Toward a Better Understanding of Crowdfunding, Openness and the Consequences for Innovation

Abstract

Crowdfunding is now a commonly used tool for innovating entrepreneurs, yet many unresolved questions surrounding crowdfunding's effect on innovation remain. Often, crowdfunding backers play an active role in the innovation conversation. Thus, crowdfunding can be viewed as one form of open search (actively seeking out ideas from outsiders). Beyond open search, backers also generate word of mouth awareness for the crowdfunded product. Crowdfunding backers can be thought of as the earliest possible adopters, who may be even more valuable than traditional early adopting consumers. In this study, data pertaining to crowdfunded products from the Kickstarter platform is coupled with survey data from the respective innovating entrepreneurs to better understand the effects of elements of crowdfunding on the subsequent market success of the crowdfunded product as well as the innovation focus of the crowdfunding organization. Results indicate that the amount of funding raised during a crowdfunding campaign does not significantly impact the later market performance of the crowdfunded product, while the number of backers attracted to the campaign does. Open search depth (drawing intensely from external sources) enhances product market performance, while open search breadth (drawing from many external sources) induces a radical innovation focus. Interestingly, adverse effects from over-relying on external knowledge sources are not observed. The small size of the crowdfunding organizations in this study is seen as a boundary condition to previous findings of inverse U-shaped performance effects. Finally, the portion of product development complete when crowdfunding impacts the entrepreneurs' subsequent focus on radical innovation.

Full text

Toward a Better Understanding of Crowdfunding, Openness and the Consequences for Innovation
Michael A. Stanko *
Poole College of Management
North Carolina State University
Raleigh, NC 27695-7229
Phone: (919) 515.0372
mike_stanko@ncsu.edu
David H. Henard
Poole College of Management
North Carolina State University
Raleigh, NC 27695-7229
Phone: (919) 515.8945
dhhenard@ncsu.edu
* Corresponding author
RESPOL-D-15-00964
Forthcoming at Research Policy
February 2017

1
Toward a Better Understanding of Crowdfunding, Openness and the Consequences for Innovation
Abstract
Crowdfunding is now a commonly used tool for innovating entrepreneurs, yet many unresolved questions
surrounding crowdfunding’s effect on innovation remain. Often, crowdfunding backers play an active role
in the innovation conversation. Thus, crowdfunding can be viewed as one form of open search (actively
seeking out ideas from outsiders). Beyond open search, backers also generate word of mouth awareness
for the crowdfunded product. Crowdfunding backers can be thought of as the earliest possible adopters,
who may be even more valuable than traditional early adopting consumers. In this study, data pertaining
to crowdfunded products from the Kickstarter platform is coupled with survey data from the respective
innovating entrepreneurs to better understand the effects of elements of crowdfunding on the subsequent
market success of the crowdfunded product as well as the innovation focus of the crowdfunding
organization. Results indicate that the amount of funding raised during a crowdfunding campaign does not
significantly impact the later market performance of the crowdfunded product, while the number of
backers attracted to the campaign does. Open search depth (drawing intensely from external sources)
enhances product market performance, while open search breadth (drawing from many external sources)
induces a radical innovation focus. Interestingly, adverse effects from over-relying on external knowledge
sources are not observed. The small size of the crowdfunding organizations in this study is seen as a
boundary condition to previous findings of inverse U-shaped performance effects. Finally, the portion of
product development complete when crowdfunding impacts the entrepreneurs’ subsequent focus on
radical innovation.
Keywords
Crowdfunding; Open Innovation; Open Search; Innovation; Entrepreneurship

2
Toward a Better Understanding of Crowdfunding, Openness and the Consequences for Innovation
1. Introduction
Crowdfunding has quickly evolved into a commonly used vehicle to help innovating
entrepreneurs get products developed and is one of the ways that innovative, small organizations have
been able to access capital since the financial crisis (Lee et al., 2015). Crowdfunding campaigns
conducted via Kickstarter alone have raised over $2B in pledges from millions of backers since the
platform’s inception in 2009 (Kickstarter, 2015). While crowdfunding has generated both tremendous
public interest and financial backing, questions around crowdfunding and innovation are largely
unanswered, both with respect to how elements of crowdfunding impact the success of the crowdfunded
product once it is released to the market and with respect to how crowdfunding shapes the entrepreneurial
organization’s future innovation efforts. While it might be intuitive to look at crowdfunding as a primarily
financial exercise, we view the true value of crowdfunding as the ability to learn from backers and to use
them as ambassadors for the crowdfunded product. To the authors’ knowledge, this study is the first to
investigate the effects of elements of crowdfunding on subsequent innovation outcomes. While research
on crowdfunding has progressed in recent years (e.g., Agrawal et al., 2014, Belleflamme et al., 2014,
Calic and Mosakowski, 2016, Mollick, 2014), examining the subsequent outcomes from crowdfunding is
a necessary step to better understand crowdfunding’s influence on technological innovation.
Since crowdfunding backers often take an active role in the innovation conversation (Mollick,
2016, Stanko and Henard, 2016), we view crowdfunding as one element of open search (actively seeking
ideas from outsiders). Due to an inherent lack of resources, entrepreneurs and startups often turn to open
innovation tactics out of necessity as it can be overwhelming for them to try to conduct product
development activities in isolation (Henkel, 2006, Van de Vrande et al., 2009). Thus, we look to the
growing open innovation literature to develop our conceptual framework. However, there are other
benefits from crowdfunding (beyond open search) such as the word of mouth benefit that may come from
having a large group of crowdfunding backers. Given this, we take an inclusive perspective to

3
understanding the potential benefits and drawbacks of crowdfunding, rather than restricting ourselves
exclusively to open innovation based arguments.
Backers are central to understanding crowdfunding’s potential innovation effects. Crowdfunding
allows early stage entrepreneurs access to capital; yet, it also allows them to potentially engage with a
large number of individuals in ways that were previously unavailable. Crowdfunding backers not only
offer their money, but also their opinions. Backers often want to become engaged in product development
alongside the innovating entrepreneur, as that experience is typically considered by backers to be a
rewarding part of the process (Agrawal et al., 2014, Gerber et al., 2012). This injection of large numbers
of external voices into the product development process has the potential to dramatically impact
innovation efforts. On one hand, backers’ opinions could convince entrepreneurs to develop products
closer to what is currently visible in the marketplace, effectively discouraging risk taking and dampening
innovation. Conversely, backers could voice divergent or creative ideas that might lead to a heightened
focus on radical innovation.
Given this dynamic, open search – the term used to describe the process by which organizations
actively seek out ideas from outsiders – is thought to be a driver of key outcomes for crowdfunding
innovators. Organizations draw from external entities, such as customers, suppliers, consultants and
universities to aid their innovation efforts (Chesbrough, 2003). While previous research has focused on
the outcomes of open search (e.g., Lee et al., 2010, Love et al., 2011, West et al., 2014), this research has
not yet extended to the particular case of crowdfunding – despite an acknowledged need to better
understand the effects of open innovation on startups (Lee et al., 2010). Crowdfunding can be viewed as
an embodiment of the open innovation paradigm, given that backers often offer their personal opinions to
the innovating entrepreneurs. Building on prior research into the performance effects of open search (e.g.,
Laursen and Salter, 2006, Lee et al., 2010) and informed by knowledge creation theory (KCT; Grant,
1996, Nonaka et al., 2000), we examine how both open search depth (the number of information sources
used extensively for innovation) and open search breadth (the total number of information sources turned
to for innovation) impact the innovating organization. As have other studies focused on open search (e.g.,

4
Chiang and Hung, 2010, Laursen and Salter, 2006), we ground our arguments in the organizational
learning literature related to exploration and exploitation (see March, 1991). In some ways, the very
permeable membrane between backers and the innovating organization (for instance, backers have
worked side by side on product design with innovating entrepreneurs; Lewis-Kraus, 2015) makes
crowdfunding an ideal context to explore questions regarding open search’s effects on innovative
outcomes. Extant research (e.g., Belderbos et al., 2010, Laursen and Salter, 2006, Salge et al., 2013) has
found evidence of inverse U-shaped performance effects for larger organizations, whereby an over-
reliance on openness can have negative performance implications. We believe that startups and very small
businesses (common to crowdfunding initiatives) likely constitute a boundary condition to the noted
inverse U-shaped effect. For these small businesses with limited knowledge built up inside the
organization, we argue and subsequently demonstrate that there are not significant adverse performance
effects to over-relying on external knowledge sources.
Research into the innovation implications of crowdfunding is in its infancy (Agrawal et al.,
2014). The goal of this research is to broadly understand crowdfunding’s subsequent effects on both the
success of the crowdfunded product as well as on the entrepreneurial organization. Given this, our first
outcome of interest is product market performance (De Luca and Atuahene-Gima, 2007), defined as the
degree of financial success experienced once the crowdfunded product is launched. Further, we posit that
a firm’s interaction with backers and other innovative outsiders via open search activities has the potential
to impact their focus on radical innovation in future efforts (i.e., one to two years after the crowdfunding
campaign). As such, our second outcome of interest is the organization’s radical innovation focus
(McGrath, 2001), defined as the degree that the organization focuses on radical innovation in subsequent
efforts. Importantly, radical innovation focus concerns the product development activities that occur
within the organization, and not strategy, culture, goals, or aspirations regarding future radical innovation,
which are outside of this construct’s conceptual domain. A better understanding of the implications for
both these outcomes will prove valuable to crowdfunding entrepreneurs and can inform a broader policy

5
discussion focused on encouraging innovation through crowdfunding while ensuring protection for
various stakeholders.
In the following sections, we present a primer on crowdfunding and develop the conceptual
foundation for each of the variables of interest. Hypotheses are presented, followed by a discussion of our
methodological approach, which combines webscraped data on crowdfunded products from Kickstarter
with a later survey of the respective innovating entrepreneurs. The manuscript concludes with a
presentation of the results and implications for researchers, practitioners and policy makers. In summary,
we hypothesize and empirically show that while the number of backers attracted to a crowdfunding
campaign positively impacts the market success of the crowdfunded product, the amount of funding
raised does not have a significant effect. Backers have importance beyond their financial contributions.
Further, while open search breadth fosters a radical innovation focus, depth supports product market
performance. We also show that the effect of breadth on radical focus is contingent on the portion of
crowdfunding complete at the time of crowdfunding.
2. Crowdfunding and Innovation
Crowdfunding takes several forms, such as equity-based, reward-based, lending-based and
donation-based (Belleflamme et al., 2013). While lending-based crowdfunding generates substantial
funding globally (Massolution, 2015), equity and reward-based crowdfunding have garnered the most
interest with respect to their implications for innovation (e.g., Cholakova and Clarysse, 2015). Equity-
based crowdfunding, in which investors receive an ownership stake, has grown in recent years in part due
to the Jumpstart Our Business Startups (JOBS) Act. Equity-based crowdfunding websites (e.g.,
Crowdcube, CircleUp) are expanding, but remain relatively smaller (volume-wise) than reward-based
crowdfunding sites (e.g., Kickstarter, Indiegogo). Reward-based crowdfunding offers a relatively risk free
way for entrepreneurs and startups to generate new product awareness and gauge potential market
response. In reward-based crowdfunding, backers typically pre-order the product being developed (or a
different reward) in exchange for tiered levels of financial support. Generally, there is an expectation in

