provide valuable
insights to private-sector organizations because public-
sector management has become increasingly results and
customer focused, with great attention being given to
targets and to productivity gains (Van Wart and Berman
1999, Moon and DeLeon 2001). The context of public
bureaucracies has been described as one of uncertainty,
resource scarcity, ill-defined goals, and high expectations
(Pressman and Wildavsky 1973, Lipsky 1980), which is
consistent with the challenges faced by many private-
sector firms. In addition, the City represented a wide
range of business services.
Implications and Future Directions
This study contributes to research on improvisation
in four ways. First, we build on the insights from
improvisational theater to parse the descriptive ele-
ments of improvisation (spontaneity and creativity) and
its prescriptive elements (what it takes to do it well
and innovate—expertise, teamwork quality, experimental
culture, real-time information and communication, and
memory). In doing this, we sought to help clarify the
conceptual confusion about the value of improvisation,
to provide a detailed description of what improvisation
entails, and to inform efforts to measure and observe
improvisation. We encourage researchers to shift the dis-
cussion from “idealizing” the role of improvisation in
firms, or “warning” about its dangers, to the detection
of factors associated with its effectiveness.
Second, this is one of the few studies that attempts
measurement of improvisation and, to our knowledge,
the first empirical study to test arguments based on the
improvisational theater metaphor. With our improvisation
scale we sought to help fill the gap in quantitative work
in the improvisation field, in which measures of impro-
visation are scarce and existing scales only partially rep-
resent the phenomenon. Although there is a shortage of
both field studies and empirical studies on improvisation,
prior research has largely been qualitative. In our study,
we took advantage of the benefits of both methods. While
surveys enabled the systematization of data coming from
38 teams and the testing of hypotheses, interview data
was useful in validating the results, interpreting the sta-
tistical relationships, and clarifying the nonfindings.
Third, we provided empirical evidence of the incidence
and innovative outcomes of improvisation in the munici-
pal context and showed that new insights about the rela-
tionship between improvisation and innovation could be
obtained by moving from contexts where improvisation
is expected (such as product development and crisis situa-
tions) to more counterintuitive settings (such as a public-
sector organization). Finally, we offered initial evidence
that the improvisational skill can be learned by organiza-
tional members. Because improvisation has been advo-
cated as a means to reconcile basic tensions in teams
and in firms such as those between flexibility and struc-
ture (Kamoche and Cunha 2001), and exploration and
exploitation (Crossan and Hurst 2003), acknowledging
that the skill can be developed is extremely important for
our understanding of team innovation and team training.
Although our focus in both theory development and
testing was on the construct of improvisation, improvi-
sation itself has important implications for other fields of
study. Obvious connections have been made to innova-
tion, creativity, and the literature on teams. Here we see
improvisation as a complex process that serves to reveal
some of the key challenges, tensions, and obstacles asso-
ciated with innovation, creativity, and teamwork. To fur-
ther investigate the factors that support improvisation in
firms, future research will need to take an in-depth look
at each of the teamwork dynamics that constitute an
improvisation capability. For example, we incorporated
trust among team members as an aspect of teamwork
quality. We see great potential for future research on the
specific role of trust in collective improvisation given the
tight link between trust and the conditions of risk and
interdependence (Rousseau et al. 1998) that characterize
Although not explicit, there are implicit connections
to the fields of organizational learning and strategic man-
agement. Indeed we arrived at the study of improvisation
from these fields as we sought to gain a better under-
standing of how organizations learn and adapt under
time pressure. Initially, we viewed improvisation as
the exploration (March 1991) or feed-forward (Crossan
et al. 1999) processes of organizational learning, but we
soon came to realize that improvisation was a mecha-
nism to manage the tensions between exploration and
exploitation. The tension is revealed quite clearly in this
study in terms of memory and routines. Routines, sys-
tems, structure, and strategy are elements of organiza-
tional learning that have become institutionalized. The
intent is for the firm to leverage and exploit prior learn-
ing. However, institutionalized learning may impede new
learning when it no longer serves the situation. As men-
tioned by one of the respondents, “when we are trained,
we are told to go by the book   But, at the same time,
we are trained to react, and rules get you to some
point  ” It appears that in the municipal context this
Vera and Crossan: Improvisation and Innovative Performance in Teams
Organization Science 16(3), pp. 203–224, © 2005 INFORMS 221
tension is quite pronounced, with respondents suggesting
that following standard procedures was the “right” and
“safe” thing to do and simultaneously providing exam-
ples of cases when they used their discretion to deal
with novel events. This supports Maynard-Moody and
Musheno’s (2003) observation that when the rules and
procedures fit, street-level judgment is not problematic—
there is no conflict, no dilemma. In contrast, discretion
is used when dealing with ambiguous and nonroutine
situations. Improvisation calls for identifying the mini-
mal constraints or rules that must be adhered to, rather
than building layers of routines and systems that become
ossified and are eventually tagged as “red tape.” The pro-
cess of improvisation and the training exercises serve to
magnify what we see as the thorny problems of team-
work, creativity, innovation, and organizational learning.
