Digital Innovation Units: An Empirical Investigation of Performance Implications

Abstract

Digital innovation is both a necessary and a challenging endeavor for most firms. To achieve progress in this regard, firms across contexts increasingly set up digital innovation units (DIUs). Despite its popularity in practice, the prospects of this initiative are to date unclear. Building upon dynamic capabilities theory, we hypothesize the performance implications of DIUs and employ panel data regressions to a longitudinal and cross-industry data-set to investigate our predictions. We find that DIUs increase performance and that this effect is strengthened by the presence of digital ventures in the industry as well as the degree to which the industry relies on tangible assets. Our additional analyses provide a nuanced perspective on the implications of DIU establishments. On the base of these findings, we derive important implications for IS research about digital innovation and transformation as well as DIUs and provide recommendations for managerial practice.

Full text

Performance Implications of Digital Innovation Units
Forty-Third International Conference on Information Systems, Copenhagen 2022
1
Digital Innovation Units: An Empirical
Investigation of Performance Implications
Completed Research Paper
Max Schumm
University of Kassel
Chair of Digital Transformation
Management
Henschelstraße 2, 34109 Kassel
max.schumm@uni-kassel.de
André Hanelt
University of Kassel
Chair of Digital Transformation
Management
Henschelstraße 2, 34109 Kassel
hanelt@uni-kassel.de
Sebastian Firk
University of Groningen
Department of Accounting
Nettelbosje 2, 9747 AE Groningen
s.firk@rug.nl
Abstract
Digital innovation is both a necessary and a challenging endeavor for most firms. To
achieve progress in this regard, firms across contexts increasingly set up digital
innovation units (DIUs). Despite its popularity in practice, the prospects of this initiative
are to date unclear. Building upon dynamic capabilities theory, we hypothesize the
performance implications of DIUs and employ panel data regressions to a longitudinal
and cross-industry data-set to investigate our predictions. We find that DIUs increase
performance and that this effect is strengthened by the presence of digital ventures in the
industry as well as the degree to which the industry relies on tangible assets. Our
additional analyses provide a nuanced perspective on the implications of DIU
establishments. On the base of these findings, we derive important implications for IS
research about digital innovation and transformation as well as DIUs and provide
recommendations for managerial practice.
Keywords: Digital Innovation Units, Digital Innovation, Digital Transformation,
Dynamic Capabilities, Firm Performance, Panel Data Regression
Introduction
Competitive advantage without embracing digital innovation (DI), defined „as the creation of [and
consequent change in] market offerings, business processes, or models that result from the use of digital
technology” (Nambisan et al. 2017, p. 224), appears to be unattainable in recent times (Nambisan et al.
2017; Yoo et al. 2012). Inactivity with regards to DI exposes firms to the risk of losing customers, market
shares, and reputation to emerging digital players in their industry (Gregory et al. 2018). Yet, DI comes with
specific hurdles as firms must build the capabilities to engage in the development of new, digital market
offerings (Piccinini et al. 2015; Yoo et al. 2012). One popular initiative to overcome these hurdles is the
establishment of digital innovation units (DIUs). DIUs can be defined as autonomous entities that aid their
respective main organization (MO) in the development of digital capabilities and in the search for and
creation of new digital products, services, and processes (e.g. Raabe et al. 2020a; Svahn et al. 2017).
Although this initiative is popular in business practice (e.g. Lau et al. 2021) and has recently garnered some
attention in IS research (e.g. Jöhnk et al. 2022; Raabe et al. 2020a; Svahn et al. 2017), there is no large-
scale empirical evidence to verify its presumed beneficial impact.

Performance Implications of Digital Innovation Units
Forty-Third International Conference on Information Systems, Copenhagen 2022
2
IS research documents a variety of initiatives aimed at building capabilities for DI (e.g. Jöhnk et al. 2022).
Among other initiatives, including the installment of digital institutional entrepreneurs (e.g. Firk et al.
2021), investment in digital M&A (Hanelt et al. 2021b), and the formation of external partnerships (Chanias
et al. 2019), DIUs are of special interest for research and practice (e.g. Barthel et al. 2020; Svahn et al. 2017).
Compared to other initiatives, DIUs are unique since they can, as organizational entities, internally enhance
and aid the development of novel, idiosyncratic capabilities via research, production, and integration of
digital products (Barthel et al. 2020; Raabe et al. 2021). Thereby, sourcing novel capabilities internally is
considered as competitive advantage (Teece 2007), as it enables an intrinsic link to the firm's strategy and
activities (Teece 2018). Further, in contrast to other initiatives, DIUs can be characterized as structural
alterations, which represent manifestations of a firm’s digital transformation (DT), defined “as
organizational change that is triggered and shaped by the widespread diffusion of digital technologies”
(Hanelt et al. 2021a, p. 1160). However, research frequently leaves unexplained if DIUs are more than a
media phenomenon or signaling effort by the respective firms and whether they achieve substantial and
quantifiable benefits (e.g. Mayer et al. 2021; Raabe et al. 2020b). Further, research reveals significant
hurdles and tensions when it comes to integrating DIUs' digital outputs in a pre-digital context (Svahn et
al. 2017). Relatedly, a practitioner study, for instance, describes a significant number of closures (10%) on
over 250 monitored European DIUs (Lau et al. 2021). Thus, while previous research on DIUs provides
valuable qualitative insights into their areas of activities, objectives (Raabe et al. 2021), and purpose
(Barthel et al. 2020; Fuchs et al. 2019), we have less empirical evidence on their performance impacts, their
ramifications for building competitive advantage and associated driving forces (Mayer et al. 2021).
To build and sustain competitive advantage in the digital era, a solid foundation of digital capabilities is
necessary (Yoo et al. 2012), which must be compatible with the new logics underlying digital technologies
(Hund et al. 2021). Karimi and Walter (2015) point out that building digital capabilities requires so called
dynamic capabilities, which foster the creation and modification of ordinary (daily business) capabilities
(Teece 2007). Dynamic capabilities are defined “as the firm’s ability to integrate, build, and reconfigure
internal and external competences to address rapidly changing environments” (Teece et al. 1997, p. 516)
and are associated with three key elements of sensing, seizing, and transforming (Teece 2007). They are
described as the fundamental criteria that determine whether a business can survive or thrive in an era of
growing digital challenges, turbulence, and pressure (Ellström et al. 2021). In line with prominent examples
of IS research on the subjects of DI and DT, we apply the established perspective of dynamic capabilities to
understand the influence of DIUs (Karimi and Walter 2015; Steininger et al. 2022; Warner and Wäger
2019).
An organizational foundation consisting of "distinct skills, processes, procedures, organizational structures,
decision rules, and disciplines" is required to enable the emergence and advancement of dynamic
capabilities (Steininger et al. 2022). Initial research associates the establishment of DIUs with the
emergence of dynamic capabilities since DIUs’ desired objectives, entrepreneurial setup, and post-
bureaucratic organizational setting can be related to the required characteristics to build dynamic
capabilities (Hellmich et al. 2021; Warner and Wäger 2019). Further, DIUs may be used to modernize firms'
ordinary capabilities for the digital era (Hellmich et al. 2021; Warner and Wäger 2019), as indicated by
prior DIU research (Fuchs et al. 2019; Hellmich et al. 2021). Despite the fact that some research suggests
that DIUs can build and increase their MO's dynamic capabilities (Hellmich et al. 2021), other studies
highlight challenges and hurdles with regard to their integration (Svahn et al. 2017). In the absence of large-
scale empirical investigations regarding these assertions, our study aims to clarify this pertinent instance.
Further, there is a dearth of empirical data demonstrating the dynamic capability framework's favorable
impacts in DI contexts when built and realized in a distinct organizational entity (Ellström et al. 2021;
Karimi and Walter 2015; Teece 2018; Warner and Wäger 2019). Additionally, recent research on DIUs is
mostly based on the distilled expertise of employees directly associated with DIUs (e.g. Holotiuk and
Beimborn 2019; Raabe et al. 2020a). This may result in an exaggerated and glossed-over - thus maybe
biased – positive impression of DIUs’ advantages, which is currently not validated nor confirmed by
quantifiable empirical investigations. By leveraging independent and quantifiable empirical data, we seek
to overcome this research gap on the true impacts of DIUs. Aiming to close these gaps, our investigation
focuses on the following research question: How does the establishment of DIUs influence firm
performance?
To address this question, we employ panel data regressions to a longitudinal dataset of Standard & Poor’s
500 (S&P 500) firms in the years 2011 and 2019. We observed that establishing a DIU is associated with