6
reward-based crowdfunding that backers will be exposed to the product development process through
ongoing updates and have the opportunity for direct communication with the innovating entrepreneurs
(Agrawal et al., 2014, Gerber et al., 2012). Thus, with reward-based crowdfunding, backers represent a
potential source of knowledge flow to innovating entrepreneurs that can be facilitated by open search
tactics. While all forms of crowdfunding may be relevant to innovating entrepreneurs and startups, the
focus of this research is on the relatively larger population of organizations conducting reward-based
crowdfunding to support product development, specifically in a technological innovation domain.
Several aspects of reward-based crowdfunding make it potentially appealing to innovating
entrepreneurs and startups. Reward-based crowdfunding platforms offer organizations – even startups and
very small businesses – a chance to reach a broad base of interested early adopters without having to part
with equity. These backers can prove to be important advocates as well as a valuable source of feedback
and ideas as the product progresses to market launch. Reward-based crowdfunding can also serve as a
market validation (Gerber et al., 2012), demonstrating the viability of a product concept, which might be
useful as sign of credibility should the innovating entrepreneur later pursue forms of equity investment
(e.g., venture capital).
Though research on crowdfunding is in its early stages (Bouncken et al., 2015), early work has
focused on addressing practical concerns for crowdfunding entrepreneurs. Some of the early
crowdfunding research understandably focused on the question of “what determines the likelihood of a
crowdfunding campaign achieving its funding goal?” Unsurprisingly, a social orientation,
communications tactics, and professionalism impact the propensity for campaigns to reach their funding
goal (Calic and Mosakowski, 2016, Müllerleile and Joenssen, 2015). Another practical question of
concern to crowdfunders is the effort that goes into managing a crowdfunding campaign. Crowdfunding
campaigns require more effort and a greater variety of skills than new crowdfunders expect (Hui et al.,
2012). The question of whether equity or reward-based crowdfunding is optimal for particular
circumstances is also quite important to crowdfunding entrepreneurs (Belleflamme et al., 2014). Some
researchers have examined backer behavior, such as the tendency for backers to support campaigns which

7
are geographically close to them (Lin and Viswanathan, 2015). Still other researchers have taken a
community perspective, examining the efforts of crowdfunding veterans to mentor new crowdfunders
(Hui et al., 2014). Despite progress across the crowdfunding literature, questions pertaining to
crowdfunding’s influence on innovation are yet to be well addressed. As such, many of the relationships
hypothesized here represent open theoretical questions.
2.1. Role of Backers
Backers are an important source of inbound open innovation for crowdfunding organizations.
Beyond their financial contribution, backers make an important contribution to knowledge creation
through their cooperation. In the parlance of KCT, crowdfunding organizations continuously articulate
and amplify ideas generated by individuals such as backers (McFadyen and Cannella, 2004, Nahapiet and
Ghoshal, 1998, Nonaka, 1994). Knowledge creation through interactions with backers (and other external
parties) is a key factor in understanding subsequent innovation. Specifically, backers are typically
engaged, early adopters who offer advice, design ideas and even criticism throughout the product
development process. In some cases, backers have been called upon to play an active role in making
design choices (Diallo, 2014, Lewis-Kraus, 2015).
While innovation research suggests that interactions with users during product development
should have a positive influence (e.g., Lilien et al., 2002, Spaeth et al., 2010), there is also the potential
for interactions with backers to be harmful over time (see Nahapiet and Ghoshal, 1998). It may be
overwhelming for small startups to try to interact with so many opinionated backers. Attempting to
incorporate their input may slow product development or divert the innovating entrepreneurs from their
intended priorities (relatedly, Faems et al., 2010 show that open innovation efforts significantly increase
short-term costs). It is also conceivable that backers may prod innovating entrepreneurs towards
developing a product more consistent with their entrenched expectations, hindering risk taking and acting
as a form of inertia, pressuring the innovating entrepreneur to be consistent with current market offerings
(Moreau et al., 2001). From a knowledge creation perspective, increasing the number of backers as direct
exchange partners increases the amount of information, resources, and ideas available to innovating

8
entrepreneurs (McFadyen and Cannella, 2004). Theoretically, this should have positive implications for
innovation and knowledge creation as these ideas are articulated and amplified through the organization’s
product development process. Yet, KCT research also notes that continual interaction with specific
individuals leads to increasingly similar knowledge stocks (Coleman, 1988), which leads to less radical
innovation over time. Thus, backers’ potential to both help and hinder the innovation process raises
important unanswered questions with regarding to crowdfunding.
Beyond feedback, backers also play an important role as product evangelists, spreading word of
mouth about the product via both traditional and contemporary (i.e., social media) means (Scholz, 2015).
Word of mouth from early adopters has long been thought to be extremely influential in determining
product launch success, as it reduces uncertainty for subsequent product adopters (Rogers, 2003). Since
crowdfunded products are not yet commercially available through typical channels, crowdfunding backers
can be viewed as a preliminary category of innovation adopters, prior to other groups (i.e., innovators and
early adopters) valued for their ability to spread word of mouth awareness from an influential position
within a network of potential later product adopters (see Kozinets et al., 2010, Rogers, 2003).
2.2. Open Search
Open search includes all tactics intended to uncover new ideas and knowledge from entities
external to the organization (Katila and Ahuja, 2002). These tactics identify knowledge within external
individuals and organizations, expose the firm to new ideas and possibilities, and augment the current
knowledge base of the innovating organization (Salge et al., 2013). A common refrain of open innovation
is that “not all the smart people work for us” (Chesbrough, 2003, p. xxvi). Knowledge, creativity, and
innovation often arise from interactions with individuals external to the firm. Echoing this, Dahlander and
Gann (2010, p. 699) note that an organization cannot “innovate in isolation”. Open search is the
embodiment of this principle; firms conducting open search draw from customers (including
crowdfunding backers), contractors, universities, professional organizations, competitors and others to aid
in their product development efforts.

9
Extant research on open search has found it insightful to conceptualize it in two dimensions:
depth and breadth (Laursen and Salter, 2006, van Wijk et al., 2012, Zang et al., 2014). Open search depth
is defined as the extent to which organizations draw intensively from external sources of innovative ideas
(Laursen and Salter 2006) and is associated with exploitation and efficiency (March, 1991, Zang et al.,
2014). Search depth generally implies that the organization is building on existing knowledge and
relationships, enabling efficiency and routinization by exploiting familiar knowledge sources to develop
new products (Katila and Ahuja, 2002). From a KCT perspective, deeply developing current knowledge
capabilities narrows the scope and direction of future firm innovation activities as routinization and
embedded rigidities constrain the creation of new knowledge (Nonaka et al., 2000). The repeated use of
similar information sources will likely lead to familiar relationships based on respective competencies, in
which neither party has incentive to expand their set of technological competencies – instead focusing on
efficiently moving products towards commercialization (Rothaermel and Deeds, 2004).
Open search breadth is defined as the number of different external sources of innovative ideas
that a firm draws upon in its innovative activities. (Laursen and Salter, 2006). It is associated with
exploration and divergent information search (Chiang and Hung, 2010, Katila and Ahuja, 2002). With
respect to KCT, open search breadth allows innovative entrepreneurs to develop their own stock of unique
knowledge as they are exposed to more diverse experiences and viewpoints from numerous backers (see
Henard and McFadyen, 2008). In a practical sense, breadth boosts the entrepreneurs’ absorptive capacity
(Cohen and Levinthal, 1990) and fosters more innovative, expansive thinking. Open search breadth
allows an innovating organization to broaden its base of knowledge beyond its existing boundaries,
allowing for meaningful departures through the consideration of divergent knowledge sources (Katila and
Ahuja, 2002). Breadth also acts against the development of core rigidities within the firm, allowing for a
more adaptable innovation organization (Leonard-Barton, 1992, Zang et al., 2014).
2.3. Stage to Crowdfund
When is the optimal time in the new product development (NPD) process (i.e., early, mid, or late
stage) to inject backer voices into the innovation conversation? All else being equal, equity-based

10
investors would likely prefer to invest in a product that is at a later stage in the development process. This
would better ensure that any technical and organizational hurdles could be overcome. Conversely,
reward-based backers often value “being along for the ride” of product development, preferring to
experience the entire product development process (Agrawal et al., 2014). For reward-based backers,
crowdfunding has a definite social element, with part of the personal benefit often being direct
communication with the innovating entrepreneur (Agrawal et al., 2014, Gerber et al., 2012). These
crowdfunding backers take a participative role by offering suggestions and feedback to innovating
entrepreneurs (Lewis-Kraus, 2015). Some innovating entrepreneurs have encouraged this by soliciting
input from backers at key decision points (e.g., Diallo, 2014).
The portion of product development complete at the time of crowdfunding is likely to affect our
outcomes of interest. Salge et al. (2013) argue that open search is particularly meaningful in earlier stages
of product development since at this stage novel insights can be generated by integrating or recombining
divergent ideas from outside the organization. Backers’ desire for involvement throughout the NPD
process pushes many innovating entrepreneurs to crowdfund early in the NPD process when there is a
greater chance for backers’ input to impact the direction of the project. Conversely, there is also evidence
to suggest that open search can be beneficial at a later stage of development, once the innovating
entrepreneur has a more developed concept to collaborate on with outsiders (Huizingh, 2011).
Understanding the timing implications of backer involvement is crucial to a more comprehensive
understanding of crowdfunding and innovation, and is yet to draw the attention of researchers.
3. Hypotheses
A conceptual overview of our research is detailed in Figure 1. Recall that there are two outcomes
of interest: product market performance (the degree of financial success experienced once the
crowdfunded product is launched) and radical innovation focus (the degree that the organization focuses
on radical innovation in subsequent efforts). In terms of product market performance, we anticipate that
the effect of backers is positive but that backers’ true value to innovating entrepreneurs lies in their ability

11
to provide ideas and feedback during the product development process as well as to act as product
evangelists, spreading awareness to other potential adopters. First, backers are known to play an active
role in the development of crowdfunded products (Diallo, 2014, Lewis-Kraus, 2015). Two-way
interaction between firms and innovative outside actors is critical for product development success (Enkel
et al., 2009, West and Bogers, 2014) and despite the potential costs of interacting with backers, we expect
firms with a larger number of backers to have more enriching engagement, getting input and valuable
ideas from interested backers as well as quickly receiving feedback from backers at key points during
product development.
Figure 1
Conceptual Overview.
Note: Solid lines are hypothesized and empirically tested. Dashed lines are empirically tested only. For simplicity,
only direct effects shown.
Further, firms with a larger number of backers are able to harness a larger network of product
evangelists to spread worth of mouth, assisting the market launch of the product (relatedly, the size of the
crowdfunding entrepreneur’s personal network has also been shown to be related to campaign success;

12
Mollick, 2014). As crowdfunded products have, by definition, not yet reached the mainstream
marketplace, backers are the earliest possible adopters and might thus be even more valuable than
traditional early adopters (Scholz, 2015). Theoretically, these crowdfunding backers can be viewed as a
category of adopters prior to typical innovators and early adopters (see Rogers, 2003) since crowdfunding
backers choose to commit to the product before it is commercially available. Based on this preliminary
position on the adoption curve, valuable characteristics of early adopting consumers such as their
venturesomeness, desire to learn about technology and influential position with potential later adopters
are likely intensified for crowdfunding backers.
While we believe that the number of backers attracted will be important to the market success of
the product being crowdfunded, we do not believe that the amount of funding raised will be consequential
for subsequent product market performance (i.e., performance once the product is released to the
mainstream market). Essentially, innovating entrepreneurs are not able to effectively financially exploit
the crowdfunding process. Delays are commonplace when reward-based crowdfunding is employed (over
75% of technology and design projects are delayed; Mollick, 2014). The delays that are common is this
space are typically accompanied by cost over-runs in product development, manufacturing and logistics
(Agrawal et al., 2014). Crowdfunding also often requires the development of new skills for entrepreneurs,
and the substantial costs of developing these skills (Hui et al., 2012). For instance, the logistics of
shipping to many individual small customers is challenging, costly and time consuming (Greenberg and
Gerber, 2014). The result of this is that margins on crowdfunding pre-orders tend to be much smaller than
innovating entrepreneurs anticipate due to cost over-runs in development, production and logistics.
Theoretically, since exploitation is fostered by repeatedly drawing on refined, efficient
organizational routines and knowledge (March, 1991, van Wijk et al., 2012), which are not yet developed
for these startups and very small organizations, their inability to exploit pre-orders is predictable (as is
their inability to predict relevant hurdles). From an organizational learning perspective, consider that the
capabilities developed by a startup organization prior to crowdfunding may quickly become both
inadequate and inappropriate to meet the challenges they face after crowdfunding (see Zahra and