In doing so, it is our hope that researchers and managers
will be able to delve more deeply in the black box of
process to advance theory and practice in each of these
Ultimately, we are interested in one particular aspect
of organizational learning, which is the strategic renewal
of the organization. The institutionalized learning in the
form of systems, structures, processes, and routines are
aligned with the strategy and all take on a dynamic char-
acteristic as adjustments are made based on the con-
stant learning about the environment (e.g., customers,
suppliers, competitors, general trends). We view impro-
visation as a critical element of strategic renewal for
the same reasons outlined for teamwork and organiza-
tional learning in general; improvisation is a process that
helps reveal and enables researchers and managers to
deal with the tension between exploration and exploita-
tion. The improvisation of teams within organizations
is the raw material that either supports or impedes the
strategic renewal of the entire organization. Therefore,
while we view the study of improvisation in teams as
a critical area of research, we do not see it as the
end point. In fact, an important research stream is the
development of a multilevel theory of improvisation that
defines improvisational processes at different levels and
links them to each other in a business context. The the-
ory developed in this paper focused on the team level,
where team improvisation is the function of the impro-
visational capability of individual members and team
and contextual factors. An extrapolation of our work to
the organizational level would suggest that an organi-
zational capability to improvise is more than the sum
of the teams’ competence in improvisation; it results
from the interaction of team competences in improvi-
sation with contextual influences such as organizational
routines, reward systems, structure, culture, and strategic
The foregoing discussion of improvisation as it relates
to teamwork, creativity, innovation, organizational learn-
ing, and strategic renewal points to a common theme—
the dialectical nature of improvisation. Weick (1998,
p. 551) proposes that improvisation helps to reconcile
organizational tensions because “it is a mixture of the
pre-composed and the spontaneous, just as organizational
action mixes together some proportion of control with
innovation, exploitation with exploration, routine with
non-routine, automatic with controlled.” Given the train-
ing effects arising from this study, important implications
arise for future research in examining the possibilities
improvisation might hold for dealing with some of the
more vexing issues in the various fields we have noted.
If effective improvisation can be learned, research can
move from the description of the difficulties of managing
organizational tensions or paradoxes to the prescription
of how the development of effective improvisational
skills helps individuals to overcome the natural tendency
toward polarization, and can reconcile these tensions.
As well as academic contributions, this study presents
findings of interest to managers. Our results show that
improvisation can positively impact team innovation if
certain conditions are present. Managers can manage
factors such as the degree of experimentation in a cul-
ture, the frequency and diversity of real-time informa-
tion and communication, and the quality of teamwork
dynamics. In the case of culture, managers can leverage
improvisation by establishing the boundaries and mini-
mal constraints within which people are free to experi-
ment and take controlled risks. In the case of real-time
information and communication, managers could moti-
vate work teams to reflect on the kind of information
they need to be responsive in their jobs, and promote
mechanisms to establish fluid communication flows. It is
also important for managers to recognize that effective
improvisational dynamics in teams may need some time
and practice to be developed, in particular because trust
grows out of the experience of working together. Addi-
tionally, the effective improvisation of a whole division
should be linked to training and promotion systems that
reward initiative and experimentation. Finally, this study
provides senior managers initial evidence about the abil-
ity to develop the improvisational skill in individuals and
teams, and shows that this requires more than attendance
in training workshops; it requires the development of a
culture and a context that supports spontaneous and cre-
ative processes in firms.
The authors would like to thank Henrich Greve, Senior Edi-
tor, and the three anonymous reviewers for their very helpful
comments and suggestions during the review process. They
also want to acknowledge Chris Higgins, Amy Hillman, John
Hulland, Anne Miner, Charlene Nicholls-Nixon, Rod White,
and Margaret Ann Wilkinson for their insightful input on ear-
lier versions. This research received funding from the Social
Sciences and Humanities Research Council of Canada and the
Richard Ivey School of Business.
Vera and Crossan: Improvisation and Innovative Performance in Teams
222 Organization Science 16(3), pp. 203–224, © 2005 INFORMS
Appendix. Questionnaire Items
Variable Items
Innovation The team frequently introduces new product/service innovations. The team is fast in introducing new
product/service innovations.
Improvisation The team deals with unanticipated events on the spot. Team members think on their feet when carrying out
actions. The team responds in the moment to unexpected problems. The team tries new approaches
to problems. The team identifies opportunities for new work processes. The team takes risks in terms of
producing new ideas in doing its job. The team demonstrates originality in its work.
Expertise Team members are aware of the critical issues that affect their work. Team members are current and
knowledgeable about their field of work. Team members have knowledge in diverse fields.
Teamwork quality Team members accept ideas coming from others and build on them. Team members trust each other. Team
members encourage different points of view. Team members are alert to the opportunities presented
by the situation and help to move the action forward. Team members take the lead at different times
depending on the needs of the situation.
Experimental culture In the team, errors are considered a source of learning. In the team, there is room for initiative. In the team,
there is freedom to experiment. In the team, we are encouraged to take risks when trying new ideas.
Real-time information
and communication
In the team, our meetings are a good source of up-to-date information. Information about what is going
on within the organization is readily shared at all levels. In the team, we regularly receive information
about other departments’ activities. In the team, we have the necessary information about our external
Memory The team has well-defined procedures. The team keeps records of past projects. The team has information
systems to support the work. The team has files and databases that are up to date.
Job characteristics The job requires team members to deal with unexpected events. The job requires team members to
come up with new ways of doing things. The job requires team members to deal with ambiguous
assignments, for which no previously established procedures exist. The job requires team members
to perform repetitive duties. The job requires team members to perform structured tasks (the steps to
perform the task are clear).
Note. In the scales for individual innovation and improvisation, “the team” was replaced with “I.”
Agrell, A., R. Gustafson. 1996. Innovation and creativity in work
groups. M. A. West, ed. Handbook of Work Group Psychology.
Wiley & Sons, Chichester, UK, 317–343.
Aiken, L. S., S. G. West. 1991. Multiple Regression: Testing and
Interpreting Interactions. Sage, Newbury Park, CA.
Amabile, T. M. 1988. A model of creativity and innovation in orga-
nizations. Res. Organ. Behavior 10 123–167.
Amabile, T. M. 1996. Creativity in Context. Westview Press, Boulder,
Barrett, F. 1998. Creativity and improvisation in jazz and organiza-
tions: Implications for organizational learning. Organ. Sci. 9(5)
Bentler, P. 1990. Comparative fit indexes in structural models. Psych.