Performance Implications of Digital Innovation Units
Forty-Third International Conference on Information Systems, Copenhagen 2022
3
higher future firm performance on average. Further, our findings suggest that the degree of tangible assets
and the presence of digital ventures in the industry environment positively moderates the influence of DIUs
on firm performance. Finally, additional tests reveal that the influence on firm performance may stem from
three developments, each of which is intimately linked to the dynamic capability perspective. Specifically,
we discovered a beneficial effect (1) of DIUs on the number of produced digital patents (sensing capability),
(2) of DIUs on the number of digital market offerings (seizing capability), and (3) of DIUs on the
firm’s digital transformation (transforming capability), showing that the establishment of DIUs can serve
as a valuable foundation to build and realize dynamic capabilities.
This research contributes to the increasing body of knowledge on DIUs by presenting the first large-scale
empirical investigation on their performance implications (Mayer et al. 2021; Raabe et al. 2020a). We
further provide insights about moderating contextual factors that positively influence DIUs’ performance
implications. In addition, we contribute to the framework of dynamic capabilities by providing an indicator
of how they are built and realized within an organizational entity (i.e., a DIU) (Ellström et al. 2021; Karimi
and Walter 2015; Teece 2007). By providing applicable metrics to assess the advantages of DIUs and
dynamic capabilities our additional analysis contributes to both streams of literature. Additionally, our
results contribute to both DI and DT research by demonstrating and examining the impact of one initiative
in depth (Hanelt et al. 2021a; Vial 2019). Finally, this study has significant implications for management
practice.
Background
Digital Innovation and Transformation
Given the pervasiveness of digital technology in practically every product and service, being at the forefront
of DI is becoming increasingly critical for competitiveness (Nambisan et al. 2017; Yoo et al. 2012).
Specifically, new digital ventures compete for customers, market share, and reputation, also in established
industries, and may thereby alter or disrupt the competitive landscape (Gregory et al. 2018). Examples are
the famous Kodak-Moment (Lucas and Goh 2009) or the disruptive introduction of mobility providers such
as UBER or Lyft into the car sector (Bohnsack et al. 2021). In established contexts, DI requires “a new
organizational form that departs dramatically from traditional industrial production” (Berente 2020, p.
92), since creating DI puts pressure on both the offered products and services and institutionalized
organizational forms (Yoo et al. 2012). Accordingly, organizational determinants (Hund et al. 2021), such
as the establishment of a DIU, can have a favorable influence on enabling and fostering DI (e.g. Barthel et
al. 2020; Svahn et al. 2017). Therefore, while DIUs are dedicated to create DI (Svahn et al. 2017a), their
establishment, as structural alteration of the MO, represents a manifestation of a firm’s digital
transformation (DT) (Göbeler 2020; Jöhnk et al. 2022). DT propels organizations into a continuous state
of change and reconfiguration with the purpose of successfully developing and supplying DI as well as
preserving continual adaptability within digital business ecosystems (Hanelt et al. 2021a). Considering the
turbulent and challenging context of DT, triggered by the widespread diffusion of digital technology (Hanelt
et al. 2021a), researchers argue that the dynamic capabilities perspective provides an effective lens through
which to perceive a firm’s transformation endeavor (Warner and Wäger 2019). Firms have been urged to
develop dynamic capabilities capable of swiftly creating, implementing, and transforming business models
in order to remain ahead in the emerging digital environment (e.g. Karimi and Walter 2015; Teece 2018).
More broadly, the continual adaptation, implementation, and refinement of an organization's business and
organizational structures necessitates the development of dynamic capabilities (Teece 2007; Teece 2018).
Due to the inherent dynamism of digital innovation (Nambisan et al. 2017; Yoo et al. 2012), the demand for
dynamic capabilities is more of an ongoing requirement than a selective one (Warner and Wäger 2019). To
stay competitive in the digital age, firms must provide organizational foundations to build and realize
dynamic capabilities (Karimi and Walter 2015; Steininger et al. 2022). Despite their crucial role in firms'
DI and DT activities, there is a scarcity of research on how firms can build and realize dynamic capabilities
(Warner and Wäger 2019).
Digital Innovation Units
DIUs have been illustrated as an important initiative to carry out digital research and development (Dremel
et al. 2017; Svahn et al. 2017), to enable the implementation of a DT strategy (Chanias et al. 2019), to

Performance Implications of Digital Innovation Units
Forty-Third International Conference on Information Systems, Copenhagen 2022
4
overcome DT inertia (Haskamp et al. 2021c) or to enable organizational hybridity (Schumm and Hanelt
2021). While different types of DIUs may have distinct attributes, they all share an emphasis on digital
technologies at their core (Fuchs et al. 2019; Raabe et al. 2020a), which distinguishes this phenomenon
from those discussed in other streams of literature, such as new business incubators (e.g. Gassmann and
Becker 2006). DIUs operate with a focus on flexibility (Barthel et al. 2020), with creative methodologies
(Svahn et al. 2017), using agile principles and a close customer orientation in the context of DI (Fuchs et al.
2019; Raabe et al. 2020a). Thus, establishing a DIU to foster DI activities can be understood as a firm’s
effort to change its organizational structure, routines, and culture to become more adaptable to digital
environments (Göbeler 2020; Schumm and Hanelt 2021). Recent IS research identifies multiple objectives
for DIUs, all of which stem from a strong emphasis on DI, including research, selection, development, and
dissemination of digital offerings (Holotiuk and Beimborn 2019). Creating and providing novel digital
products or services, DIUs’ outcomes entail various beneficial ramifications for their MOs (Holotiuk and
Beimborn 2019; Raabe et al. 2021). Research associates the creation and integration of novel digital
offerings into an existing product landscape as superior to gain new customers and market share (Gregory
et al. 2018), which can be related to an improvement in (financial) performance (Hanelt et al. 2021b).
Further, by participating actively in DI, DIUs are related to building up digital capabilities, for example in
the field of big data analytics (Dremel et al. 2017). Consequently, the establishment of DIUs is deemed to
serve as a foundation to build and realize dynamic capabilities (Fuchs et al. 2019; Jöhnk et al. 2022; Raabe
et al. 2021). Concluding, DIUs’ purpose includes the research on or the development of digital offerings
(Barthel et al. 2020; Fuchs et al. 2019), the enablement of the MO’s DT endeavors (Dremel et al. 2017), as
well as, at a higher abstraction level, the building and realization of new, dynamic capabilities (Hellmich et
al. 2021; Svahn et al. 2017). Yet, despite the fact that research on DIUs is generally associated with a positive
effect (e.g. Dremel et al. 2017), these studies have not determined whether the establishment of the
observed DIUs results in a measurable improvement (e.g. Mayer et al. 2021; Raabe et al. 2020b). Further,
the establishment of DIUs into a pre-digital environment has been shown to have a generally beneficial
impact (Smith and Beretta 2021), yet numerous tensions and hurdles during their integration have been
noted (Svahn et al. 2017). Relatedly, there have been reports about relatively high closure rates among DIUs
in business practice (Lau et al. 2021). Thus, large-scale empirical insights on DIUs’ performance
implications are currently missing and would contribute to a greater understanding of their underlying
impacts and benefits. Regarding our theoretical lens, additional research is required to determine whether
DIUs can serve their MOs as an organizational foundation for building and realizing dynamic capabilities.
Hypotheses Development
Prior research on DI has shown the increased need for digital capabilities (Hund et al. 2021; Yoo et al. 2012).
These can be built by drawing on dynamic capabilities, aiding organizations in adjusting their ordinary
capabilities to the digital era (Karimi and Walter 2015). Dynamic capabilities, with their key elements of
sensing, seizing and transforming (Teece 2007), enable organizations to adapt their business models to
substantial changes in their environment (Steininger et al. 2022). The presence of dynamic capabilities is
therefore said to be related to competitive advantage in changing and turbulent business environments
(Teece 2007; Teece et al. 1997). Organizational foundations to build and realize dynamic capabilities should
be internally incorporated, not bought, to enable an intrinsic link to the firm's strategy and activities (Teece
2018). Therefore, DIUs may be thought of as foundations where dynamic capabilities may be built, and
realized (Hellmich et al. 2021), since they can provide the structural, processual, and entrepreneurial
framework deemed essential for the emergence of dynamic capabilities (Teece 2007).
Moreover, DIUs as dedicated DI engines are aimed at searching for, developing, and integrating novel
digital offerings into the existing product landscape of an incumbent (Svahn et al. 2017). Therefore,
expanding the current business with digital products and services is considered to generate additional
income and thus can be advantageous to the firm's performance (Hanelt et al. 2021b). Concluding, DIUs
may have a positive influence on firm performance, yet their inherent advantages may vary, since they are
based on conditions that render benefits particularly valuable (Göbeler 2020; Holotiuk 2020). Their impact
may be amplified, particularly in contexts where the internal hurdles and external threats to participating
in DI are comparably high.
Performance Implications. The continuous and increased utilization of digital technology places firms
in the position of requiring novel capabilities to generate new digital products and services to foster firm