13
Filatotchev, 2004). Put another way, the two major phases that crowdfunding organizations typically
experience (capital raising and project implementation) require distinctly different skillsets, knowledge
and relationships (Xu et al., 2016). For instance, recent research links creativity and social orientation to
funding success (Calic and Mosakowski, 2016), though the competencies required for project
implementation are likely quite different. The knowledge and organizational competencies established
during and prior to crowdfunding leave entrepreneurs unprepared to complete the implementation phase
while financially exploiting crowdfunding sufficiently to bolster their mainstream market launch (for
instance, through increased promotional spending). This results in reward-based crowdfunding’s financial
impact on the product’s subsequent mainstream market performance being mitigated by the cumulative
(often unexpected) costs in the implementation phase. Thus, we argue that funding raised through
crowdfunding is inconsequential to the product’s subsequent mainstream market performance. In essence,
we argue that the organization typically emerges from a crowdfunding experience with backers
themselves as their only new asset. That is, backers’ value in contributing to the market success of the
product lies beyond their ability to contribute financially. More formally stated,
H1A: The number of backers attracted during a crowdfunding campaign is positively
related to product market performance.
H1B: The amount of funds raised during a crowdfunding campaign is not related to
product market performance.
We expect that open search depth (the extent to which firms draw intensively from external
sources of innovative ideas) will have a positive effect on product market performance. The exploitative
nature of search depth enhances market performance by allowing organizations to be more effective and
efficient in the NPD process (Henard and McFadyen, 2005, Katila and Ahuja, 2002). Repeatedly drawing
from the same information sources (e.g., backers, suppliers, retailers) leads to a stronger understanding of
the knowledge area and more efficient knowledge transfer through structured communication
mechanisms and the development of a shared innovation vocabulary (van Wijk et al., 2012). These more
deeply embedded relationships allow for a greater quantity of complex product development information
to be conveyed to the innovating organization due to heightened trust and emotional closeness

14
(Krackhardt, 1992, Rindfleisch and Moorman, 2001). More sizable and complex flows of knowledge
accessed through embedded relationships allow for greater product development effectiveness through in-
depth, systematic learning that is readily applied to NPD (Bonner and Walker, 2004). Repeatedly relying
on familiar information sources leads to routinization of innovation-related tasks, which can have
efficiency benefits (van Wijk et al., 2012), but may also render the information search somewhat
predictable and unlikely to generate unique and novel outcomes (Katila and Ahuja, 2002). Open search
depth is notably associated with efforts to address customers’ currently expressed needs (Benner and
Tushman, 2002, Salge et al., 2013), which should lead to faster, more assured product development
efforts.
Other studies examining the performance effects of openness have theorized and detected an
inverse U-shaped relationship between openness and performance (Belderbos et al., 2010, Laursen and
Salter, 2006, Salge et al., 2013). In the case of open search depth, relying too deeply on external
knowledge sources can lead to knowledge redundancy, a lack of new learning, as well as the time and
expense (e.g., staff, travel) that comes with relying heavily on external relationships (Salge et al., 2013).
For these reasons, over-relying on open search depth may be detrimental to market performance for most
firms. However, we view the case of startups and very small businesses as a boundary condition negating
the potentially harmful performance effect of high levels of depth. In the case of startups and very small
businesses, knowledge stores (and virtually all other resources) are relatively undeveloped within the
organization (Chesbrough et al., 2014). Given the lack of internal knowledge, it will not be possible to
over-rely on open search depth; at times, external ideas and knowledge may be the only ones available to
the crowdfunding entrepreneur (Henkel, 2006, Van de Vrande et al., 2009). Further, open search requires
internal resources (e.g., relationship building, technology scouting). The limited resources available to
these small firms to be deployed to open search makes it improbable that performance will be harmed by
drawing too intensively on external sources. Thus, we hypothesize a positive effect of open search depth
on product market performance, with no curvilinear effect:
H2: Open search depth is positively related to product market performance.

15
We expect that the effect of open search breadth (the number of different external sources of
innovative ideas that a firm draws upon in its innovative activities) on radical innovation focus will be
positive. From a KCT perspective, moving from a focus on depth to breadth implicitly implies a shift in
focus from value appropriation (i.e., exploitation) to value creation (i.e., exploration; Moran and Ghoshal,
1996). Open search breadth allows crowdfunding entrepreneurs and startups to expand beyond the
knowledge resident in the organization by interacting with a variety of outsiders (e.g., research institutes
and industry associations), allowing for the possibility of new variations through the consideration of
divergent knowledge bases and broad-based, expansive conversations (Katila and Ahuja, 2002, Tripsas
and Gavetti, 2000).
As noted, increasing the number of information sources drawn upon raises the amount of
knowledge available to the innovating entrepreneur to articulate and amplify through product
development (McFadyen and Cannella, 2004, Nonaka et al., 2000). This recombination of previously
disconnected ideas fosters a focus on creativity and radicalness, generally favoring exploration over
exploitation within product development (March, 1991, Salge et al., 2013). The explorative nature of
search breadth lends itself better to longer-term innovation initiatives and differs from search depth in
much the same way as basic research is distinct from applied research (Chiang and Hung, 2010, Henard
and McFadyen, 2005). Accessing a variety of external knowledge leads to recombining the firm’s
knowledge with distant knowledge elements in creative ways to broaden the scope of future product
development possibilities (van Wijk et al., 2012), while preserving the organization’s independence from
any specific knowledge source (Moran and Ghoshal, 1996). Conceptually, open search breadth should
lead to new innovation opportunities being explored in the future (Zang et al., 2014).
At the same time, accessing, evaluating and integrating new knowledge from disparate sources
can be difficult, costly and time consuming; thus search breadth can effectively inhibit exploitation
(Katila and Ahuja, 2002, Salge et al., 2013). While search depth can foster rigidity and inertia within an
organization, search breadth inoculates an organization against these efficiency traps (Leonard-Barton,
1992, Stanko et al., 2013, Zang et al., 2014). In essence, the search breadth activities that might be less

16
consequential to product market performance are posited to lead to crowdfunders having greater focus on
radicalness in subsequent innovation efforts.
There are also findings suggesting that, beyond a given point, open search breadth can decrease
creativity or radicalness (Laursen and Salter, 2006, Salge et al., 2013). These researchers have theorized
that beyond a certain point of open search breadth, knowledge attained will become redundant,
conflicting, or irrelevant. They further posit that additional breadth requires resources that would be more
impactfully deployed to internal activities central to radical innovation, such as integrating technologies,
considering customer needs and generally “concentrat[ing] their energy, effort and mindfulness on a
limited number of issues” (Ocasio, 1997, p. 203). Similar to our argument supporting H2, we argue that
there will not be negative implications of over-relying on openness (in this case breadth) in the case of
startups and very small organizations. For these companies (with very limited internal knowledge stores
developed), we argue that it is not yet possible to expose the small organization to so much breadth as to
discourage a focus on radical innovation. These small companies lack the resources (i.e., financial, people
and relational) to search out external knowledge sources so broadly that they could decrease their own
focus on radical innovation by doing so. That is, these small firms’ resources do not enable them to reach
the point where breadth of search becomes redundant and diminishes in its ability to foster radicalness.
Thus, we view this as a boundary condition to findings that for firms generally, breadth diminishes
radicalness beyond a certain threshold (Laursen and Salter, 2006, Salge et al., 2013). We consequently
hypothesize a positive effect of open search breadth on radical innovation focus, with no curvilinear
effect:
H3: Open search breadth is positively related to an organization’s radical innovation
focus.
Crowdfunding campaigns come to potential backers’ attention at various stages of product
development completion. Some innovating entrepreneurs launch crowdfunding campaigns for concepts
that are only at the beginning the product development process while others crowdfund nearly completed
products. Generally, exposing a product development process to an influx of backer opinion while

17
development is at an early, impressionable stage indicates a conscious structuring of innovation efforts to
include customer opinions as well as a tolerance for ambiguity and risk, which should result in higher
radical innovation focus into the future (Slater et al., 2014). At the early stages of product development,
there exist a rich set of possible ways backer engagement can impact the product development process.
Backers may suggest entirely new technologies that were not previously considered by the crowdfunding
entrepreneur; a large, diverse group of backers is very likely to generate novel solutions (Hong and Page,
2004). In terms of product design, at early stages of product development, entirely new product designs
can be developed based on input from interested backers (whereas design elements often become frozen
at later stages of development; Ulrich and Eppinger, 2003). Innovating entrepreneurs who have completed
most or all NPD activities prior to the crowdfunding campaign are likely less interested in listening to
backers. For these entrepreneurs, crowdfunding might be a purely financial exercise. Thus,
H4: The portion of new product development completed at the time of crowdfunding is
negatively related to an organization’s radical innovation focus.
We also believe that the interaction between the portion of NPD complete at the time of
crowdfunding and open search may be consequential in determining an organization’s radical innovation
focus. We hypothesize that for those organizations crowdfunding development efforts that are already at
an advanced stage, open search breadth will have a stronger, positive impact on subsequent radical
innovation focus. Open search breadth may be even more effective at fostering a radical innovation focus
in latter stages of product development when the entrepreneur or startup has something concrete to show
outsiders (Huizingh, 2011). To learn from external sources and focus on big ideas, crowdfunding
innovators likely need to demonstrate a reasonably advanced concept. Backers and other sources of
inbound open innovation might not take the time to critique and offer creative suggestions unless the
innovating entrepreneur has first taken substantial effort to develop the concept. This is consistent with
design and creativity research indicating that well framing a problem - akin to a more advanced concept
in this application - enables focused creativity (Dorst and Cross, 2001). Absent the direction that comes

18
from an advanced concept, outsiders may be hampered in efficiently developing ideas that will be valued
by the innovating crowdfunder.
Given that backers are pragmatically motivated to improve the product that they will eventually
use (just as user innovators are; Hienerth et al., 2014), backers, along with other outsiders, may have
greater motivation to meaningfully contribute when product development is nearing completion - their
payoff from improving the innovation is more immediate. Further, when shown an advanced concept,
outsiders will be better able to offer suggestions that represent a valid consumer viewpoint and mirror
those that future customers might have (both of these are important contributors to radical product
innovation; Slater et al., 2014).
Beginning the innovation conversation with backers by showing them a relatively advanced
concept will foster the two-way dialogue that is critical to effective NPD learning (Enkel et al., 2009,
West and Bogers, 2014). Though this late stage discourse may not always be conducive to efficiently
progressing through the product development process (thus we do not hypothesize this effect with regard
to product market performance), bringing in numerous external opinions during the late stages of product
development is likely to foster creative, divergent efforts to improve the products being developed, rather
than mechanical execution of pre-ordained product development plans. Combining the fine tuning that
happens in the latter stages of NPD with open search breadth allows organizations to combine a pre-
existing stock of knowledge with external voices, leading to creativity, divergent thinking and a generally
heightened focus on radical innovation (Zang et al., 2014). As such, we hypothesize,
H5: The positive effect of open search breadth on radical innovation focus is greater in
instances where a higher portion of product development has been completed at the time
of crowdfunding.
4. Methods
4.1. Data Collection and Composition
To test the hypotheses, we obtained webscraped data (i.e., data systematically harvested from the
internet) pertaining to funded Kickstarter campaigns. While other crowdfunding platforms take varying