Bull. 107 238–246.
Bontis, N., M. Crossan, J. Hulland. 2002. Managing an organizational
learning system by aligning stocks and flows. J. Management
Stud. 39(4) 437–469.
Brown, S. L., K. M. Eisenhardt. 1998. Competing on the Edge:
Strategy as Structured Chaos. Harvard Business School Press,
Boston, MA.
Caldwell, D. F., C. A. O’Reilly. 2003. The determinants of team-based
innovation in organizations. Small Group Res. 34 497–517.
Cannon-Bowers, J. A., E. Salas. 1998. Making Decisions Under
Stress: Implications for Individual and Team Training. American
Psychological Association, Washington, D.C.
Cannon-Bowers, J. A., E. Salas, S. Converse. 1993. Shared mental
models in expert team decision making. N. J. Castellan, ed. Indi-
vidual and Group Decision Making. Lawrence Erlbaum Asso-
ciates, Hillsdale, NJ, 221–246.
Chow, G. C. 1960. Tests of equality between sets of coefficients in
two linear regressions. Econometrica 28(3) 591–605.
Claxton, G. 1984. Live and Learn: An Introduction to the Psychol-
ogy of Growth and Change in Everyday Life. Harper and Row
Publishers, London, UK.
Cohen, S. G., D. E. Bailey. 1997. What makes teams work: Group
effectiveness research from the shop floor to the executive suite.
J. Management 23(3) 239–290.
Cohen, J., P. Cohen, S. G. West, L. S. Aiken. 2003. Applied Multi-
ple Regression/Correlation Analysis for the Behavioral Sciences.
Erlbaum, Mahwah, NJ.
Cook, S. D. N., D. Yanow. 1993. Culture and organizational learning.
J. Management Inquiry 2(4) 373–390.
Crossan, M. 1997a. Improvise to innovate. Ivey Bus. J. 62(1) 36–42.
Crossan, M. 1997b. Improvise to Innovate Training Program. Ivey
Business School, London, Ontario, Canada.
Crossan, M. 1998. Improvisation in action. Organ. Sci. 9(5) 593–599.
Crossan, M., D. Hurst. 2003. Strategic renewal as improvisation: Rec-
onciling the tension between exploration and exploitation. Work-
ing paper, Ivey Business School, London, Ontario, Canada.
Crossan, M., M. Sorrenti. 1997. Making sense of improvisation.
J. Walsh, A. Huff, eds. Advances in Strategic Management,
Vol. 14. JAI Press, Greenwich, CT, 155–180.
Crossan, M., H. Lane, R. White. 1999. An organizational learning
framework: From intuition to institution. Acad. Management
Rev. 24(3) 522–538.
Crossan, M., M. Cunha, D. Vera, J. Cunha. 2005. Time and organi-
zational improvisation. Acad. Management Rev. 30(1) 129–145.
Vera and Crossan: Improvisation and Innovative Performance in Teams
Organization Science 16(3), pp. 203–224, © 2005 INFORMS 223
Cunha, M., J. Cunha, K. Kamoche. 1999. Organizational improvisa-
tion: What, when, how, and why. Internat. J. Management 1(3)
De Bono, E. 1973. Lateral Thinking: Creativity Step By Step. Harper
and Row, New York.
Drach-Zahavy, A., E. Somech. 2001. Understanding team innova-
tion: The role of team processes and structure. Group Dynam. 5
Drazin, R., M. A. Glynn, R. K. Kazanjian. 1999. Multilevel theorizing
about creativity in organizations: A sensemaking perspective.
Acad. Management Rev. 24(2) 286–307.
Edmonson, A. 1999. Psychological safety and learning behavior in
work teams. Admin. Sci. Quart. 44 350–383.
Eisenhardt, K. M. 1989. Making fast strategic decisions in high-
velocity environments. Acad. Management J. 32(3) 543–576.
Eisenhardt, K. M., B. N. Tabrizi. 1995. Accelerating adaptive pro-
cesses: Product innovation in the global computer industry.
Admin. Sci. Quart. 40(1) 84–110.
Faraj, S., L. Sproull. 2000. Coordinating expertise in software devel-
opment teams. Management Sci. 46 1554–1568.
Frost, A., R. Yarrow. 1990. Improvisation in Drama. St. Martin’s
Press, New York.
Fry, A. 1987. The Post-It Note: An intrapreneurial success. SAM Adv.
Management J. 52 4–9.
Gardner, H. 1982. Art, Mind, and Brain: A Cognitive Approach to
Creativity. Basic Books, New York.
Gerbing, D. W., J. C. Anderson. 1988. An updated paradigm for scale
development incorporating unidimensionality and its assess-
ment. J. Marketing Res. 25 186–192.
Gilson, L. L., C. E. Shalley. 2004. A little creativity goes a long way:
An examination of teams’ engagement in creative processes.
J. Management 30 453–470.
Gregory, K. L. 1983. Native-view paradigms: Multiple cultures
and culture conflicts in organizations. Admin. Sci. Quart. 28
Hackman, J. R., C. G. Morris. 1975. Group tasks, group interaction
process, and group performance effectiveness: A review and pro-
posed integration. L. Berkowitz, ed. Advances in Experimental
Social Psychology, Vol. 8. Academic Press, New York, 45–99.
Halpern, C., D. Close, K. Johnson. 1994. Truth in Comedy. Meri-
wether Publishing Ltd., Colorado Springs, CO.
Hanlon, S. C., R. R. Taylor. 1991. An examination of changes in
work group communication behaviors following installation of
a gain-sharing plan. Group Organ. Stud. 16(3) 238–267.