Performance Implications of Digital Innovation Units
Forty-Third International Conference on Information Systems, Copenhagen 2022
5
performance (Yoo et al. 2012). Establishing DIUs may therefore have a positive effect on firms’ performance
(Hellmich et al. 2021; Warner and Wäger 2019), as they can serve as foundation to build and realize
dynamic capabilities, which may lead to competitive advantage (Teece 2007; Teece et al. 1997). More
specifically, via sensing and seizing, DIUs may “serve as a fast lane for developing new ideas and products“
(Haskamp et al. 2021a, p. 1), resulting in the exploration of new client groups and the establishment of new
or advanced revenue streams, which can improve the firm's performance (Hanelt et al. 2021b). Smith and
Beretta (2021), for example, describe a DIU inside an industrial pump manufacturer that is responsible for
identifying (sensing) and developing (seizing) digital components with the goal of extending existing
physical goods to create "intelligent pumps”. These new, digital features provide the MO with a plethora of
technical data that can be used to optimize internal development and maintenance processes, while also
paving the way for new business models in platform-based ecosystems (Smith and Beretta 2021). This
digital enhancement can result in new digital income streams, increasing the overall firm performance
(Hanelt et al. 2021b). Further, the establishment of DIUs may facilitate critical DT efforts (Barthel et al.
2020; Fuchs et al. 2019), such as by supplying and supporting the MO with digital knowledge and expertise
(Fuchs et al. 2019; Raabe et al. 2021). Dremel et al. (2017) present an example of DIUs developing and
integrating new (digital) capabilities in the field of big data analytics in order to change (transforming) the
organizations' capabilities to current digital requirements. Another example illustrates how a DIU may
serve as a best practice for new methods of working, cultural changes, and DT strategy enhancements
(Chanias et al. 2019). These transforming capabilities pave the way for integrating novel capabilities and
DIs into established business models (Barthel et al. 2020; Fuchs et al. 2019). To conclude, we suggest that
DIUs contribute to performance improvement by producing DIs and assisting in their integration. Further,
they can be seen as the foundation to build and realize vital dynamic capabilities (e.g. Fuchs et al. 2019;
Hellmich et al. 2021) relevant to increasing competitive advantage (Teece 2007; Teece et al. 1997). This
leads us to the following hypothesis:
H1: The establishment of DIUs is positively associated with future firm performance.
Contextual conditions
The need to change in response to the diffusion of digital technology and therefore the importance of
dynamic capabilities for firm performance may differ from firm to firm depending on the respective
contextual conditions (Singh and Hess 2020). DIU outcomes are characterized to be especially
advantageous when external environmental "pressure [...] is exacerbated by pervasive digitalization"
(Göbeler 2020, p. 10) and internal hurdles to creating DI are significant (Barthel et al. 2020). Thus, DIUs’
influence on firm performance and thus their inherent advantages may vary. Their impact on firm
performance may be influenced in two ways: (1) internally, by the current organizational hurdles to develop
DIs, characterized by the extent to which an industry relies on physical, tangible assets; and (2) externally,
by environmental concerns prompted by new actors threatening traditional industrial roles.
Internal DT Hurdles. The internal DT hurdles may be linked to the (in)ability to produce digital
components and the prospect of digitizing a firm's present business model (Firk et al. 2022). DI is
considered as a central requirement for all industries and firms to preserve or expand competitive
advantage in recent times (Nambisan et al. 2017; Yoo et al. 2012). Yet, the ability and effort to create them
seems to differ between industries (Hanelt et al. 2021b; Karimi and Walter 2015; Piccinini et al. 2015).
While business models depending on non-physical, intangible assets are rather associated with digitization
(Yoo et al. 2012), business models based substantially on tangible assets come with specific pitfalls and
hurdles in DI as to the semantic difference in the underlying knowledge and the static nature of physical
materiality (e.g. Hanelt et al. 2021b; Piccinini et al. 2015). Given the fact that firms in physical-product
industries have a long history of manufacturing physical, non-digital goods (Bohnsack et al. 2021; Hylving
and Schultze 2020), they often lack critical digital capabilities and organizational preconditions necessary
to build DI (Yoo et al. 2012). This is plausible, given their "strong hardware legacy, where development
processes and organizational structures are typically adjusted and reflected in the physical product, i.e., the
car" (Hylving and Selander 2012, p. 2). It presents distinct internal hurdles since there is not only a scarcity
of knowledge for obtaining digital capabilities, but also a scarcity of organizational flexibility and autonomy
for innovative activities (Smith and Beretta 2021). Accordingly, we argue that in industries that rank high
in tangible assets, firms face more internal hurdles in creating DI (Dremel et al. 2017; Svahn et al. 2017).
Thus, DIUs have a greater impact on performance as the need for the dynamic capabilities they instantiate
is comparatively higher (Teece 2007; Teece 2018) and digital products and services are more vital to be