19
funding approaches, campaigns on the Kickstarter platform must reach a preset funding goal to receive
the funding pledged from backers (an 'all or nothing' approach; Riedl, 2013). Kickstarter is the largest
reward-based crowdfunding platform and is used by many innovating entrepreneurs. While Kickstarter
began with the intention of helping to fund creative projects in categories such as theater, literature and
publishing, it has matured into a viable platform for technology entrepreneurs (Riedl, 2013). We focus on
four Kickstarter categories: technology, product design, hardware and video games. These categories are
populated with products (rather than services) and are consistent with our research interest in
technological innovation. Other Kickstarter categories that were not selected focus to a substantial degree
on services (e.g., dance, food) or are inconsistent with our emphasis on technological innovation (e.g.,
fashion, comics).
To allow for sufficient time between the crowdfunding campaign and our measurement of various
outcomes, we began with all funded campaigns raising at least $10,000 commencing between July 2013
and February 2014 (n=1,060). A search was then conducted for contact email addresses for the person(s)
responsible for each campaign (typically a founder of the organization). Information was sourced from
either the Kickstarter campaign page or by searching the links provided on that page. Projects were
deleted from the sample if: 1) they were not product innovations (e.g., an event that is categorized within
the technology category, n=6); 2) the same person(s) had multiple funded campaigns (i.e., only the first
campaign per person or group was included in our sample, n=18); or 3) the initiative no longer appeared
to be a going concern (n=27). Of the remaining 1,009 campaigns, email addresses were identified for 944
campaigns (93.6%).
Survey invitations were emailed at least one year, but not more than two years, after the
respective crowdfunding campaign. This enables us to gather survey data regarding the product
development process as well as providing an indication of market success of the product. Our approach is
consistent with survey based research in the field of product development relying on a survey “at least
one year after” product development to measure variables related to both product development and
market performance (e.g., Langerak et al., 2004, Moorman, 1995, Stanko et al., 2013, Swink and Song,

20
2007). There was a mean of 512 days between the start of the crowdfunding campaign and survey
completion. This is adequate to allow an indication of subsequent market performance of the
crowdfunded product, while staying consistent with established measurement approaches to ensure
entrepreneurs can accurately recall product development. To encourage their participation, innovating
crowdfunders were offered an advanced copy of the study results as well as a chance to win one of
several $100 Amazon.com gift cards. Two hundred twenty four individuals (23.7%) started the web-based
survey (i.e., they completed at least some information), with 173 completing all information required for
this study (18.3%). Twenty three individuals left the survey prior to completing all items pertaining to the
organization’s subsequent radical innovation focus, which explains the slight difference in sample size
across the study’s two dependent variables. The survey response rate is comparable to studies using
detailed web-based surveys and surpasses that of many published studies (cf. Sauermann and Roach,
2013). We attribute this to a high degree of interest the crowdfunding community has in better
understanding crowdfunding’s consequences for innovation.
As a check for representativeness of our sample, we compared the population of campaigns (i.e.,
the sampling frame) with our sample. Examining the population as a whole reveals that the population
contains a very small number of campaigns having an exceedingly disproportionate quantity of backers,
funding, updates or comments. For instance, the campaign drawing the most funding support in the
population garnered $3.85m (nearly 50 times the mean funding within our sample) from over 67,000
backers (over 60 times the sample mean). Understandably, given our research design (attempting to
survey the founders of these campaigns), these extremely large outliers are not well represented in our
sample. These founders are presumably difficult to reach through electronic communication (in part due
to the tens of thousands of backers who may be trying to communicate with them). Accordingly, to check
whether survey respondents are representative of the population, we first remove all outliers (i.e., those
observations more than four standard deviations away from the mean) on any of the four variables used in
our analysis that are available for the entire population (funding raised, number of backers, updates and
comments). This results in 23 campaigns being removed of the 1009 campaigns in the population

21
(2.28%). The 95% confidence interval of the mean for the remaining 986 campaigns always includes the
mean for survey respondents for each of the four variables (funding raised, number of backers, updates
and comments). Additionally, there is no statistically significant difference for any construct of interest (p
> .10) when comparing early vs. late survey respondents (Armstrong and Overton, 1977).
The campaigns in our sample had a mean funding goal of $28,417 and actually raised $78,726,
drawn from an average of 1,078 backers. Respondents’ organizations are typically small, reporting a
mean of 4.76 employees at the time of our survey. Respondent organizations’ small size makes this
sample an ideal context to test the theorized boundary condition at play for startups and very small
companies (H2 and H3). Interestingly, survey respondents report having 61.8% of product development
efforts completed at the time of the crowdfunding campaign. Thus, for most innovating entrepreneurs in
our sample there was still a substantial portion of the product development effort remaining, which could
be influenced by the input of backers attracted during crowdfunding. Eighty-six percent of the campaigns
in our sample were conducted in US dollars, with the remainder conducted in British Pounds, Canadian
Dollars or Australian Dollars. For all campaigns in our sample, currencies were converted to US dollars
based on exchange rates at the close of their respective Kickstarter campaigns.
4.2. Measures
As indicated in Figure 1, the first portion of the dataset used for this study was webscraped from
kickstarter.com. Funds raised represents the total funding raised through the Kickstarter campaign ($USD
in thousands). Number of backers is the number of people contributing to the specific Kickstarter
campaign. Several control variables are also drawn from the webscraped data. The number of updates that
innovating entrepreneurs provide to their backers (through the campaign’s Kickstarter page) as well as the
number of comments that backers post regarding the product (again, on the campaign’s Kickstarter page)
are used as control variables. Comments and updates are considered to be proxies for online activity from
both sides of the backer/entrepreneur relationship. Theoretically, it is important to be able to distinguish
these levels of online communication activity from constructs of interest. We also employ three indicator
variables to control for the category of the innovation (i.e., product design, technology, video game),

22
using hardware as the baseline condition. It is conceivable that entrepreneurs across these categories may
have meaningful differences, for instance, in terms of their radical innovation focus. Finally, the number
of days elapsed between the crowdfunding campaign and survey completion is also used as a control
variable. Controlling for the number of days elapsed between crowdfunding and the survey allows us to
ensure that there are not performance or innovativeness effects captured elsewhere within our models that
are due to the amount of time elapsed.
Each of the remaining constructs was measured using a key respondent survey conducted
between one and two years after crowdfunding (see Appendix A). For some sections of the survey,
respondents were instructed to recall the time of their specific crowdfunding campaign, whereas in other
sections respondents were instructed to consider current conditions at the time of the survey (consistent
with other survey research examining both product development and performance, e.g., Swink and Song,
2007). The portion of NPD complete at the time of crowdfunding was assessed by asking the
crowdfunding principals what percentage of NPD efforts (including activities such as developing the
product’s feature set, conducting business analysis, prototyping, engineering / design / coding, etc.) had
been completed for the specific product at the completion of the crowdfunding campaign.
Using standard measurement approaches (Laursen and Salter, 2006, Zang et al., 2014), open
search depth and breadth variables were both captured by asking respondents to what extent NPD ideas
are drawn from each of nine potential sources (see Appendix A) using a five-point scale anchored by
‘none’ (one) and ‘very much’ (five). From those responses, open source depth was calculated as the
number of sources heavily drawn from (i.e., four or five) and open source breadth was calculated as the
total number of sources consulted (i.e., more than ‘none’). In this way, single item measures are
calculated for open search depth and breadth based on the nine potential idea sources (Appendix A).
Importantly, the measures of openness pertain to the sources that NPD ideas are drawn from. The sources
of these ideas can be influential in determining the organization’s focus on radical innovation in their
innovation efforts (consistent with the causal flow indicated by H3 and H5, as well as prior research, e.g.,
Salge et al., 2013, Zhou and Li, 2012). For the two outcomes of interest (product market performance and

23
radical innovation focus), respondents were explicitly asked to consider the current state (i.e., at the time
of the survey). Product market performance
1
was measured using five items evaluating financial returns
of the specific crowdfunded product as of the time of the survey, with a seven point scale anchored by
(one) ‘far underperformed’ and (seven) ‘far exceeded’ (De Luca and Atuahene-Gima, 2007). Individual
scale items refer to (for instance) sales, return on investment and profitability for the specific
crowdfunded product at the time of the survey.
Radical innovation focus was measured using five items concerning how much of the
organization’s current (i.e., at the time of the survey) internal innovation efforts are devoted to radical
innovation, with a seven point scale anchored by (one) ‘very little’ and (seven) ‘a great deal’ (De Luca
and Atuahene-Gima, 2007, McGrath, 2001). Consistent with the definition of radical innovation focus
(the degree that the organization focuses on radical innovation in subsequent efforts), the radical
innovation focus measure specifically addresses innovation activities actually happening in the
organization at the time of the survey; our measurement is again congruent with the causal flow
hypothesized (H3 to H5). Importantly, the measure employed does not encompass strategy, culture, goals,
or aspirations regarding future radical innovation. The source of product development ideas can be
influential in determining the organization’s focus on radical innovation in their subsequent internal
innovation efforts. Multi item scales are detailed in Appendix A. We note that all variables involved in
squared or interaction terms were mean centered prior to constructing the interaction or quadratic terms.
4.3. Reliability and Validity
To assess the measurement properties of the data, a confirmatory factor analysis (CFA) of all
constructs of interest in this study was conducted using AMOS 22.0 (Appendix A). All multi item
constructs show composite reliability greater than .80, exceeding the traditional benchmark of .70.
Convergent validity was verified several ways. First, all items load significantly on their construct of
interest (p < .001) and second, the data indicate a strong fit between the measurement model and the data
1
Note that De Luca and Atuahene-Gima (2007) use the term “product innovation performance”, rather than “product market
performance”. We opt to refer to this construct as “product market performance” to avoid confusion with our other outcome
of interest, “radical innovation focus”.