Hatch, M. J. 1997. Jazzing up the theory of organizational improvisa-
tion. J. Walsh, A. Huff, eds. Advances in Strategic Management,
Vol. 14. JAI Press, Greenwich, CT, 181–191.
Hatch, M. J. 1998. Jazz as a metaphor for organizing in the 21st cen-
tury. Organ. Sci. 9556–557.
Hollingshead, A. B. 1998. Group and individual training: The impact
of practice on performance. Small Group Res. 29 254–280.
Hutchins, E. 1991. Organizing work by adaptation. Organ. Sci. 2
James, L. R., R. G. Demaree, G. Wolf. 1984. From micro to
meso: Critical steps in conceptualizing and conducting multi-
level research. Organ. Res. Methods 3(3) 211–236.
Johnstone, K. 1979. Improv: Improvisation and the Theatre. Theatre
Arts Books, New York.
Jöreskog, K. G., Sörbom. 1986. Lisrel VI: Analysis of Linear Struc-
tural Relationships by Maximum Likelihood and Least Square
Methods. Scientific Software, Inc., Mooresville, IN.
Kamoche, K., M. Cunha. 2001. Minimal structures: From jazz impro-
visation to product innovation. Organ. Stud. 22(5) 733–764.
Kamoche, K., M. Cunha, J. Cunha. 2003. Towards a theory of orga-
nizational improvisation: Looking beyond the jazz metaphor.
J. Management Stud. 40 2023–2051.
Kunda, G. 1992. Engineering Culture. Temple University Press,
Philadelphia, PA.
Lawrence, K. 2001. Playing by the rules: A role for improvisa-
tion in groups. Working paper, University of Michigan Business
School, Ann Arbor, MI.
Liang, D. W., R. Moreland, L. Argote. 1995. Group versus individual
training and group performance: The mediating role of transac-
tive memory. Personality Soc. Psych. Bull. 21 384–393.
Lindsay, R. M. 1993. Incorporating statistical power into the test of
significance procedure: A methodological and empirical inquiry.
Behavioral Res. Accounting 5211–236.
Lipsky, M. 1980. Street-Level Bureaucracy. Russel Sage, New York.
March, J. G. 1991. Exploration and exploitation in organizational
learning. Organ. Sci. 271–87.
Martin, J., C. Siehl. 1983. Organizational culture and counterculture:
An uneasy symbiosis. Organ. Dynam. 12(2) 52–65.
Maynard-Moody, S., M. Musheno. 2003. Cops, Teachers, Counselors:
Stories from the Front Lines of Public Service. University of
Michigan Press, Ann Arbor, MI.
Meyer, A. 1998. Organizing for improvisation: The backstage story
of the Vancouver jazz concert and symposium. Organ. Sci. 18
Miner, A., P. Bassoff, C. Moorman. 2001. Organizational impro-
visation and learning: A field study. Admin. Sci. Quart. 46
Mirvis, P. H. 1998. Practice improvisation. Organ. Sci. 9586–592.
Mohammed, S., B. C. Dumville. 2001. Team mental models in a
team knowledge framework: Expanding theory and measure-
ment across disciplinary boundaries. J. Organ. Behavior 22
Moon, J., P. DeLeon. 2001. Municipal reinvention: Managerial val-
ues and diffusion among municipalities. J. Public Admin. Res.
Theory 11(3) 327–351.
Moorman, C., A. Miner. 1997. The impact of organizational memory
on new product performance and creativity. J. Marketing Res.
34 91–106.
Moorman, C., A. Miner. 1998a. Organizational improvisation
and organizational memory. Acad. Management Rev. 23(4)
Moorman, C., A. Miner. 1998b. The convergence of planning and exe-
cution: Improvisation in new product development. J. Marketing
62(3) 1–20.
Moreland, R. A., L. Myaskovsky. 2000. Exploring the performance
benefits of group training: Transactive memory or improved
communication? Organ. Behavior Human Decision Processes
82 117–133.
Orlikowski, W. J. 1996. Improvising organizational transformation
over time: A situated change perspective. Inform. Systems Res.
Pascale, R. T. 1984. Perspectives on strategy: The real story behind
Honda’s success. California Management Rev. 26(3) 47–72.
Vera and Crossan: Improvisation and Innovative Performance in Teams
224 Organization Science 16(3), pp. 203–224, © 2005 INFORMS
Poolton, J., H. Ismail. 2000. New developments in innovation. J. Man-
agerial Psych. 13 795–811.
Pressman, J. L., A. Wildavsky. 1973. Implementation. University of
California Press, Berkeley, CA.
Preston, A. 1991. Improvising order. I. L. Mangham, ed. Organiza-
tion Analysis and Development, Vol. 22. John Wiley & Sons,
New York, 308–325.
Rosenberg, M. 1998. The Flexible Thinker. Orange You Glad Inc.,
Brampton, Canada.
Roth, A. 1993. Performance dimensions in services: An empiri-
cal investigation of strategic performance. T. A. Swartz, D. E.
Bowen, S. W. Brown, eds. Advances in Services Marketing and
Management, Vol. 2. JAI Press Inc., Greenwich, CT, 1–47.
Rousseau, D., S. Sitkin, R. Burt, C. Camerer. 1998. Not so different
after all: A cross-discipline view of trust. Acad. Management
Rev. 23(3) 393–404.
Sawyer, K. 1999. The emergence of creativity. Philos. Psych. 12(4)
Sawyer, K. 2000. Improvisation and the creative process: Dewey,
Collingwood, and the aesthetics of spontaneity. J. Aesthetics Art
Criticism 58(2) 149–161.
Schein, E. H. 1992. Organizational Culture and Leadership. Jossey-
Bass, San Francisco, CA.