Performance Implications of Digital Innovation Units
Forty-Third International Conference on Information Systems, Copenhagen 2022
6
provided by DIUs (Dremel et al. 2017; Svahn et al. 2017). Concluding, firms with a strong emphasis on
tangible assets, may experience a positive influence on DIUs’ performance implications. This leads us to the
following hypothesis:
H2a: The level of tangible assets positively moderates the positive association between the establishment
of DIUs and future firm performance.
External DT Threats. The external DT threats may be linked to the presence of digital ventures entering
the respective industry (Firk et al. 2022). The existence of digital ventures underscores the critical need to
create and develop DI (e.g. Huang et al. 2017). Previous research demonstrated their influence on
established industries via large-scale empirical studies and observed a triggering impact on incumbents to
respond (Firk et al. 2021; Zapadka et al. 2022). Flexible and agile digital ventures that can leverage digital
technologies successfully and swiftly (Huang et al. 2017) are infiltrating current competitive environments
or creating new ones that supersede previously dominating ones (Skog et al. 2018). These digital
ventures can rapidly test new business models and scale up successful concepts globally (Kelestyn et al.
2017). Accordingly, the presence of digital ventures equals an increase in environmental change that drives
the importance and benefits of dynamic capabilities (Teece 2018). Firms could particularly profit from DIUs
digital outcomes since they are described as an explicit initiative when “new market entrants from the
technology industry increase the competitive pressure” (Holotiuk 2020, p. 7). When organizations face
larger external threats as a result of a strong presence of digital ventures, DIUs' ability to predict digital
disruptive challenges may create more value, since firms must respond even quicker. Being able to supply
digital offerings on their own may have a positive impact on firm performance in sectors where digital
competition is comparably strong (Holotiuk 2020). Furthermore, when the external environment is more
threatening, DIUs may have a stronger influence on competitive advantage since the demand for the
dynamic capabilities they can aid to build (Teece 2007; Teece 2018) and the MO’s pressure to accept and
integrate DIUs’ digital outcomes is comparatively higher. Concluding, we argue that DIUs’ impact on firm
performance may be even higher, in the presence of intense digital competition. As a consequence, the
following hypothesis is proposed:
H2b: The number of digital ventures in the industry positively moderates the positive association between
the establishment of DIUs and future firm performance.
Methodology
Sample Selection
We focus on a longitudinal sample of S&P 500 firms over the years 2011 to 2019. Data related to the
establishment of DIUs was hand-collected from press releases and media articles in newswires on the
LexisNexis database. We further retrieved data on firm performance, tangible assets, and further financial
and industry characteristics from Datastream, whereas data on digital ventures was retrieved from the
CrunchBase database. Based on the availability of this data, we end up with a final sample covering 618
firms and 4,823 firm-years.
Variables
Dependent Variable – Firm performance. To measure firm performance, we use, similar to prior
research in the IS economics literature (Banker et al. 2011; Ho et al. 2017; Van Peteghem et al. 2019), a
market-based performance measure. Specifically, we focus on the firm’s stock returns (market return 2fy)
over the next two years (t+1 and t+2). We decided for a market-based measure instead of an accounting-
based measure as accounting measures are often criticized for not appropriately capturing the performance
of digital business models (e.g., due to long lead times) (Govindarajan et al. 2018). To calculate the two-
year stock returns, we use the RI variable as specified in Datastream.
Independent Variable – DIU establishment. To measure the establishment of DIUs, we exploited
press and company announcements regarding DIUs. In order to gather all relevant search strings, we
carefully studied the current DIU literature. It contains several distinct descriptions, archetypes, and names
(e.g. Barthel et al. 2020; Raabe et al. 2020a). We derived a list of 12 unique descriptions for DIUs, like
"Digital Lab", "Digital Innovation Lab”, “Innovation Lab”, “Innovation Hub”, “Digital Innovation Hub"
etcetera, to compile a list of their descriptions that considers different terms. We then manually hand-

Performance Implications of Digital Innovation Units
Forty-Third International Conference on Information Systems, Copenhagen 2022
7
collected and coded all press releases and business announcements from newswires on the LexisNexis
database, based on our S&P 500 firms, between 2011 and 2019. We verified the DIU's assignment to the
MO to ensure that it is directly related to the MO and not, for example, mentioned in terms of a consortium
of firms cooperating. Our final DIU variable (DIU) takes the value of one if a MO established a DIU in the
respective year, and zero otherwise.
Moderator variable – Internal DT hurdles. To measure internal DT hurdles, we focus on the firm’s
reliance on tangible assets such as machinery or other equipment. Similar to prior research (Antia et al.
2010; Custódio et al. 2019), we measure the reliance of a firm on tangible assets by dividing the level of
property, plant and equipment by a firm’s total assets (Tangible assets).
Moderator variable – External DT threats. To measure the threats that digital ventures place on
incumbent firms, we follow prior research and use a measure that captures the number of new digital
ventures per industry incumbent. Specifically, Firk et al (2021) extracted all ventures in the Crunchbase
database and evaluated whether it were a digital venture by evaluating the descriptions. They further
classified each digital venture’s industry affiliation. In line with prior studies (e.g. Zapadka et al. 2022), we
carefully matched this industry-level construct to our firm-level abstraction. Based on the classification of
Firk et al. (2021), we counted the number of new digital ventures in each industry and year and divided this
number by the number of industry incumbents (i.e., all firms that had been listed for more than 3 years in
the Datastream database) to measure our external DT threat variable (digital ventures). Finally, similar to
Firk et al. (2021) we calculated the average of this variable over a 3 year period (t, t-1 and t-2) as we expect
that the pressure of new ventures may evolve over time.
Control variables. We selected several variables that may drive both the firm’s decision to establish a
DIU and firm performance to control for confounding effects. We screened prior research focus on the
relation between IT-related constructs and firm performance to select control variables (Banker et al. 2011;
Ho et al. 2017; Mithas et al. 2012). We include firm size as the natural logarithm of net sales. We then
included a firm’s leverage proxied by total debt in relation to total assets and firm risk as the standard
deviation of the return of equity over the last 3 years divided by its mean. We included R&D intensity
measured as R&D expenditure by net sales. Moreover, we captured the current performance by including
ROA as a measure for profitability and sales growth as the 3 year growth in sales prior to the respective
year. Finally, we followed Mithas et al. (2012) an included the yearly average of the industry’s Tobin’s q as
well as the industry concentration calculated via a Herfindahl index. We used for both a Fama-French 30
classification.
Empirical Method
To examine performance implications of establishing DIUs, we follow the IS economics literature by
estimating a firm-fixed effects regression (Mithas et al. 2012; Pan et al. 2016). This decision was supported
by running a Hausman (1978) test. The firm-fixed effects regression assigns each firm an individual effect
to control for any firm-specific unobservable factors. We further specified the firm-fixed effects model as a
treatment effects model. The treatment effect model is a type of selection model that can be used to mitigate
self-selection concerns (Certo et al. 2016). To apply the treatment effect model, we first employ a probit
model that estimates the likelihood of establishing a DIU and calculate an inverse Mills’ ratio based on its
results. Second, we include the inverse Mills’ ratio as an additional variable in our firm-fixed effects model
to control for self-selection. The literature outlines the importance of including an appropriate exclusion
criteria in the first-stage probit model (Lennox et al. 2012). We use the peer industry average of firms that
have at least one DIU established as our exclusion variable. We believe that this variable could be
appropriate because peer industry average of firms could influence the firms’ decisions to establish a DIU
(relevance condition). We further expect that this variable is rather exogenous from a focal firm’s
performance and that it is valid to exclude this variable from our second stage estimation (exclusion
restrictions). In the first stage probit model (untabulated), we further include all our second stage control
variables and added the market performance in the two years before the DIU establishment. Based on the
results, we then calculated the inverse Mill’s ratio and estimate the following regression to test our first
hypothesis.
I.
𝑀𝑎𝑟𝑘𝑒𝑡'𝑅𝑒𝑡𝑢𝑟𝑛'2𝑓𝑦!"#$%&"$ %'( = '𝛼 + '𝛽&(𝐷𝐼𝑈)!"$ +' 𝛾&𝐼𝑀𝑅!"$ ' + '𝛾'
(
𝐶𝑂𝑁𝑇𝑅𝑂𝐿𝑆
)
!"$ ' + 𝑌
$+ 𝜂!+' 𝜀!"$ '