24
(RMSEA = .036; CFI = .984; TLI = .977). The average variance extracted (AVE) for multi item scales
exceeds .50 in all cases and consistently is greater than the squared correlation between any two
constructs, thus providing evidence of discriminant validity (Fornell and Larcker, 1981). Multi item scales
were averaged and used with single item measures in the regression models reported in the following
section (see Table 1 for means and correlations).
Table 1
Pearson Correlations.
Mean
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
1. Product Market
Performance
4.19
1
2. Radical
Innovation Focus
4.41
-.003
1
3. Funding Raised
78.73
.139**
.081
1
4. Number of
Backers
1077.56
.259***
.074
.487***
1
5. NPD Complete
61.80
.143**
-.105
-.067
-.138**
1
6. Open Depth
1.33
.171**
.220***
-.008
-.008
.059
1
7. Open Breadth
4.04
.073
.365***
.011
-.050
-.124*
.522***
1
8. Updates
10.10
.017
.088
.061
.189***
-..309***
.017
.085
1
9. Comments
156.37
.138**
-.071
.324***
.515***
-.108
-.082
-.111
.280***
1
10. Days Elapsed
512.22
.065
-.066
-.103
.025
.011
.022
.017
.246***
-.027
1
11. Funding Goal
28.42
-.033
.199***
.651***
.167**
-.148**
.018
.195***
.155**
.232***
-.007
1
* p < .10 ** p < .05 *** p < .01

25
5. Analyses and Results
5.1. Product Market Performance
In the first set of regression analyses (Table 2), the dependent variable is specified to be product
market performance. Ordinary least squares estimation (conducted using Stata 14.0) is used for all
regression analyses. Model 1, which can be considered a base model, includes only the linear terms with
no interaction effects modeled. Model 2 includes quadratic terms for both open search depth and breadth
to test for potential curvilinear performance effects of openness and determine whether there is support
for our contention that startups and very small organizations constitute a boundary condition that negates
potential curvilinear effects. Finally, Model 3 includes interaction terms between both open search depth
and breadth with the portion of NPD complete. While not specifically hypothesized with respect to
product market performance, Model 3 is included in Table 2 to allow for contrasts to be drawn with the
equivalent model in Table 3. For Models 1-3, sample size (n=196) exceeds standard guidelines (i.e., ten
observations per independent variable).
The independent variables included in the first two models explain a moderate amount of the
variance in product market performance (R2 =12% and 13%, respectively). Model 1 provides support for
several hypothesized relationships. The number of backers has a positive and statistically significant (p <
.05) effect on product market performance, thus supporting H1A. Open search depth has a positive,
significant (p = .0509) effect on product market performance, providing support for H2. In contrast, open
search breadth does not significantly impact product market performance (p > .10). Model 2 results mirror
those of Model 1 in that both the number of backers and open search depth are statistically significant
predictors of product market performance (both p < .05) providing further support for H1A and H2.
Consistent with H1B, funding raised is not significantly related to product market performance in either
model (p > .10). No statistically significant curvilinear relationships were detected in the second model (p
> .10), which is consistent with our argument that startups and very small organizations constitute a
boundary condition to the curvilinear performance effects of openness found to apply to larger firms.

26
As is good practice in regression analyses (Deligonul et al., 2009, Neter et al., 1996), several
diagnostics were employed. Residual plots do not reveal any patterns indicative of heteroscedasticity.
Further, the Breusch-Pagan test does not detect heteroscedasticity (p > .10 for all models). Residuals
appear to be normally distributed (D'Agostino et al.'s 1990 joint test for skewness and kurtosis finds the
residuals' distribution to not be significantly different from normal; p > .10 for all models). For all three
models, Hamilton’s (2012) interquartile range procedure shows that the residuals’ pseudosigmas closely
mirror their standard deviations (again indicating residual normality), with no extreme outliers.
Multicollinearity does not appear to be a concern to the interpretation of our results as the highest
variance inflation factor (VIF) for Models 1-3 is 2.7, well below the standard benchmark of 10 (Neter et
al., 1996). White’s (1980) general test for specification error does not indicate evidence of
misspecification (p > .10 for all models).
5.2. Additional Product Market Performance Analyses
As previously mentioned, Kickstarter campaigns establish a funding goal, which must be
surpassed for pledges to be collected from backers. Given this, it is reasonable to question whether or not
the performance impact of funding raised from backers should be viewed in context of the funding goal.
To address this, we estimate nine additional models to investigate whether funding raised impacts
subsequent market performance when viewed in context of the funding goal. First, we controlled for the
funding goal by adding this variable to the previous regressions, again using product market performance
as a dependent variable (the regressions are otherwise identical to Models 1-3 in Table 2). Both funding
raised and the funding goal are non-significant (both p > .10) in all three regressions, with all other terms
materially unchanged. That is, in all three additional regressions, all other coefficients’ 95% confidence
intervals include the original coefficients from Models 1-3.
Next, we assembled a surplus funding variable (i.e., funding raised less funding goal) and used
this variable in place of the funding raised variable in regressions otherwise identical to Models 1-3 in
Table 2. The surplus funding variable is non-significant (p > .10) in all three models, with all other terms
materially unchanged in all three additional models. That is, all other coefficients’ 95% confidence

27
intervals include the original coefficients from Models 1-3. If both the surplus funding and the funding
raised variables are used together, with the rest of the regressions identical to Models 1-3, both the surplus
funding and funding raised terms are again non-significant (p > .10) in all three models, with all other
coefficients materially unchanged (95% confidence intervals always include the original coefficients from
Models 1-3). These additional analyses strongly support the perspective that funding raised is not
significantly impacting product market performance, even when viewed in context of the funding goal.
5.3. Radical Innovation Focus
In Table 3, radical innovation focus is specified as the dependent variable. Since entrepreneurs’
level of ambition may impact their subsequent focus on radical innovation, we control for the campaigns’
funding goal set at the beginning of crowdfunding ($USD in thousands). It is intuitive that a relatively
low funding goal may limit entrepreneurs’ time and effort devoted to pursuing radical innovation, while
those entrepreneurs with a higher funding goal may demonstrate a greater level of ambition with regard to
their product development efforts, more exhaustively pursuing radical innovation. As with Table 2, Model
1 serves as a base model and includes only the linear terms with no interaction effects modeled. Also,
consistent with Table 2, Model 2 includes quadratic terms for both open search depth and breadth,
allowing us to examine our contention that there will not be negative implications of over-relying on
openness. Model 3 includes interaction terms between both open search depth and breadth with the
portion of NPD completed and is included to test H5. Sample size (n=173) for all models in Table 3
meets traditional sample size requirements.
The independent variables in Models 1-3 explain a substantial amount of the variance in radical
innovation focus (R2 =34%, 35% and 38%, respectively). Open search breadth is positively related to
radical innovation focus (p < .01 in models 1 and 2, p < .05 in model 3), providing support for H3. H4 is
also supported as the portion of NPD complete at the time of crowdfunding is negatively related to radical
innovation focus (p < .10 in models 1 and 2, p < .05 in model 3). Within Model 3, there is a statistically
significant interaction effect between open search breadth and the portion of NPD complete on radical
innovation focus (p = .010), which supports H5. Note that this significant interaction relationship

28
contrasts with a nonsignificant interaction effect on product market performance (tested in Table 2 Model
3 but not hypothesized). To sum, all hypothesized relationships are supported by these data (with no
significant curvilinear effects detected).
Interestingly, while the funding goal control variable has a positive impact on radical innovation
focus (p < .10 in models 1 and 2, p < .05 in model 3), funding raised has a negative effect on radical
innovation (p < .10 in all three models). While funding raised through crowdfunding supports an ongoing
focus on radical innovation up to the point of the funding goal (i.e., the positive effect of the funding goal
is greater than the absolute effect of funding raised), funding raised beyond the goal hampers a radical
innovation focus, presumably as crowdfunding entrepreneurs become overwhelmed with logistical and
other concerns. Though not hypothesized, this finding is consistent with the logic of our other
hypothesized relationships; its implications will be discussed in the following section.
As with the previous set of analyses, several diagnostics are performed. First, neither residual
plots nor the Breusch-Pagan test indicate heteroscedasticity (p > .10 for models 1-3). D'agostino et al.'s
(1990) test for normality finds the residuals' distribution is not significantly different from a normal
distribution (p > .10 for all models). The residuals’ pseudosigmas closely mirror their standard deviations
and no extreme outliers are present (Hamilton, 2012). Multicollinearity does not appear to be an issue in
interpreting the results (the highest VIF for models in Table 3 is 3.0). White’s (1980) general test for
specification error does not detect significant evidence of misspecification (p > .10 for all models).

29
Table 2
Product Market Performance Regression Analyses.
Model
1
2
3
Linear Model
Quadratic Model
Interaction Model
B (Std. Error)
β
B (Std. Error)
β
B (Std. Error)
β
Funding Raised
.00015
(.00081)
.015
.00017
(.00081)
.018
.00021
(.00081)
.021
Number of Backers
.00018**
(.000071)
.220
.00018**
(.000072)
.224
.00017**
(.000073)
.204
NPD Complete
.0051
(.0037)
.108
.0046
(.0038)
.097
.0048
(.0038)
.102
Open Depth
.147*
(.0747)
.161
.221**
(.103)
.242
.137*
(.079)
.150
Open Breadth
.0026
(.052)
.004
-.012
(.056)
-.019
.0048
(.053)
.008
Open Depth Squared
--
--
-.024
(.022)
-.114
--
--
Open Breadth Squared
--
--
.0099
(.017)
.048
--
--
Open Depth x NPD complete
--
--
--
--
-.0016
(.0027)
-.055
Open Breadth x NPD complete
--
--
--
--
.0017
(.0021)
.075
Control Variables
Updates
.0094
(.018)
.044
.011
(.018)
.049
.0087
(.018)
.041
Comments
.00017
(.00021)
.071
.00016
(.00021)
.067
.00018
(.00021)
.077
Product Design Category
.165
(.249)
.060
.159
(.250)
.057
.184
(.251)
.066
Technology Category
.190
(.397)
.039
.165
(.399)
.034
.162
(.402)
.033
Video Game Category
-.495
(.391)
-.124
-.523
(.393)
-.131
-.541
(.403)
-.136
Days Elapsed
.0013
(.0014)
.069
.0013
(.0014)
.066
.0015
(.747)
.077
R2
.124
.130
.128
Adjusted R2
.072
.068
.065
F
2.38***
2.09**
2.05**
n
196
196
196
* p < .10 ** p < .05 *** p < .01 B = unstandardized coefficient with significance level noted β = standardized coefficient

30
Table 3
Radical Innovation Focus Regression Analyses.
Model
1
2
3
Linear Model
Quadratic Model
Interaction Model
B (Std. Error)
β
B (Std. Error)
β
B (Std. Error)
β
Funding Raised
-.0020*
(.0012)
-.185
-.0021*
(.0012)
-.189
-.0022*
(.0011)
-.199
Number of Backers
.00013
(.000081)
.140
.00013
(.000081)
.138
.00011
(.000081)
.112
NPD Complete
-.0079*
(.0041)
-.141
-.0074*
(.0042)
-.133
-.0096**
(.0041)
-.170
Open Depth
-.019
(.084)
-.018
-.073
(.118)
-.069
.018
(.086)
.017
Open Breadth
.167***
(.062)
.226
.205***
(.066)
.276
.138**
(.061)
.186
Open Depth Squared
--
--
.020
(.024)
.084
--
--
Open Breadth Squared
--
--
-.030
(.019)
-.123
--
--
Open Depth x NPD complete
--
--
--
--
-.00038
(.0029)
-.012
Open Breadth x NPD complete
--
--
--
--
.0061**
(.0023)
.231
Control Variables
Updates
.049**
(.020)
.184
.045**
(.020)
.170
.050**
(.020)
.189
Comments
-.000060
(.00026)
-.018
-.000029
(.00026)
-.0090
-.000020
(.00025)
-.0062
Product Design Category
-.301
(.288)
-.091
-.281
(.288)
-.085
-.192
(.284)
-.058
Technology Category
.127
(.433)
.023
.155
(.434)
.028
.134
(.423)
.024
Video Game Category
-2.52***
(.460)
-.515
-2.42***
(.463)
-.495
-2.87***
(.461)
-.586
Days Elapsed
-.0011
(.0016)
-.050
-.0011
(.0016)
-.048
-.00078
(.0015)
-.034
Funding Goal
.0093*
(.0049)
.183
.0090*
(.0049)
.179
.011**
(.0048)
.213
R2
.339
.350
.381
Adjusted R2
.290
.293
.326
F
6.85***
6.08***
6.94***
n
173
173
173
* p < .10 ** p < .05 *** p < .01 B = unstandardized coefficient with significance level noted β = standardized coefficient

31
6. Conclusions
Research at the intersection of crowdfunding and innovation is in its infancy. These findings can
help innovating entrepreneurs and startups understand how engaging with backers and other outsiders
supports the market performance of the crowdfunded product and a focus on radical innovation in
subsequent efforts. They also provide insights for policy makers.
6.1. Focus on Backers
While raising capital is obviously important and necessary to entrepreneurs and startups, the
amount of funding raised through reward-based crowdfunding is not found to be significantly related to
the later market success of that product. This research suggests that much of the value of crowdfunding to
innovating entrepreneurs lies in the non-financial benefits that come with attracting backers. We find that
the number of backers involved in a crowdfunding campaign is a key driver of the market performance of
the crowdfunded product. Crowdfunding backers can be viewed as a preliminary category of innovation
adopters (i.e., prior to all other groups on Rogers' classic adoption curve). Given that valuable traits of
early adopting consumers (such as their ability to spread word of mouth awareness from an influential
position within a network of potential later product adopters) may be intensified for crowdfunding
backers, entrepreneurs should be mindful of this source of backer value. To provide further evidence, we
asked survey respondents how valuable they found various potential non-financial benefits of
crowdfunding (Figure 2). Their responses confirm that the awareness generated from backers, as well as
feedback regarding the product, are tremendously valuable to crowdfunding entrepreneurs.
2
2
Note that awareness is rated as being significantly (p < .01) more valuable than all other listed benefits, and product feedback
is more valuable (p < .01) than all benefits except awareness.