Seham, A. 2001. Whose Improv is it Anyway? Beyond Second City.
University Press of Mississippi, Jackson, MS.
Shalley, C. E., J. Zhou, G. R. Oldham. 2004. The effects of personal
and contextual characteristics on creativity: Where should we go
from here? J. Management 30 933–958.
Spolin, V. 1963. Improvisation for the Theatre. Northwestern Univer-
sity Press, Evanston, IL.
Sternberg, R. 1999. Handbook of Creativity. Cambridge University
Press, Cambridge, UK.
Tierney, P., S. M. Farmer, G. B. Graen. 1999. An examination of
leadership and employee creativity: The relevance of traits and
relationships. Personnel Psych. 52(3) 591–620.
Tucker, L., C. Lewis. 1973. The reliability coefficient for maximum
likelihood factor analysis. Psychometrika 38 1–10.
Unger, L. S., J. B. Kernan. 1983. On the meaning of leisure: An
investigation of some determinants of the subjective experience.
J. Consumer Res. 9(March) 381–392.
Van Wart, M., E. Berman. 1999. Contemporary public sector pro-
ductivity values. Public Productivity Management Rev. 22(3)
Vera, D., M. Crossan. 2004. Theatrical improvisation: Lessons for
organizations. Organ. Stud. 25(5) 727–749.
Walsh, J. P., G. Rivera. 1991. Organizational memory. Acad. Manage-
ment Rev. 16(1) 57–91.
Wegner, D. J. 1987. Transactive memory: A contemporary analysis
of the group mind. B. Mullen, G. R. Goethals, eds. Theories of
Group Behavior. Springer-Verlag, New York, 185–208.
Weick, K. E. 1993. The collapse of sensemaking in organizations:
The Mann Gulch disaster. Admin. Sci. Quart. 38(4) 628–652.
Weick, K. E. 1998. Improvisation as a mindset for organizational
analysis. Organ. Sci. 9(5) 543–555.
Weick, K. E., K. H. Roberts. 1993. Collective mind in organizations:
Heedful interrelating on flight decks. Admin. Sci. Quart. 38 357–
West, M. A. 1990. The social psychology of innovation in groups.
M. A. West, J. L. Farr, eds. Innovation and Creativity at Work:
Psychological and Organizational Strategies. Wiley, Chichester,
UK, 309–333.
West, M. A. 1994. Effective Teamwork. British Psychological Society,
London, UK.
Zack, M. H. 2000. Jazz improvisation and organizing: Once more
from the top. Organ. Sci. 11(2) 227–234.
... However, the scientific works of the 21 st century (Hadida et al., 2015;Fisher and Barrett, 2019) speak of organisational improvisation as an inseparable part of management research. Organisational improvisation has been studied by Brown and Eisenhardt (1997), Barrett (1998), Weick (1998), Dennis and Macaulay (2003), Vera and Crossan (2005), Dennis and Macaulay (2007), Johnson (2014) and other scholars. Organisational improvisation is analysed as an inseparable part of innovation. ...
... In today's world, many organisations should forget stability and start looking for new solutions and "jazzing" (Kuura and Sandoval, 2019); however, detailed plans are still required followed by complex bureaucratic procedures, and improvisation is still seen as a sign of failure and a temporary "fire-extinguishing" measure as well as a risky activity that should be avoided (Alterhaug, 2004). Author, year Improvisation described as Brown and Eisenhardt, 1997 Product development adapting to ever-changing markets and technologies Barrett, 1998 Analysis, constant experimentation, consideration of opportunities without knowing how an action will occur Weick, 1998 Combination between created in advance and spontaneous, control and novelty Dennis and Macaulay, 2003 Creative action that changes the nature of a plan Vera and Crossan, 2005 Creative and spontaneous process that helps to achieve a goal in a new way Dennis and Macaulay, 2007 An opportunity to create solutions based on the main values of an organisation rather than following a certain pre-determined order or a strict plan Johnson, 2014 Innovation, learning and fast response Many leaders think that improvisation is an abominable rather than valuable process even if it leads to success (Fisher and Barrett, 2019). It is no surprise that such underestimation of improvisation has inspired many scientific studies related to this topic. ...
... Creative and spontaneous process that helps to achieve a goal in a new way (Vera and Crossan, 2005) Jazz Academy is established on the first floor of a private house in the Šilainiai district that has nothing to do with jazz. This is a venture calculated by two cold yet creative minds (Kaunas pilnas kultūros, 2018) Analysis, constant experimentation, consideration of opportunities without knowing how an action will occur ( Barrett, 1998) Selection is carried out through interviews, the focus is on students' motivation (Kaunas pilnas kultūros, 2018) Innovation, learning and fast response (Johnson, 2014) The work is not based on private lessons, "As an educator, I have always tried to avoid private lessons. ...
Full-text available
The article discusses the concept of organisational improvisation and reveals why it is important for contemporary organisations. Organisational improvisation is more and more acknowledged as a relevant field of management research; however, heads of most organisations still believe that detailed plans accompanied by various bureaucratic procedures are important, and that improvisation is a sign of failure, is risky and is to be avoided. The article discusses the three levels of improvisation (individual, interpersonal and organisational) pointing out its possibilities and advantages. Peculiarities of organisational improvisation are provided along with the results of a case study of public institution Jazz Academy. In this way, this article is the first study of organisational improvisation in Lithuania. The aim of the article is to reveal the peculiarities of organisational improvisation.
... Therefore, improvisation is particularly important in the face of uncertainty, and it has become a way to cope with change and uncertainty [8]. According to Vera and Crossan, improvisational behavior is a new approach to solving objective problems with spontaneity and creativity [9], which is important for any individual [10] and any organization [8]. The related study noted that different organizations have responded differently [7], health systems and organizations need to "proactively adapt and recover from the pandemic" [7], and suggest related measures such as wearing masks and keeping safe distances [4]. ...