Performance Implications of Digital Innovation Units
Forty-Third International Conference on Information Systems, Copenhagen 2022
8
Besides the dependent variable (Market Return 2fy) and our independent variable (DIU) the item
𝐶𝑂𝑁𝑇𝑅𝑂𝐿𝑆
represents a matrix reflecting the selected control variables. The item
𝐼𝑀𝑅
stands for the
inverse-Mill’s ratio that addresses selection concerns. The remaining items are the year fixed effects
(𝑌
$)
,
the constant term (α), the firm-fixed effects (
𝜂!
), and the error term (
𝜀!"$)
.
To test Hypothesis 2a and Hypothesis 2b, we further add our moderator variables as well as interaction
terms between the moderator variables and the DIU variable. The moderator variables were standardized
before adding them to the model. We use the following regression to estimate Hypothesis 2a and Hypothesis
2b.
II.
𝑀𝑎𝑟𝑘𝑒𝑡'𝑅𝑒𝑡𝑢𝑟𝑛'2𝑓𝑦!"#$%&"$ %'( = '𝛼 + '𝛽&(𝐷𝐼𝑈)!"$ + 𝛽'(𝐷𝐼𝑈 ∗𝑇𝑎𝑛𝑔𝑖𝑏𝑙𝑒'𝑎𝑠𝑠𝑒𝑡𝑠)!"$ ' + 𝛽)(𝐷𝐼𝑈 ∗
𝐷𝑖𝑔𝑖𝑡𝑎𝑙'𝑣𝑒𝑛𝑡𝑢𝑟𝑒𝑠)! "$ +' 𝛽*(𝑇𝑎𝑛𝑔𝑖𝑏𝑙𝑒'𝑎𝑠𝑠𝑒𝑡𝑠)!"$ + 𝛽+(𝐷𝑖𝑔𝑖𝑡𝑎𝑙'𝑣𝑒𝑛𝑡𝑢𝑟𝑒𝑠)!"$ + '𝛾&𝐼𝑀𝑅!"$ '+
'𝛾'
(
𝐶𝑂𝑁𝑇𝑅𝑂𝐿𝑆
)
!"$ ' + 𝑌
$+ 𝜂!+' 𝜀!"$ '
Findings
Descriptive Results
Table 1 illustrates the diffusion of DIUs in our sample. The establishment of DIUs has grown significantly
in recent years. While we observed barely twenty DIU establishment in the first three years of our sample
period (i.e., 2011-2013), we observed seventy DIU establishments in the last three years of our sample
period (i.e., 2017-2019). In total, we found 131 DIU establishment over the entire sample period.
Furthermore, about fifteen percent of all firms established at least one DIU in our sample period.
Table 1. DIU Establishments
Year
Obs.
DIU
establishments
DIU
establishments (%)
DIU users
DIU users (%)
2011
536
9
1.7%
9
1.7%
2012
544
7
1.3%
15
2.8%
2013
544
4
0.7%
18
3.3%
2014
542
11
2.0%
27
5.0%
2015
541
14
2.6%
38
7.0%
2016
532
16
3.0%
46
8.6%
2017
523
24
4.6%
57
10.9%
2018
531
29
5.5%
74
13.9%
2019
530
17
3.2%
81
15.3%
Total
4823
131
2.7%
365
7.6%
Table 2 displays the mean, standard deviation, as well as the bivariate correlations of all our regression
variables. The means and SDs are comparable to prior research. All continuous variable have been
winsorized at the 1st and 99th percentiles thus mitigating any issue due to exceptional outliers. The
correlations further indicate a positive and small correlation between DIUs and firm performance.
Moreover, the correlations did not indicate multicollinearity issues as all correlations were below 0.5 and
most far below 0.2.
Table 2. Descriptive statistics and correlation matrix
Variable names
Mean
SD
1
2
3
4
5
6
7
8
9
10
11
12
13
(1) Market ret. 2fy
0.13
0.20
1.00
(2) DIU
0.03
0.16
-0.01
1.00
(3) Tangible assetsb
0.28
0.27
-0.09
-0.07
1.00
(4) Digital venturesb
0.05
0.09
0.03
0.01
-0.22
1.00
(4) Firm sizea
15.67
1.29
-0.11
0.17
-0.18
-0.08
1.00

Performance Implications of Digital Innovation Units
Forty-Third International Conference on Information Systems, Copenhagen 2022
9
(5) Leverage
0.26
0.18
-0.06
-0.03
0.20
0.03
-0.02
1.00
(6) R&D intensity
0.02
0.04
0.13
0.06
-0.17
0.07
-0.19
-0.16
1.00
(7) Sales growth
0.33
0.54
0.14
-0.03
-0.03
0.04
-0.14
-0.05
0.23
1.00
(8) ROAc
7.46
7.18
0.18
0.03
-0.15
0.06
-0.02
-0.05
0.07
0.01
1.00
(9) Firm risk
0.28
1.65
0.02
0.02
0.00
-0.03
0.02
-0.02
0.00
0.02
0.02
1.00
(10) Ind. Tobin’s q
1.97
0.87
0.12
0.08
-0.43
0.42
-0.03
0.04
0.36
0.08
0.33
0.00
0.40
1.00
(11) Ind. concentr.
0.12
0.09
-0.03
-0.01
-0.14
0.07
0.15
0.11
-0.10
-0.03
0.07
-0.01
0.16
0.10
1.00
Notes:
N = 4,823; a = natural logarithm; b = standardized variable is used in the regression model; c = measured in percent; all
variables except from the binary variable DIU have been winsorized at the 1st and 99th percentiles; correlations exceeding the value of
0.02 are significant at the 10 percent level.
Regression Results
The results of firm fixed effects models estimating the influence of establishing a DIU and future market
returns are displayed in Table 3. Model 1 shows the direct influence of establishing a DIU. We find a positive
and significant coefficient (p < 0.05) for the DIU variable. In practical terms, the establishment of DIU is
associated with a 9 percentage point increase in market returns over the next two years. Hence, the results
support our first hypothesis suggesting that establishment of DIUs is positively associated with firm
performance.
Model 2 to Model 4 of Table 3 test the moderating influence of DT hurdles (tangible assets) as well as the
moderating influence of DT threats (digital ventures). With regard to internal DT hurdles, we find a positive
interaction term between the level of tangible assets and DIU establishment in Model 2 (p < 0.05) and
Model 4 (p < 0.01). In practical terms, the results of Model 2 suggest that a 1SD increase in the level of
tangible assets strengthens the positive association between establishing DIUs and future market returns
by 38 percent. This result supports our hypothesis H2a suggesting that firms with more tangible assets that
have a higher DT hurdle could particularly benefit from the establishment of DIUs. With regard to external
DT threats, we find a positive interaction term between digital ventures and DIU establishment in Model 3
(p < 0.1) and Model 4 (p < 0.01). The results of Model 4 indicate that a 1SD increase in the level of digital
ventures in the industry strengthens the positive association between establishing DIUs and future market
returns by 18 percent. Hence, this results supports hypothesis H2b suggesting that firms in industries with
more digital ventures, and thus a higher external DT threat, benefit more from DIUs’ presence. We also
plotted both interaction terms (untabulated). The plot supported the positive moderation of tangible assets
as well as digital ventures.
Table 3: Regression results: DIU establishments and firm performance
Model
1
2
3
Dependent Variable
Market return 2fy
Market return 2fy
Market return 2fy
Market return 2fy
Method
Firm-fixed effects
Firm-fixed effects
Firm-fixed effects
Firm-fixed effects
DIU (H1)
0.090**
0.121***
0.090**
0.124***
(0.034)
(0.009)
(0.036)
(0.007)
DIU * Tangible assets (H2a)
0.045**
0.054**
(0.039)
(0.010)
DIU * Digital ventures (H2b)
0.016*
0.021***
(0.059)
(0.003)
Tangible assets
-0.038***
-0.038***
(0.007)
(0.007)
Digital ventures
0.006
0.005
(0.732)
(0.772)
Inverse mills ratio
0.029*
0.034*
0.030*
0.035**
(0.087)
(0.051)
(0.082)
(0.048)
Firm size
-0.076***
-0.085***
-0.076***
-0.085***
(0.000)
(0.000)
(0.000)
(0.000)