32
Figure 2
Beyond financial backing, which of these potential benefits from crowdfunding proved valuable to your
organization?
In a follow-up interview, one survey respondent told us, “Backers seem to enjoy the behind-the-
scenes ‘story’, since many are not creators/manufacturers themselves”. This is consistent with Bitterl and
Schreier’s (2017) contention that living vicariously is a compelling motivator for backers. Another
respondent told us that “the feedback from the Kickstarter community was invaluable, (even though a lot
of it was quite negative)”. Crowdfunding entrepreneurs described communicating with backers in several
ways (social media, email, in person and through the Kickstarter comment boards), with many of these
conversations being backer initiated. Other respondents mentioned the value backers have in helping to
generate awareness of the product. For example, “A large number of backers provides those projects a
very inexpensive audience that can help overcome inadequate marketing…some backers become
incredible assets.” Backers have led to media coverage of the products, and interest from retailers. In
another interview, a respondent told us how backers have become part of the network that he draws on:
“We have hired two of our backers, have one of our board members from our backers, and one of our
investors as well. [There are] a lot of benefits on top of just funding the product.”
While many crowdfunding entrepreneurs understandably focus on funding levels, our results
indicate that this funding does not significantly impact the crowdfunded product’s later market success.

33
Unforeseen costs and delays are widespread among startups, especially technology initiatives. Previous
research has found that the large majority of crowdfunded technology and design products are delayed
(Mollick, 2014), which often coincides with increased costs and lower margins than anticipated on
backers’ pre-orders. Crowdfunding entrepreneurs often underestimate the effort, time and cost that is
needed for development, marketing and fulfillment tasks (Hui et al., 2012). So, while capital is a logical
focus of many innovating entrepreneurs, funding obtained through reward-based crowdfunding typically
does not change the trajectory of the crowdfunded product’s market launch. One respondent commented,
“Having a successful campaign will not necessarily translate into a successful business - especially if the
company is unable to execute properly after the campaign.” Viewed another way, the size of the
collective “backers’ bet” is not necessarily a good indicator of the eventual market success of the product.
This is consistent with recent media reports of high profile crowdfunded products lagging in terms of their
business performance once campaigns are completed (Lomas, 2016, Mudhar, 2016).
Beyond the lack of impact on products’ market performance, our results also show that funding
raised (beyond the entrepreneurs’ funding goal) hampers entrepreneurs’ subsequent focus on radical
innovation. As entrepreneurs emerge from campaigns where pre-order expectations have been
dramatically surpassed, they may become particularly overwhelmed with logistical and customer service
concerns, as well as suffer an increased financial tie to pre-orders, which, again, often have diminished
margins. This is logically consistent with Mollick’s (2014) finding that overfunded projects are especially
vulnerable to fulfillment delays. The commitment to an unexpectedly high number of pre-orders
understandably acts to constrain entrepreneurs’ future innovation activities.
The implication of our research for crowdfunding entrepreneurs is clear. Reward-based
crowdfunding should be viewed as a way to engage with interested early adopters rather than purely as a
mechanism to raise funds. According to one Kickstarter alumni, “the feedback we get is ten-fold more
valuable than the money we raise” (Diallo, 2014). Structuring a campaign to attract as many backers as
possible – and to seek out deep involvement in the development effort – is generally an effective method
to improve the odds of market success for the product. Some crowdfunding entrepreneurs have begun to

34
recognize the potential value of backers, structuring their campaigns to offer low contribution thresholds
for receiving progress updates on product development as well as for digital rewards such as designs, e-
books and interactive downloads; each of which have negligible variable costs and minimal logistical
demands when fulfilling rewards to backers.
One way that crowdfunding entrepreneurs initiate conversations with backers is through updates.
We used the number of updates as a control variable in our analyses; while not a hypothesized variable of
interest, the results are nonetheless insightful. Notably, the number of updates to backers is significantly,
positively related to radical innovation focus (see Table 3). A high number of updates to backers indicates
that crowdfunders are playing an active role in initiating dialogue with backers. Taking this type of
participative stance towards crowdfunding is conducive to an ongoing focus on radical innovation.
6.2. When Depth and Breadth Matter Most
Extending previous research findings (Salge et al., 2013, van Wijk et al., 2012, Zang et al., 2014),
we demonstrate that for innovating entrepreneurs pursuing reward-based crowdfunding, open search
depth is positively associated with a product’s market performance. Exploiting familiar knowledge
sources to predictably advance a product toward launch leads to more assured market response. Open
search breadth, while not diminishing product market performance, was shown to foster a radical
innovation focus on subsequent efforts. This focus on radical innovation is desirable to enable adaptation
to a changing environment, develop products that are difficult to imitate and reap the larger financial
rewards that stem from radical innovation (Rubera and Kirca, 2012, Slater et al., 2014). By understanding
the outcome differences between open search depth and breadth, crowdfunding entrepreneurs can more
consciously structure their interactions with backers, suppliers, contractors and other innovative outsiders.
As predicted, we do not observe the inverse U-shaped performance effect of the two dimensions
of openness reported by other studies (Belderbos et al., 2010, Laursen and Salter, 2006, Salge et al., 2013)
for either of our dependent variables. Early stage entrepreneurs and startups often look to open search out
of necessity (Van de Vrande et al., 2009) and our results indicate that the small entrepreneurs in our
sample have not yet built up enough internal knowledge to become over-reliant on external voices. This

35
constitutes an important, yet previously unobserved, boundary condition to the curvilinear performance
effects of openness theorized and empirically detected by other researchers examining larger
organizations (Belderbos et al., 2010, Laursen and Salter, 2006, Salge et al., 2013).
Our results have clear implications (in terms of openness) for reward-based crowdfunders. Those
entrepreneurs who are primarily interested in the market performance of the particular crowdfunded
product should focus on interacting with relatively few categories of innovative outsiders (e.g., backers,
industry organizations, universities), but interact extensively with the selected outsiders. This deeper,
likely repetitive, level of interaction is more likely to lead to NPD efficiencies, which can enhance near-
term performance. Conversely, reward-based crowdfunders with an eye toward a radical innovation
orientation should focus on interacting with as many categories of innovative outsiders as feasible. This
breadth of external input encourages risk taking and experimentation in future efforts, which can lead to
the development of breakthrough products (Mascitelli, 2000).
6.3. When to Seek Backer Input
Generally, we observe that entrepreneurs who crowdfund later in the product development
process tend to have less focus on radical innovation in subsequent innovation efforts. It appears likely
that such entrepreneurs treat the crowdfunding process more as a transactional financial tactic rather than
an opportunity to meaningfully engage with backers. This could be a viable reward-based crowdfunding
tactic for those organizations not prioritizing radical innovation.
The interaction found between open search breadth and the portion of NPD completed on
subsequent radical innovation focus constitutes an interesting research contribution. While some scholars
have found that search openness can be of most benefit to product development efforts during the early
(i.e., ideation) stage (e.g., Aloini and Martini, 2013, Salge et al., 2013), our results challenge this in the
case of innovating crowdfunders. By broadly consulting external voices even relatively late in the
development process, entrepreneurs and startups can achieve an increased focus on radical innovation in
subsequent efforts (with no significant impact on product market performance). Seen in connection with
other findings from this study, the negative radical innovation focus implications of crowdfunding late in

36
the development process can be somewhat mitigated by the use of open search breadth. By being open to
a variety of external voices, even when technological challenges have narrowed and launch deadlines
loom, innovating organizations can foster divergent thought within all stages of product development –
even late stage fine tuning (Smed and Salomo, 2012, Zang et al., 2014).
7. Policy Implications
This research may also inform the perspective of policy makers toward reward-based
crowdfunding. While there has been much discussion regarding the need to protect backers from delayed,
cancelled and fraudulent campaigns (Agrawal et al., 2014, Mollick, 2014) and, even more recently
regarding intermediaries’ unwillingness to offer purchase protection to crowdfunding backers (Morris,
2016), our results indicate that there may also be a need to better educate entrepreneur crowdfunders.
While previous research suggests that crowdfunders underestimate the effort required for successful
crowdfunding (Hui et al., 2012) and often fail to fulfill pre-ordered products on time (Mollick, 2014), our
research casts doubt on whether funds raised through reward-based crowdfunding are impactful in
enabling the later widespread market success of the product.
In follow-up interviews, respondents commented on the disconnect between their expectations
from crowdfunding a product development effort and the reality they experienced. One crowdfunding
entrepreneur noted, “My sense after running the Kickstarter is that many people launching the
crowdfunding campaigns have wildly incorrect expectations.” Similarly, another respondent commented,
“Most of the Kickstarter community has no idea what goes into product development and getting a
product to market.” Relatedly, Mollick and Kuppuswamy (2014) find that developing a more complete
initial schedule leads to on time product fulfilment for technology crowdfunders. It may be important to
educate entrepreneurs on potential pitfalls of crowdfunding technology products (most saliently, the
notion that delays and cost overruns should be expected), just as it is to educate potential backers on these
campaigns’ risks. Crowdfunding entrepreneurs would benefit from an early understanding of the non-
financial benefits of backers (both in terms of product feedback and as conduits for word of mouth