... Related studies have shown that there are individual differences in improvisation development [15], as the process is complex and difficult [10]. When faced with the uncertainty of COVID-19, we know what to do and how to respond at the organizational level, but it is not yet clear how individuals at the micro level develop improvisation skills or generate improvisational behavior [9]. ...
... Cunha argues that there is no good or bad improvisation and that improvisation is a continuous process [8], from low-level interpretation to moderate modification to full improvisation [30]. Scholars have used the concept of improvisation in a variety of situations at the organizational level, including product innovation [12] and new product development [9]. Organizational improvisation starts from the core definition of improvisation, which is "the deliberate integration of designing and executing new work" [12], and it is a process that occurs in teams or throughout the organization [8,12], enabling managers to continuously and creatively adapt to change and continually expand their products and services outward [9]. ...
Full-text available
Organizations and individuals are unprepared for an unexpected outbreak of COVID-19. While most of the literature focuses on improvised reactions at the organizational level, this paper focuses on understanding improvised reactions at the individual level. This paper draws on previous research applying improvisation to the field of consumer behavior and introduces consumer knowledge acquisition as a mediating variable and tightness-looseness culture as a moderating variable from the perspective of mixed emotions of awe and anxiety to explain the mechanism of consumers with mixed emotions of awe and anxiety on improvisation behavior based on the environment of a COVID-19 outbreak. Data from 330 participants in Study 1 examined the effect of mixed emotions of awe and anxiety on improvisation behavior through knowledge acquisition, and data from 434 participants in Study 2 examined the moderating effect of relaxed culture. The findings suggest that consumers with mixed emotions report a higher willingness to acquire knowledge and report higher levels of improvisational behavior. Consumers behaved differently in different environments. Consumers with mixed emotions responded more strongly to improvisation in the loose-culture environment than in the tight-culture environment, and the mixed emotions of awe and anxiety had a positive effect on individual consumers’ improvisational behavior through the mediating role of knowledge acquisition.
... When inevitable mishaps occur, they are applauded and embraced instead of shamed and hidden. The purpose of these exercises is to desensitise individuals to the fear of failure and to reframe mistakes merely as events, whereby something happens other than what was anticipated or expected (Barrett, 1998;Vera & Crossan, 2005). The emotional connotation of mistakes shifts from negative to positive, 14 and becomes crystallised as 'a mistake is a gift', a common motto among improvisers (Weis & Arnesen, 2014). ...
... Indeed, improvisation encourages risk-taking and engaging in an intuitive, reciprocal and creative process. This process is supported by psychological safety, where making mistakes is not punished and experimentation is allowed (Vera & Crossan, 2005). The change in creativity has been assessed with a variety of alternate use tasks (Guilford, 1967), whereby participants are asked to list as many uses for a common object as possible within a limited time, or through creative verbal or problem-solving tasks. ...
Full-text available
Improvisation is commonly understood as a performance or creating something without preparation. As an art form, improvised theatrical plays are created spontaneously on stage without a script. As an applied form of theatre, improvisation has been utilised in fields requiring collaboration and a tolerance for uncertainty, such as in the business and education sectors. This dissertation contributes to the literature in educational research by investigating applied improvisation as a tool to promote student teachers’ interpersonal competence. Applied improvisation enables individuals to explore and practise teaching-related encounters in a fictional and psychologically safe context. Psychological safety is particularly important when practising challenging interactions. Despite the fictionality of the context, bodily experiences during improvisations may promote experiential learning. The research summarised in this dissertation was guided by two primary research questions. First, I asked whether improvisation training influenced student teachers’ interpersonal competence and social stress. Student teachers (n = 19) participated in a 7-week (17.5-h) improvisation intervention, comprising the fundamentals of theatre improvisation and status expression (verbal and nonverbal behaviours indicating the social dominance of a person). The impact of the intervention was measured using subjective self-reports (interpersonal confidence, i.e., belief regarding one’s capability related to effective social interactions, self-esteem and experienced stress) and a large array of physiological measurements (heart rate, heart rate variability, skin conductance, facial muscle activity, frontal electroencephalogram (EEG) alpha asymmetry and stress hormone cortisol). Self-reports, physiological measurements and Trier Social Stress Tests (TSST; including public speaking) were performed before and after the improvisation intervention. An improvisation course was arranged for the control group (n = 20) following the intervention study. One year later, the long-term effects of improvisation training on self-reported interpersonal confidence were measured in a follow-up study. Second, I asked how real versus fictional social rejections impact experienced stress and psychophysiological responses. Student teachers (n = 39) participated in an experiment including both real (interview) and fictional (improvisation exercises) dyadic interactions. In the real condition, student teachers were unaware that the interviewer was an actor trained to include subtle social rejections during the interview by using three types of social rejections: devaluing, interrupting and nonverbal rejections. In the fictional condition, student teachers were informed in advance which social rejection type would be used during a later improvisation exercise. Experienced stress and psychophysiological reactivity during social rejections were measured under both experimental conditions. Following an improvisation intervention, interpersonal confidence and its components of performance confidence and a tolerance for failure increased relative to controls, whilst one year later the improved performance confidence persisted. Furthermore, a heterogeneous treatment effect was found. Those with the lowest pretest interpersonal confidence score benefited most from the improvisation intervention. No between-group differences in self-esteem were observed. Psychological and physiological indications of relief from performance-related stress were also observed following