Performance Implications of Digital Innovation Units
Forty-Third International Conference on Information Systems, Copenhagen 2022
10
Leverage
0.012
0.006
0.011
0.006
(0.817)
(0.902)
(0.826)
(0.911)
R&D intensity
-0.667*
-0.615
-0.662*
-0.613
(0.079)
(0.116)
(0.083)
(0.119)
Sales growth
-0.003
-0.004
-0.003
-0.004
(0.792)
(0.713)
(0.787)
(0.708)
ROA
0.003***
0.003***
0.003***
0.003***
(0.000)
(0.000)
(0.000)
(0.000)
Firm risk
0.002
0.002
0.002
0.002
(0.301)
(0.287)
(0.310)
(0.294)
Industry Tobin’s q
0.101***
0.101***
0.100***
0.100***
(0.000)
(0.000)
(0.000)
(0.000)
Industry concentration
-0.003
-0.005
0.012
0.009
(0.985)
(0.979)
(0.947)
(0.960)
Intercept
1.088***
1.199***
1.088***
1.200***
(0.000)
(0.000)
(0.000)
(0.000)
Firm effects
yes
yes
yes
yes
Year effects
yes
yes
yes
yes
N
4823
4823
4823
4823
Adjusted R-Square
0.308
0.310
0.308
0.310
Notes:
*p<0.10; **p<0.05; ***p<0.01. Robust standard errors clustered at the firm-level. P-values in parentheses. The moderator
variables Tangible assets and Digital ventures are standardized (i.e., mean value of zero and a standard deviation of one) to
facilitate the interpretation of the moderating effects.
Robustness of Results
Endogeneity concerns. The establishment of DIUs and firm performance could be simultaneously
driven by an unobservable factor. To address such endogeneity concerns, we used in our main analysis a
firm-fixed effects regression and a correction factor that accounts for self-selection issues. To further tackle
endogeneity concerns, we also employ three matching procedures: propensity score matching, coarsened
exact matching as well as entropy balancing. All these matching procedures aim to account for the fact that
observations with DIU establishments (i.e., treatment group) are not randomly distributed in our dataset.
Matching procedures adjust the control group (i.e., observation with no DIU establishments) to reduce
systematic differences in observable characteristics between the treatment and the control group. The idea
is that this likely reduces differences in unobservable characteristics between the two groups (Hainmueller
and Xu 2013; Shipman et al. 2017) and thus endogeneity issues. For the propensity score matching, we
specified a one-to-one match and thus matched the control observation with the most similar propensity
score to a respective treatment observation (Shipman et al. 2017). In the coarsened exact matching, we
divided our controls into “bins” (quartiles) and then matched treatment observations and control
observations that are in the same bin (Iacus et al. 2012). Both the propensity score matching and the
coarsened exact matching reduces the number of total observations. Entropy balancing, however, does not
exclude dissimilar control observations. It reweights the observations in the dataset to diminish any
systematic differences between observations in the control and the treatment group (Hainmueller and Xu
2013). For all matching procedures, we used all of our control variables as well as the market performance
over the last two years. Table 4 displays regression results on the matched samples. The results consistently
support our previous findings.
Table 4: Regressions results on matched samples
PS matched
CEM matched
Entropy balanced
Model
1
2
3
4
5
6
Dependent Variable
Market
return 2fy
Market
return 2fy
Market
return 2fy
Market
return 2fy
Market
return 2fy
Market
return 2fy
Method
OLS
OLS
OLS
OLS
OLS
OLS
DIU (H1)
0.046***
0.065***
0.037***
0.040***
0.032***
0.051***
(0.005)
(0.003)
(0.004)
(0.003)
(0.008)
(0.001)
DIU * Tangible assets (H2a)
0.055*
0.021**
0.055**
(0.093)
(0.074)
(0.015)

Performance Implications of Digital Innovation Units
Forty-Third International Conference on Information Systems, Copenhagen 2022
11
DIU * Digital ventures (H2b)
0.050***
0.022**
0.028***
(0.002)
(0.046)
(0.001)
Tangible assets
-0.034*
-0.004
-0.027*
(0.094)
(0.539)
(0.072)
Digital ventures
-0.011
0.001
0.013**
(0.448)
(0.948)
(0.016)
Control variables
yes
yes
yes
yes
yes
yes
Year effects
yes
yes
yes
yes
yes
yes
N
255
255
2501
2501
4823
4823
Adjusted R-Square
0.440
0.462
0.375
0.376
0.371
0.390
Notes:
*p<0.10; **p<0.05; ***p<0.01. Robust standard errors clustered at the firm-level. P-values in parentheses. The moderator
variables Tangible assets and Digital ventures are standardized (i.e., mean value of zero and a standard deviation of one) to
facilitate the interpretation of the moderating effects.
Alternative specifications. We run several (untabulated) robustness tests. First, we alternatively run a
random effects model and found consistent results. Second, we tested alternative time-windows for our
market return variable and found that the effects also hold for a shorter window of one-year as well as a
longer window of three years. Third, we also tested Tobin’s Q as alternative performance measure and found
similar results to those reported. Finally, we also tested an alternative specification of the DIU variable. We
used the number of DIUs as well as the first DIU establishment as alternative variable and found again
support for our hypotheses.
Additional Analysis
Our investigation revealed that establishing DIUs has a considerable effect on the MOs’ firm performance.
Our arguments for the positive performance effect of DIUs are based on the idea that DIUs might be able
to serve as a foundation to build and realize dynamic capabilities (Fuchs et al. 2019; Hellmich et al. 2021).
Specifically, this implies that DIUs may lead to increased dynamic capabilities with regard to their key
elements of - sensing, seizing, and transforming (Teece 2007). Hence, in additional tests, we focused on
three outcome variables that could indicate an increase in terms of these three key elements.
First, DIUs can be thought of as an organizational entity tasked with performing research and screening for
(external) digital possibilities (Barthel et al. 2020; Raabe et al. 2020a). From the dynamic capability
perspective, this may position firms in a leading role when it comes to sensing the environment for possible
future business opportunities, putting them one step ahead of the competition in terms of gaining new
digital expertise and meeting new customer expectations (Teece 2007; Teece 2018). Building this dynamic
capability likely has a beneficial influence on digital knowledge (Warner and Wäger 2019). Hence, we follow
earlier research using digital patents as a proxy for a firm's internal digital knowledge base (e.g. Firk et al.
2022; Hanelt et al. 2021b) and test whether DIUs are positively associated with the filings of digital patents
in the next two years, and if they can serve as a foundation to build the dynamic capability of sensing. We
revised all patent profiles for digital ones and measured them similar to Firk et al. (2022) as the natural
logarithm of the number of digital patents in the next two years. The data is collected from the USPTO
database.
Second, DIUs aim to assist their MOs by developing new digital components, such as digital products and
services that may result in new income streams (Svahn et al. 2017). Applying the dynamic capability
perspective, seizing emerging opportunities is critical when attempting to build and execute new (digital)
business models and may lead to a competitive advantage (Teece 2018). Regarding our theoretical lens,
earlier studies observed a theoretical connection between this dynamic capability and digital product and
service innovation (Limaj et al. 2016). We hence investigate whether DIUs are associated with more digital
market offerings (i.e., digital products and services) in the next two years, and if they can serve as a
foundation to build the dynamic capability of seizing. We focus on press releases and news articles
regarding product and service releases and then classified these product and service releases as a new digital
market offering or not. Again, we counted the number of digital market offerings and calculated the natural
logarithm of digital market offerings over the next two years.
Third, establishing DIUs can be associated with facilitating a firm's internal organizational and cultural
transformation (Göbeler 2020; Jöhnk et al. 2022). When it comes to aligning and enhancing existing
business models and capabilities with new ones, the dynamic capability of transforming may be seen as