37
awareness). More positively, these results suggest that entrepreneurs who have conducted funded reward-
based crowdfunding campaigns are often focused on radical innovation well after crowdfunding is
complete (which can have valuable economic implications; e.g., Lindič et al., 2012). Alongside recent
findings that over 90% of successful technology crowdfunding initiatives are still ongoing ventures one
year later (with substantially higher revenue than prior to crowdfunding; Mollick and Kuppuswamy,
2014), this points to long term value enabled by reward-based crowdfunding, which should be noteworthy
to policy makers.
8. Limitations and Future Research Directions
To our knowledge, this study represents the first research to couple data from an online
crowdfunding platform with survey data collected from the respective innovating entrepreneurs. There are
limitations that require mentioning and ample opportunity for related future research. Given that our
sample is composed of funded crowdfunding campaigns, it is worth mentioning that our results and
subsequent inferences are garnered from an efficient frontier of crowdfunding campaigns (i.e., survivor
bias). These results cannot be readily applied in the case of unfunded crowdfunding campaigns. While
there is a range of funded campaigns (in terms of size) represented in our study, we caution against
extending our findings to the case of extreme outlier campaigns. Questions surrounding these massive
campaigns (those attracting millions of dollars from tens of thousands of backers) may also prove
intriguing to researchers. Further, it is conceivable that collecting data one to two years after
crowdfunding is not a sufficient time span to allow for all contributors to a radical innovation focus to be
at play. While this study measured variables relating to both product development and outcomes using a
single survey, surveying at multiple points in time would allow for stronger determination of causality
and may also generate insights not possible using our approach, such as examining very long term
outcomes, or measuring the evolving use of open search tactics over time. While outside this project’s
scope, there are many potentially interesting questions regarding crowdfunding’s impact on openness. It

38
may also prove valuable to examine alternative conceptualizations of radicalness, such as a radical
innovation strategy or culture (see Leifer et al., 2001), which have the potential to impact open search.
It is noteworthy that this study did not detect the inverse U-shaped performance effects of
openness found elsewhere (e.g., Belderbos et al., 2010, Laursen and Salter, 2006, Salge et al., 2013).
Future scholars should look to replicate this finding with other small innovating organizations to better
understand where search openness does and, importantly, does not have performance drawbacks (i.e.,
understanding where this boundary condition holds). We also show that updates from innovating
entrepreneurs to backers foster a lingering focus on radical innovation. Future researchers should
investigate whether particular types of updates (e.g., technical updates, calls for backer action, disclosure
of setbacks) are more supportive of a radical focus than others. It might also prove interesting to examine
the content of backer comments. From these comments, it might be possible to determine (for instance)
whether the level of backer expertise is critical. While we have shown that backers (as a group) are
valuable, questions relating to which of them are most valuable are yet to be addressed.
So as to not allow individual entrepreneurs to unduly bias our analysis, we included only one
campaign per person or organization in our sample. This does not allow us to deeply explore the
particular case of serial crowdfunders – those entrepreneurs repeatedly using crowdfunding as a way to
fund, launch and promote new products. There are many potentially interesting research questions
regarding serial crowdfunders, including understanding the changes in information flow over the length of
a relationships with a retained backer. Presumably, just as in other innovation partnerships, the relational
embeddedness that comes with repeated entrepreneur-backer dealings would lead to more extensive
information acquisition (Rindfleisch and Moorman, 2001), though repeatedly relying on the same backers
could lead to inertia (Stanko et al., 2013). While we focused on reward-based crowdfunding, there may be
other interesting questions surrounding other forms of crowdfunding, such as donation and lending-based.
Potential motivational differences between reward-based and lending-based backers and equity-based
investors may shape entrepreneurs’ innovation tendencies (for a discussion of motivational differences
between reward-based backers and equity-based investors see Cholakova and Clarysse, 2015).

39
Through this research project, we provided a starting point for practitioners, researchers and
policy makers to understand the innovation outcomes of reward-based crowdfunding. There remain many
additional interesting questions, and we hope that future researchers will build upon this foundation.
Acknowledgements
The efforts of Editor Keld Laursen as well as the three anonymous reviewers are greatly appreciated.
Thank you to Himanshu Agrawal, Amir Lowery, James Strickland and Jeremy Tracz for excellent
research assistance. Thanks also to Water Haas for assistance with some of the data used for this project.
Importantly, thanks to all the crowdfunding innovators who graciously gave of their own time to benefit
this research.

40
Appendix A
Confirmatory Factor Analysis, Survey Scales, Loadings and Reliability Estimates
The CFA performed included all constructs of interest as outlined in §4.2, including variables webscraped
from kickstarter.com, as well as data obtained through an online survey. Since all webscraped variables
are single item measures (for which composite reliability and factor loadings are unavailable), they are
not shown below.
Constructs, Survey Instructions, Items
Composite
Reliability
Standardized
λ
Product Market Performance
For the specific product involved in this Kickstarter campaign, rate the extent to which
your organization has achieved the following product development objectives:
.904
Market share relative to your organization’s objectives.
Sales relative to your objectives.
Return on assets relative to your objectives.
Return on investment related to your objectives.
Profitability relative to your objective
.605
.761
.909
.916
.822
Radical Innovation Focus
Currently, how much of your organization’s innovation efforts are devoted to:
.835
Developing products that are new to your industry.
Developing products based on revolutionary changes in technology.
Developing products that challenge the way customers traditionally perform tasks
Developing products that require expertise not normally found in your industry
Efforts to greatly improve product effectiveness (i.e., 5x-10x improvements)
.609
.824
.746
.687
.655
Open Search Breadth and Open Search Depth*
In your organization to what extent are new product development ideas drawn from:
---
Customers
Suppliers
Competitors
Consultants
Contracted R&D or design firms
Distributors/retailers
Universities or other research institutes
Industry technical/trade associations
Investors (i.e., equity investors)
---
---
---
---
---
---
---
---
---
Portion of New Product Development Completed at the time of Crowdfunding
---
At the time the Kickstarter campaign was completed, what percentage of new
product development efforts (including activities such as developing the product’s
feature set, conducting business analysis, prototyping, engineering/design/coding
etc.) had been completed for this specific product?
---
χ2 (131) = 90.64; RMSEA = .036; CFI = .984; TLI = .977
*The breadth and depth measurements were calculated from the openness items as discussed in §4.2.
Since these two constructs must be calculated from the same complete set of openness items (consistent
with Laursen and Salter, 2006), only the breadth and depth measures (not the individual items) are
included in the CFA.

41
References
Agrawal, A., Catalini, C., Goldfarb, A., 2014. Some Simple Economics of Crowdfunding. Innovation
Policy and the Economy, 14, 63-97.
Aloini, D., Martini, A., 2013. Exploring the Exploratory Search for Innovation: A Structural Equation
Modelling Test. International Journal of Technology Management, 61, 23-46.
Armstrong, J. S., Overton, T. S., 1977. Estimating Nonresponse Bias in Mail Surveys. Journal of
Marketing Research, 14, 396-402.
Belderbos, R., Faems, D., Leten, B., Looy, B. V., 2010. Technological Activities and Their Impact on the
Financial Performance of the Firm: Exploitation and Exploration within and between Firms.
Journal of Product Innovation Management, 27, 869-882.
Belleflamme, P., Lambert, T., Schwienbacher, A., 2013. Individual Crowdfunding Practices. Venture
Capital, 15, 313-333.
Belleflamme, P., Lambert, T., Schwienbacher, A., 2014. Crowdfunding: Tapping the Right Crowd.
Journal of Business Venturing, 29, 585-609.
Benner, M. J., Tushman, M., 2002. Process Management and Technological Innovation: A Longitudinal
Study of the Photography and Paint Industries. Administrative Science Quarterly, 47, 676-707.
Bitterl, S., Schreier, M., 2017. When Consumers Become Project Backers: The Psychological
Consequences of Participation in Crowdfunding. Working Paper.
Bonner, J. M., Walker, O. C., 2004. Selecting Influential Business to Business Customers in New Product
Development: Relational Embeddedness and Knowledge Heterogeneity. J Prod Inno Manag, 21,
155-169.
Bouncken, R. B., Komorek, M., Kraus, S., 2015. Crowdfunding: The Current State Of Research. The
International Business & Economics Research Journal (Online), 14, 407.
Calic, G., Mosakowski, E., 2016. Kicking Off Social Entrepreneurship: How A Sustainability Orientation
Influences Crowdfunding Success. Journal of Management Studies, 53, 738-767.
Chesbrough, H., 2003. Open Innovation: The New Imperative for Creating and Profiting from
Technology, Boston, Harvard Business Press.
Chesbrough, H., Vanhaverbeke, W., West, J., 2014. New frontiers in Open Innovation, OUP Oxford.
Chiang, Y. H., Hung, K. P., 2010. Exploring Open Search Strategies and Perceived Innovation
Performance. R&D Management, 40, 292-299.
Cholakova, M., Clarysse, B., 2015. Does the Possibility to Make Equity Investments in Crowdfunding
Crowd Out Reward-Based Investments? Entrepreneurship Theory and Practice, 39, 145-172.
Cohen, W. M., Levinthal, D. A., 1990. Absorptive Capacity: A New Perspective on Learning and
Innovation. Administrative Science Quarterly, 35, 128-152.
D'Agostino, R. B., Belanger, A., D'Agostino Jr, R. B., 1990. A Suggestion for using Powerful and
Informative tests of Normality. The American Statistician, 44, 316-321.
Dahlander, L., Gann, D. M., 2010. How Open is Innovation? Research Policy, 39, 699-709.
De Luca, L. M., Atuahene-Gima, K., 2007. Market Knowledge Dimensions and Cross-Functional
Collaboration: Examining the Different Routes to Product Innovation Performance. Journal of
Marketing, 71, 95-112.
Deligonul, Z. S., Chabowski, B. R., Seggie, S. H., Xu, S., Cavusgil, S. T., 2009. Methodological Rigor of
Ordinary Least Squares Regression Studies in mainstream Marketing: An Assessment of
Research reported in the Journal of Marketing. In: Rudolf R. Sinkovics, P. N. G. (ed.) New
Challenges to International Marketing (Advances in International Marketing, Volume 20)
Emerald Group Publishing Limited.
Diallo, A., 2014. Crowdfunding Secrets: 7 Tips For Kickstarter Success. Forbes.
Dorst, K., Cross, N., 2001. Creativity in the Design Process: Co-evolution of Problem–Solution. Design
studies, 22, 425-437.
Enkel, E., Gassmann, O., Chesbrough, H., 2009. Open R&D and Open Innovation: Exploring the
Phenomenon. R&D Management, 39, 311-316.