improvisation training. In addition, interpersonal confidence moderated self-reported and cardiovascular stress responses. Thus, interpersonal confidence may be worth controlling for in future research which examines the effects of interventions aimed at relieving social stress. The results also support the notion that repetition may also diminish performance-related stress, since the control group exhibited decreases in cardiovascular stress during some of the test conditions. The primary finding regarding the second research question emerged through the absence of any systematic attenuation of the psychophysiological reactivity to fictional versus real-world social rejections. In other words, although student teachers knew that improvised social rejections were fictional, their psychophysiological responses during improvisation remained relatively similar and associated with those of real-world rejections. It appears as though personal relevance and engagement during improvisation explain the relatively similar bodily responses. This result suggests that interpersonal encounters can be realistically modelled through applied improvisation. In this dissertation research, I also produced a validated self-report measure, the Interpersonal Confidence Questionnaire (ICQ), to evaluate the impact of social interaction training relying on applied improvisation. Using an additional dataset (n = 208), I validated the questionnaire and examined the impact of improvisation training on a larger sample. A confirmatory factor analysis identified six factors—performance confidence, flexibility, listening skills, a tolerance for failure, collaboration motivation and presence—that contribute to interpersonal confidence. Thus, the ICQ appeared valid and reliable as a self-report measure of interpersonal confidence. In summary, the findings from this research indicate that a relatively brief improvisation intervention promotes interpersonal confidence, specifically amongst those with low interpersonal confidence. Furthermore, improvisation training serves as an intervention against performance anxiety and generates long-term improvements to performance confidence. This dissertation provides a theoretical framework and empirical support for the application of improvisation as a tool to develop interpersonal competence skills, particularly within professions requiring face-to-face interactions. Regardless of the fictionality of the improvisational context, genuine emotions and experiences may emerge, serving as experiential learning experiences. The significance of these findings may extend to theatre-based practices and drama education in general, which rely on holistic action and personal engagement in fictional contexts. The findings agree with previous research, suggesting that including the improvisation method in teacher education curricula can enhance student teachers’ interpersonal competence as well as their skills related to sensitive and responsive teaching. Finally, this dissertation contributes to social neuroscience by recommending an ecologically valid experimental design wherein naturally unfolding social interactions can be achieved using improvisation techniques. ________________________________________ Keywords: experiential learning, fictionality, improvisation, interpersonal confidence, intervention, psychophysiology, social interaction, social rejection, social stress, teacher education, theatre-based practices
... -The moderating variable: The researchers have relied on the following references: (Vera & Crossan, 2005;Pavlou & El Sawy, 2010;e Cunhaa et al., 2020;Hains-Wesson et al., 2017;XUE & SUN, 2019;Ladd, 2016). ...
Full-text available
Many studies across different eras have dealt with the topic of leadership and from different aspects, despite the large number of researches over time, management researchers have recently begun to define modern concepts of leadership, this study aims to review the theoretical literature on identifying Authentic Leadership “AL” represented by (Self-Awareness “SA”, Balanced Processing “BP”, Internalized Moral and Ethical Perspectives “IMEP” and Relational Transparency “RT”) and its impact on the Smart Organization “SO” represented by (Understanding the Environment “UE”, Continuous Learning “CL”, Resources Mobilization “RM”, and Finding Strategic Alternatives “FSA”) and to examine the Mediating role of Strategic Ambidexterity “SAM” and the moderating role of Improvisational Capabilities “IC”.
... The improvisation technique is a unique technique where there is nothing wrong and right in it, because improvisation is a process for students to increase self-confidence in presenting ideas, concepts, principles, and theories that form-critical, analytical, logical and rational thinking (Romanelli & Tishby, 2019). Moreover, Barrett (1998); Vera and Crossan (2005) mentioned that the abundantly cultivated slogan "a mistake is a gift" among the practitioners of improvisation, means that the surprise element of mistakes may also offer imaginative turns, thus modulating mistakes into "gifts". This means students need to build courage, confidence, motivation on social skills to make learning active. ...
... Whereas narrowly defined and ponderous structures can stall spontaneity and creativity as central aspects of improvisation, the lack of minimal structures can lead to organizational chaos and inefficiency (Brown and Eisenhardt 1997). Minimal structures hence serve as the boundaries within which improvisation happens and represent the elements on which improvisation is built (Barrett 1998;Kamoche and Cunha 2001;Vera and Crossan 2005). Although this approach bears many advantages, such as a timely response by empowered employees under time pressure, it is not without downsides. ...
Conference Paper
Full-text available
While current research on digital transformation strategy (DTS) promotes the usefulness of planning rationality, the role of serendipity and its impact on the DTS formation has been neglected so far. Given the environmental volatility and disruptions executives face today, we contend that reconfiguring through planning alone is impossible. Instead, as organizations encounter unexpected occurrences, they need to improvise with the means immediately at hand. In our study, we explore the intersection of DTS formation and improvisation. Drawing on Mintzberg and Waters' (1985) strategy formation concept, we present a typology of three DTS formation approaches and analyze their viability for satisfying the demand for organizational improvisation in the DTS formation context. Our study contributes to the nascent discussion on strategic improvisation by shedding light on a crucial but latent layer of strategy-making that provides untapped insights for better understanding the mechanisms affecting DTS formation.
... This leadership generates greater awareness and acceptance of the organisation's goal and mission. It promotes a common vision, by reorienting the training and the work-teams construction (Kark et al., 2003;Vera and Crossan, 2005;Imran et al., 2016). It creates a persuasive vision, offers clear goals, provides support and encourages change (Chan and Mak, 2014). ...