Performance Implications of Digital Innovation Units
Forty-Third International Conference on Information Systems, Copenhagen 2022
12
unique benefit in gaining competitive advantage (Teece 2018). This capability is associated with both the
adoption of new technologies (Karimi and Walter 2015) and the change of existing business models and
structures (Ellström et al. 2021). Hence, we decided to further investigate the influence of DIUs on the
firm’s DT in the next two years and investigate if DIUs, can serve as a foundation to build the dynamic
capability of transforming. To measure DT, we follow the approach of prior studies (Li et al. 2021a), by
applying textual analysis to the firm’s annual report. Specifically, we measure the increase of digital-related
words (e.g., identified by using a word embedding model; see, Li et al. (2021b)) in the firm’s 10k report from
the time t to t+2. We first calculate the relative occurrence of digital words in a 10k report and then measure
the relative increase over the next two years.
The results of these tests are summarized in Table 5. We find that DIUs are associated with more digital
patents (Model 1) as well as more digital market offerings (Model 2) in the next two years. We also find a
positive and significant association between DIUs and the firm’s digital transformation (Model 3). Hence,
DIUs’ outcomes can be empirically associated with the subdivided outcomes of dynamic capabilities -
sensing, seizing, and transforming which leads us to the assumption that they aid in the creation and
realization of dynamic capabilities.
Table 5. Additional analyses: DIU establishments and subdivided DIU outcomes
Model
1
2
3
Dependent Variable
Digital patents 2fy
Digital market
offerings 2fy
Digital transformation
2fy
Method
Firm-fixed effects
Firm-fixed effects
Firm-fixed effects
DIU
0.636***
0.220**
0.196*
(0.004)
(0.026)
(0.066)
Firm effects
yes
yes
yes
Control variables
yes
yes
yes
Year effects
yes
yes
yes
N
3,745
4,832
3,759
Adjusted R-square
0.06
0.10
0.09
Notes
:
*p<0.10; **p<0.05; ***p<0.01. Robust standard errors clustered at the firm-level. P-values in parentheses. The control variables
included are the same as the ones used in Table 3. We also included the two moderator variables digital ventures and tangible
assets in the models. Digital Patents 2fy refers to the natural logarithm of the number of digital patents (see Firk et al. 2022) in
the next two years. For digital patents 2fy, we could only rely on data until 2019, which reduced our sample significantly. Digital
Market Offerings 2fy refers to the natural logarithm of the number of digital service and product releases in the next two years.
Digital transformation 2fy refers to the change in the relative number of digital words in a firm’s 10k report. We could here only
rely on data until the year of 2020 and to firm-years where we could match a 10k-report in the EDGAR database.
Discussion of Findings
In this study, we use a cross-industry and longitudinal dataset to provide empirical insights about the
performance implications of DIU establishments. We first demonstrate a wide distribution of DIUs in our
sample. With a DIU presence in about 15% percent of all 618 firms examined, we may deduce that DIUs are
not a theoretical phenomenon, but a real-world initiative embraced by a diverse variety of businesses across
several industries and sectors (Haskamp et al. 2021a; Raabe et al. 2020b). Second, we find a positive
association between DIU establishment and future firm performance. This demonstrates how establishing
a DIU may help firms address digital challenges and threats (Haskamp et al. 2021b; Mayer et al. 2021) and
provides first empirical support for conceptual and qualitative literature emphasizing benefits of DIUs in
digitalizing environments (Fuchs et al. 2019; Raabe et al. 2021).
Additionally, our research uncovers moderators on the performance implications of DIUs, as defined
internally by the degree to which an industry depends on physical goods and high tangible assets, and
externally by the existence of digital ventures. According to our results, a high amount of tangible assets
positively moderates the firm performance implications of DIUs. In accordance with previous research
(Svahn et al. 2017), we propose that enterprises with a long history of developing physical goods seem to
encounter greater hurdles when it comes to developing DIs within their present organizational structures
(Hanelt et al. 2021b; Piccinini et al. 2015). These firms have institutionalized tightly coupled product
architectures and hierarchical organizational structures (Hylving and Schultze 2020). Yet, to build DI and

Performance Implications of Digital Innovation Units
Forty-Third International Conference on Information Systems, Copenhagen 2022
13
their loosely coupled product architectures (Yoo et al. 2010), more adaptable, agile and flexible work
environments as well as new digital capabilities are required (Vial 2019; Yoo et al. 2012). We argue that
DIUs can serve as a foundation to build capabilities and structures for DI (Hellmich et al. 2021; Raabe et
al. 2020a). Accordingly, firms in physical product-based industries may profit even more from DIUs
performance implications. Furthermore, we discovered that firms in industries with a higher number of
digital ventures benefit more from DIUs' effects on firm performance. According to previous research, firms
in those competitive environments appear to face greater competitive threats to their business models,
since digital ventures can leverage digital technologies more quickly (Huang et al. 2017) and infiltrate
current industries or create new ones that supersede previously dominating ones (Skog et al. 2018). As
customers get more used to the impact of digital technologies provided by digital ventures (Lucas and Goh
2009; Vial 2019), firms in such industries may benefit even more from DIU outcomes and gain greater
profit from creating their own DIs. Our findings indicate that both internal and external influences have a
favorable impact on DIUs’ effects on firm performance. This may be related to the assumptions underlying
our theoretical perspective. According to the dynamic capability framework, its competitive advantage is
particularly advantageous in environments with significant turbulence and pressure for change (Teece
2018). Both the internal and external factors increase the need to change, which elevates the value of
dynamic capabilities. Taken together, the main effect and the moderating effects lend support to the idea
that DIUs can be considered as a foundation to build and realize dynamic capabilities in the digital era and
can corroborate that dynamic capabilities positively influence firm success in an environment of digital
threats and turbulence (Ellström et al. 2021; Steininger et al. 2022).
Our additional analysis further supports the view that DIUs lead to the building and realization of dynamic
capabilities. Our findings suggest that the beneficial impact of establishing DIUs may be threefold. First,
our research demonstrates a positive influence on firms' digital patents. This may be related to DIUs’ ability
to build sensing capabilities (Teece 2007). Second, our data shows a significant positive association between
the establishment of DIUs and future digital market offerings. This can be related to DIUs’ aiding in building
seizing capabilities (Teece 2007). Third, we can observe a positive influence on the digital transformation
of firms that have established DIUs. It can be related to DIUs’ ability to serve as a foundation to gain
transforming capabilities (Teece 2007). We can relate all three subdivided outcomes to our applied
perspective of dynamic capabilities. Our findings imply that, in addition to a more theoretical correlation
between DIUs and dynamic capabilities (Hellmich et al. 2021), there is also a correlation between their
objectives and outcomes, connected to sensing, seizing, and transforming capabilities (Teece 2007; Teece
2018), leading us to the assumption that DIUs can serve as a foundation to build and realize dynamic
capabilities.
Implications
Our findings contribute to prior IS research with regards to DIUs, DI, and DT. First, regarding the DIU
research stream, we identified several implications. Our study shows a significant beneficial influence on
firm performance via the establishment of DIUs. We suggest that DIUs are more than a management trend
or a nice-looking practice to pretend DI activities, but help to create real value. Apart from a number of
qualitative studies based on (possibly biased) employee impressions about DIUs (Haskamp et al. 2021b),
this quantitative analysis demonstrates the positive correlation between the establishment of DIUs and firm
performance. It allows us to acknowledge that DIUs and MOs can incorporate mechanisms to overcome the
anticipated hurdles and tensions resulting from their establishments (Svahn et al. 2017a). Moreover, our
additional analysis allows us to provide a more nuanced picture of DIUs' beneficial influence on firm
performance and their role in serving as a foundation to build and realize dynamic capabilities (Hellmich
et al. 2021). The results may be further interpreted as an attempt to provide additional metrics to
measure the effectiveness of DIUs (Haskamp et al. 2021b; Mayer et al. 2021). Both on a broad level, as
measured by firm performance, and on a more granular level, as measured by digital patents, digital market
offerings, and DT. Further, the additional analysis results contribute quantitatively to the literature's
categorization of DIU objectives and outcomes, which had hitherto been presented rather qualitatively (e.g.
Barthel et al. 2020). We contribute to the DIU literature by shedding light on possible positive moderators
on the effects of DIUs (Barthel et al. 2020; Raabe et al. 2021). Thereby, we illuminate the influence of
external and internal environments on DIUs and their effectiveness. We see that industries with a high
number of digital ventures are more receptive to DIUs’ outcomes and gain even more from their competitive
advantage. We notice the same notion in physical product-based industries, although they face distinct