42
Faems, D., De Visser, M., Andries, P., Van Looy, B., 2010. Technology Alliance Portfolios and Financial
Performance: Value-Enhancing and Cost-Increasing Effects of Open Innovation*. Journal of
Product Innovation Management, 27, 785-796.
Fornell, C., Larcker, D. F., 1981. Structural Equation Models with Unobservable Variables and
Measurement Error: Algebra and Statistics. J Marketing Res, 18, 382-388.
Gerber, E. M., Hui, J. S., Kuo, P.-Y., Crowdfunding: Why People are Motivated to Post and Fund
Projects on Crowdfunding Platforms. Proceedings of the International Workshop on Design,
Influence, and Social Technologies: Techniques, Impacts and Ethics, 2012.
Grant, R. M., 1996. Toward a Knowledge-Based Theory of the Firm. Strategic Management Journal, 17,
109-122.
Greenberg, M. D., Gerber, E. M., Learning to Fail: Experiencing Public Failure Online Through
Crowdfunding. Proceedings of the SIGCHI Conference on Human Factors in Computing
Systems, 2014. ACM, 581-590.
Hamilton, L. C., 2012. Statistics with Stata: version 12, Cengage Learning.
Henard, D. H., McFadyen, M., 2005. The Complementary Roles of Applied and Basic Research: A
Knowledge‐Based Perspective. Journal of Product Innovation Management, 22, 503-514.
Henard, D. H., McFadyen, M. A., 2008. Making Knowledge Workers more Creative. Research-
Technology Management, 51, 40-46.
Henkel, J., 2006. Selective Revealing in Open Innovation Processes: The Case of Embedded Linux.
Research Policy, 35, 953-969.
Hienerth, C., Von Hippel, E., Jensen, M. B., 2014. User Community vs. Producer Innovation
Development Efficiency: A First Empirical Study. Research Policy, 43, 190-201.
Hong, L., Page, S. E., 2004. Groups of Diverse Problem Solvers can Outperform Groups of High-Ability
Problem Solvers. Proceedings of the National Academy of Sciences of the United States of
America, 101, 16385-16389.
Hui, J. S., Gerber, E., Greenberg, M., 2012. Easy Money? The Demands of Crowdfunding Work.
Northwestern University, Segal Design Institute.
Hui, J. S., Greenberg, M. D., Gerber, E. M., Understanding the Role of Community in Crowdfunding
Work. ACM Proceedings, 2014. 62-74.
Huizingh, E. K., 2011. Open Innovation: State of the Art and Future Perspectives. Technovation, 31, 2-9.
Katila, R., Ahuja, G., 2002. Something old, Something new: A Longitudinal Study of Search Behavior
and New Product Introduction. Academy of Management Journal, 45, 1183-1194.
Kickstarter, 2015. Stats. Available: https://www.kickstarter.com/help/stats [Accessed Nov. 3, 2015].
Kozinets, R. V., De Valck, K., Wojnicki, A. C., Wilner, S. J. S., 2010. Networked Narratives:
Understanding Word-of-Mouth Marketing in Online Communities. Journal of Marketing, 74.
Krackhardt, D., 1992. The Strength of Strong Ties. In: Nohria, N. & Eccles, R. G. (eds.) Networks and
Organizations: Structure, Form and Action. Boston: Harvard Business School Press.
Langerak, F., Hultink, E. J., Robben, H. S. J., 2004. The Impact of Market Orientation, Product
Advantage, and Launch Proficiency on New Product Performance and Organizational
Performance. Journal of Product Innovation Management, 21, 79-94.
Laursen, K., Salter, A., 2006. Open for Innovation: The Role of Openness in Explaining Innovation
Performance Among U.K. Manufacturing Firms. Strategic Management Journal, 27, 131-150.
Lee, N., Sameen, H., Cowling, M., 2015. Access to Finance for Innovative SMEs since the Financial
Crisis. Research Policy, 44, 370-380.
Lee, S., Park, G., Yoon, B., Park, J., 2010. Open Innovation in SMEs—An Intermediated Network
Model. Research Policy, 39, 290-300.
Leifer, R., Gina Colarelli, O. C., Mark, R., Gina Colarelli, O. C., 2001. Implementing Radical Innovation
in Mature Firms: The Role of Hubs. Academy of Management Executive, 15, 102-113.
Leonard-Barton, D., 1992. Core Capabilities and Core Rigidities: A Paradox in Managing New Product
Development. Strategic Management Journal, 13, 111-125.

43
Lewis-Kraus, G., 2015. ZPM Espresso and the Rage of the Jilted Crowdfunder. New York Times
Magazine.
Lilien, G. L., Morrison, P. D., Searls, K., Sonnack, M., von Hippel, E., 2002. Performance Assessment of
the Lead User Idea-Generation Process. Management Science, 48, 1042-1059.
Lin, M., Viswanathan, S., 2015. Home bias in Online Investments: An Empirical Study of an Online
Crowdfunding Market. Management Science.
Lindič, J., Bavdaž, M., Kovačič, H., 2012. Higher Growth through the Blue Ocean Strategy: Implications
for Economic Policy. Research policy, 41, 928-938.
Lomas, N., 2016. Bitlock Crowdfunded a Smart Bike Lock in 2013. What Happened Next? TechCrunch
[Online]. Available: http://techcrunch.com/2016/04/28/bitlock-crowdfunded-a-smart-bike-lock-
in-2013-what-happened-next/ [Accessed May 10, 2016].
Love, J. H., Roper, S., Bryson, J. R., 2011. Openness, Knowledge, Innovation and Growth in UK
Business Services. Research Policy, 40, 1438-1452.
March, J. G., 1991. Exploration and Exploitation in Organizational Learning. Organization Science, 2,
71-87.
Mascitelli, R., 2000. Harnessing Tacit Knowledge to Achieve Breakthrough Innovation. Journal of
Product Innovation Management, 17, 179-193.
Massolution, 2015. The Crowdfunding Industry Report [Online]. Available:
http://www.crowdsourcing.org/editorial/global-crowdfunding-market-to-reach-344b-in-2015-
predicts-massolutions-2015cf-industry-report/45376 [Accessed April 20, 2016].
McFadyen, M. A., Cannella, A. A., 2004. Social Capital and Knowledge Creation: Diminishing returns of
the Number and Strength of Exchange Relationships. Academy of Management Journal, 47, 735-
746.
McGrath, R. G., 2001. Exploratory Learning, Innovative Capacity and Managerial Oversight. Academy
of Management Journal, 44, 118-131.
Mollick, E., 2014. The Dynamics of Crowdfunding: An Exploratory Study. Journal of Business
Venturing, 29, 1-16.
Mollick, E., 2016. The Unique Value of Crowdfunding Is Not Money — It’s Community. Harvard
Business Review, Online Only.
Mollick, E. R., Kuppuswamy, V., 2014. After the Campaign: Outcomes of Crowdfunding. Working
Paper.
Moorman, C., 1995. Organizational Market Information Processes: Cultural Antecedents and New
Product Outcomes. Journal of Marketing Research, 32, 318-335.
Moran, P., Ghoshal, S., Value Creation by Firms. Academy of Management Proceedings, 1996.
Academy of Management, 41-45.
Moreau, C. P., Lehmann, D. R., Markman, A. B., 2001. Entrenched Knowledge Structures and Consumer
Response to New Products. Journal of Marketing Research, 38, 14-29.
Morris, D. Z., 2016. PayPal to Drop Purchase Protection for Crowdfunding Projects. Fortune [Online].
Available: http://fortune.com/2016/05/08/paypal-to-drop-purchase-protection-for-crowdfunding-
projects/ [Accessed May 9, 2016].
Mudhar, R., 2016. Coolest Cooler leaves Kickstarter backers cold. Toronto Star, May 2, 2016.
Müllerleile, T., Joenssen, D. W., 2015. Key Success-Determinants of Crowdfunded Projects: An
Exploratory Analysis. Data Science, Learning by Latent Structures, and Knowledge Discovery.
Springer.
Nahapiet, J., Ghoshal, S., 1998. Social Capital, Intellectual Capital, and the Organizational Advantage.
Academy of Management Review, 23, 242-266.
Neter, J., Kutner, M. H., Nachtsheim, C. J., Wasserman, W., 1996. Applied Linear Regression Models,
3rd Edition, Chicago, IL, Irwin.
Nonaka, I., 1994. A Dynamic Theory of Organizational Knowledge Creation. Organization Science, 5,
14-37.

44
Nonaka, I., Toyama, R., Nagata, A., 2000. A Firm as a Knowledge-Creating Entity: A new perspective on
the Theory of the Firm. Industrial and Corporate Change, 9, 1-20.
Ocasio, W., 1997. Towards an Attention-Based View of the Firm. Strategic Management Journal, 18,
187-206.
Riedl, J., 2013. Crowdfunding Technology Innovation. Computer, 46, 100-103.
Rindfleisch, A., Moorman, C., 2001. The Acquisition and Utilization of Information in New Product
Alliances: A Strength-of-Ties Perspective. Journal of Marketing, 65, 1-18.
Rogers, E. M., 2003. Diffusion of Innovations, 5th edition, New York, Free Press.
Rothaermel, F. T., Deeds, D. L., 2004. Exploration and Exploitation Alliances in Biotechnology: A
System of New Product Development. Strategic Management Journal, 25, 201-221.
Rubera, G., Kirca, A. H., 2012. Firm Innovativeness and Its Performance Outcomes: A Meta-Analytic
Review and Theoretical Integration. Journal of Marketing, 76, 130-147.
Salge, T. O., Farchi, T., Barrett, M. I., Dopson, S., 2013. When Does Search Openness Really Matter? A
Contingency Study of Health‐Care Innovation Projects. Journal of Product Innovation
Management, 30, 659-676.
Sauermann, H., Roach, M., 2013. Increasing Web Survey Response Rates in Innovation Research: An
Experimental Study. Research Policy, 42, 273-286.
Scholz, N., 2015. The Relevance of Crowdfunding. Springer Gabler.
Slater, S. F., Mohr, J. J., Sengupta, S., 2014. Radical Product Innovation Capability: Literature Review,
Synthesis, and Research Propositions. Journal of Product Innovation Management, 31, 552-566.
Smed, M., Salomo, S., Users ability to Share Knowledged when Integrated in New Product Development
- Evidence from the Pharmaceutical Industry. DRUID Proceedings, 2012 Cambridge.
Spaeth, S., Stuermer, M., Von Krogh, G., 2010. Enabling Knowledge Creation through Outsiders:
Towards a Push Model of Open Innovation. International Journal of Technology Management,
52, 411-431.
Stanko, M. A., Bohlmann, J. D., Molina-Castillo, F.-J., 2013. Demand-Side Inertia Factors and Their
Benefits for Innovativeness. Journal of the Academy of Marketing Science, 41, 649-668.
Stanko, M. A., Henard, D. H., 2016. How Crowdfunding Influences Innovation. MIT Sloan Management
Review, 57, 15.
Swink, M., Song, M., 2007. Effects of Marketing-Manufacturing Integration on New Product
Development Time & Competitive Advantage. Journal of Operations Management, 25, 203-217.
Tripsas, M., Gavetti, G., 2000. Capabilities, Cognition, and Inertia: Evidence from Digital Imaging.
Strategic Management Journal, 21, 1147-1161.
Ulrich, K. T., Eppinger, S. D., 2003. Product Design and Development, Tata McGraw-Hill Education.
Van de Vrande, V., De Jong, J. P., Vanhaverbeke, W., De Rochemont, M., 2009. Open Innovation in
SMEs: Trends, Motives and Management Challenges. Technovation, 29, 423-437.
van Wijk, R., Jansen, J. J., Van Den Bosch, F. A., Volberda, H. W., 2012. How Firms Shape Knowledge
to Explore and Exploit: A Study of Knowledge Flows, Knowledge Stocks and Innovative
Performance across units. Technology Analysis & Strategic Management, 24, 929-950.
West, J., Bogers, M., 2014. Leveraging External Sources of Innovation: A Review of Research on Open
Innovation. Journal of Product Innovation Management, 31, 814-831.
West, J., Salter, A., Vanhaverbeke, W., Chesbrough, H., 2014. Open Innovation: The Next Decade.
Research Policy, 43, 805-811.
White, H., 1980. A Heteroskedasticity-consistent covariance matrix Estimator and a Direct Test for
Heteroskedasticity. Econometrica: Journal of the Econometric Society, 817-838.
Xu, B., Zheng, H., Xu, Y., Wang, T., 2016. Configurational paths to Sponsor Satisfaction in
Crowdfunding. Journal of Business Research, 69, 915-927.
Zahra, S. A., Filatotchev, I., 2004. Governance of the Entrepreneurial Threshold Firm: A
Knowledge‐based Perspective. Journal of Management Studies, 41, 885-897.

45
Zang, J., Zhang, C., Yang, P., Li, Y., 2014. How Open Search Strategies Align with Firms’ Radical and
Incremental Innovation: Evidence from China. Technology Analysis & Strategic Management,
26, 781-795.
Zhou, K. Z., Li, C. B., 2012. How Knowledge affects Radical Innovation. Strategic Management Journal,
33, 1090-1102.