Transformational leadership is an important factor affecting organisational performance. Many studies have shown that transformational leadership has positive and significant influence on organisational performance. Based on a literature review and previous work, this study aims to investigate the impact of transformational leadership on organisational performance and to examine whether organisational learning and open innovation is a mediator between their relationships. A quantitative analysis is performed via a questionnaire. The data analysis was conducted thanks to the structural equation modelling method. The research sample consisted of 202 respondents from Tunisian enterprises chosen through non-probabilistic convenience sampling. The findings of this study provide evidence that transformational leadership, organisational learning and open innovation have significant positive relationship influence on organisational performance. The research also demonstrates that there is a significant effect on the role of mediation in organisational learning and open innovation on the relationship between transformational leadership and organisational performance. The study suggests that if organisation principals use the strategies of transformational leadership and organisational learning at the same time, organisational learning was highly effective to achieve open innovation and subsequently a performance within the organisation.
Ever-changing consumer preferences have urged hotel properties to continuously innovate to ensure business sustainability. Particularly in the sales context, the manifestation of this threat is due to the lack of a single optimum approach to selling. Thus, being creative and generating new ideas is imperative for good salespeople to attract customers. To address this lack of an existing componential model of creativity and to uncover the hidden cognitive mechanism underlined with this theory, this study investigated the direct relationship as well as the mediating effect of employee mindfulness on the relationship between transformational leadership and creativity of salespeople. In addition, this study examined the moderating effect of improvization on the relationship between mindfulness and employee creativity. Data collected from 275 salespeople working in the hospitality industry in the Kingdom of Bahrain revealed a significant positive association between transformational leadership and employee creativity. Furthermore, the findings show that employee mindfulness plays a salient role as a mediator in this indirect relationship. However, the moderating effect of improvization on the relationship between employee mindfulness and creativity was found to be insignificant.
Purpose This study aims to investigate the causal relationships within an experimental culture of improvisation capability and firm performance, following the logic of “culture-capability-performance” and building on informal institution theory and dynamic capability theory. Design/methodology/approach Data was mainly collected via on-site questionnaires from firm managers, and 196 valid questionnaires were analyzed using structural equation modeling to test the relationship among experimental culture, improvisation capability and firms’ performance. Findings Trust and support had a positive impact on firm spontaneity, while the effect of action promotion and error tolerance was not significant. Action promotion, trust and support demonstrate substantial positive effects on the creativity of a firm. Both dimensions of improvisation capability positively and significantly influence a firm’s performance. Research limitations/implications The research respondents were firm managers. Cross-sectional data were used to analyze the model, which may cause common method variance. The research context was limited to China, and the generalizability of the results needs to be considered. Practical implications Firms can cultivate a culture of trust and support to enhance their spontaneity and improvisation capability. They can encourage cross-departmental communication, empower employees with autonomy in decision-making, provide appropriate resource support for employees’ decisions and use praise and reward incentives to spur further innovation achievements. Originality/value This study addresses the gaps in a firm’s improvisation capability within a Chinese market context by theoretically and empirically examining the role of experimental culture and assessing the relationship among each of the dimensions of improvisation capability in relation to firm performance identified in this study.
Organisational improvisation (OI), the central subject of this study, is configured as an important resource for companies when faced with unexpected and recurrent events that are typical of times of crisis. The aim of this research is to analyze the phenomenon of OI in terms of the heuristics used for this purpose, as well as the outcomes, in the context of small businesses. The grounded theory strategy was used to analyze the information obtained through narrative interviews conducted with 23 owners of small businesses. Forty-one episodes of OI were analyzed using the content analysis technique. Five categories of outcomes of OI episodes were identified. The application of seven creative tactics associated with the OI episodes in question was also evident: adaptation, copy, exaptation, frugal innovation, improvement, new combination, and nonaptation spandrels. In terms of results, emphasis is placed on the analysis of two episodes of OI associated with the category of Strategic Repositioning of the organisation, a phenomenon that was hitherto identified in the literature as rare but possible. As for the seven creative tactics that were identified, they provided evidence that the Adaptation construct used in some OI frameworks is incomplete and restrictive in terms of possible actions.
Full-text available
The field of marketing strategy often makes the important assumption that marketing strategy should occur by first composing a plan on the basis of a careful review of environmental and firm information and then executing that plan. However, there are cases when the composition and execution of an action converge in time so that, in the limit, they occur simultaneously. The authors define such a convergence as improvisation and develop hypotheses to investigate the conditions in which improvisation is likely to occur and be effective. The authors test these hypotheses in a longitudinal study of new product development activities. Results show that organizational improvisation occurs moderately in organizations and that organizational memory level decreases and environmental turbulence level increases the incidence of improvisation. Results support traditional concerns that improvisation can reduce new product effectiveness but also indicate that environmental and organizational factors can reduce negative effects and sometimes create a positive effect for improvisation. These results suggest that, in some contexts, improvisation may be not only what organizations actually practice but also what they should practice to flourish.
Introduction - Part 1 WHO? Major Practitioners of Improvisation - Improvisation in Traditional Drama - Improvisation in Alternative Drama - Beyond Drama 'Paratheatre' - Part 2 WHAT? The Practice of Improvisation - Improvisation Exercises Preparation - Working Together - Moving Towards Performance - Applied Improvisation Work - Part 3 WHY? The Meanings of Improvisation - Towards the Poetics - Enriching the Communication of Meaning - Meaning and Performance - Notes - Bibliography - Index
This wide-ranging edited volume provides a state of the art account of theory and research on modern street-level bureaucracy, gathering internationally acclaimed scholars to address the varying roles of public officials who fulfill their tasks while interacting with the public. These roles include the delivery of benefits and services, the regulation of social and economic behavior, and the expression and maintenance of public values. Questions about the extent of discretionary autonomy and the feasibility of hierarchical control are discussed in depth, with suggestions made for the further development of research in this field. Hence the book fills an important gap in the literature on public policy delivery, making it a valuable text for students and researchers of public policy, public administration and public management.