Performance Implications of Digital Innovation Units
Forty-Third International Conference on Information Systems, Copenhagen 2022
14
tensions in combining digital and physical components (Hylving and Schultze 2020), termed as digital-
physical paradox (Piccinini et al. 2015). Finally, we contribute to the DIU literature by highlighting their
widespread presence within the US. Despite a large number of European researchers and a strong research
emphasis in Europe (e.g. Dremel et al. 2017; Lau et al. 2021; Svahn et al. 2017), our data demonstrates that
DIUs are not a uniquely European phenomenon.
Second, our results contribute to the IS literature with distinct implications for the DI stream. The
substantial number of established DIUs in our samples qualifies them as a commonly utilized initiative
targeted at tackling DI-related challenges. While other initiatives, such as digital M&A (Hanelt et al. 2021b),
have been empirically investigated, this step was missing with regards to DIUs. Our additional analysis
further contributes to the stream of DI, particularly by empirically examining different positions in
theoretical stage models (e.g. Kohli and Melville 2019). By observing three kinds of outcomes, digital
patents, digital market offerings, and digital transformation, we may argue that they are comparable with
three stages of DI such as discovery (associated with digital patents), development (associated with digital
market offerings), diffusion (associated with digital transformation) - and impact (associated with overall
firm performance) by Fichman et al. (2014). Our final contribution to the DI literature regards our
moderators. By emphasizing the importance of DIUs in product-based industries, our findings may prove
the difficulties inherent in developing DIs within their hierarchical, traditional organizational contexts
(Piccinini et al. 2015; Svahn et al. 2017).
Moreover, our results contribute to research on digital transformation. In this regard, DIUs may be
considered as another step in the overall “shift towards malleable organizational designs which enable
continuous adaptation” (Hanelt et al. 2021a, p. 3) by providing a glimpse of how to design hybrid
organizational forms suitable for the digital age (Schumm and Hanelt 2021). Given the beneficial
performance implications of DIUs, we hypothesize that this phenomenon will be a persistent one, rather
than a passing management fad. Nonetheless, this may result in ongoing challenges and stumbling blocks
in contemporary hybrid organizations (Piccinini et al. 2015; Svahn et al. 2017).
Regarding our theoretical lens, our results contribute to the literature of dynamic capabilities. First, we
present a large-scale and practical example of how dynamic capabilities can be built and realized in the
context of DI and DT (Steininger et al. 2022). Additionally, our research establishes a baseline of actual
evidence on the competitive advantages associated with dynamic capabilities in today's turbulent,
digitalized economy (Warner and Wäger 2019). Although we examined only one approach how dynamic
capabilities can be built and realized, we may offer applicable metrics to evaluate the advantages of
the dynamic capabilities sensing (digital patents), seizing (digital market offerings), and transforming (DT)
in the digital age (Ellström et al. 2021; Warner and Wäger 2019). Finally, the beneficial impact on firm
performance demonstrates that the perspective of dynamic capabilities is significantly relevant and
applicable in today's digital environment (Steininger et al. 2022).
Finally, our study has significant managerial implications. To begin with, amid the prevalent conversation
around DI and DT, managers seeking viable initiatives to mitigate challenges and achieve progress may
consider establishing DIUs. Furthermore, managers determining the applicability of DIUs to their
environment and circumstances may consider our moderators as an appropriate variable. Although we
found that DIUs have a general positive effect on firm performance, we propose that managers consider
their firm's present internal and external environments. Establishing DIUs may be particularly beneficial
in physical-based product industries with high tangible assets, as well as in surroundings with a high
concentration of digital ventures. Finally, managers and organizations who are unsure about how to target
their DIU's objectives, and purpose may discover direction in terms of sensing for digital patents, seizing
new digital market offerings, and transforming the MO.
Limitations and Future Research
The findings of our study should be interpreted in light of the following limitations. Our sample is limited
to the United States and, thus, we cannot assure whether our findings hold outside this context. Future
research may focus on whether our findings also hold in-non U.S contexts. It would also be interesting to
consider differences in culture or the national digital infrastructure (Firk et al. 2021) as a country-level
moderator for DIU performance effects. In this context, also more nuanced measures for DT threats such
as investments in digital ventures would be interesting to explore. Second, while our study considers, in

Performance Implications of Digital Innovation Units
Forty-Third International Conference on Information Systems, Copenhagen 2022
15
addition to market returns, also several outcomes related to DI and DT, these proxies might not be able to
capture the full picture of DI and DT outcomes. For example, while we are able to measure the number of
digital market offerings, we cannot track the specific performance of these offerings or while digital patents
likely capture an increase in DI expertise, it may not fully capture more tacit DI expertise in the firm. Third,
despite our best efforts (i.e., firm-fixed effects, self-selection controls and matched sample analyses), we
cannot completely rule out all endogeneity concerns. Fourth, our work makes no attempt to consider other
theoretical interpretations. For example, assumptions regarding the establishment costs of DIUs and their
real financial effects give intriguing insights for future study, particularly when considering transaction
costs and resource dependence. Further, more nuanced characteristics of DIUs could be tested to gain a
deeper understanding of the detailed effects of DIUs. Finally, we noticed that although the outcomes of
DIUs impact their MO, we currently know very little about particular processes and practices for joint
and collaborative digital innovation, innovation transfer, and digital transformation endeavors between
DIUs and MOs. Further empirical investigation is required to close this knowledge gap about how DIUs
and MOs collaborate and develop their intertwined relationship and overcome potential (paradoxical)
tensions.
Acknowledgements
We thank Christin Schäfer for her assistance in parts of the data collection.
References
Antia, M., Pantzalis, C., and Park, J. C. 2010. "Ceo Decision Horizon and Firm Performance: An Empirical
Investigation," Journal of corporate finance (16:3), pp. 288-301.
Banker, R. D., Hu, N., Pavlou, P. A., and Luftman, J. 2011. "Cio Reporting Structure, Strategic Positioning,
and Firm Performance," MIS Quarterly, pp. 487-504.
Barthel, P., Fuchs, C., Birner, B., and Hess, T. 2020. "Embedding Digital Innovations in Organizations: A
Typology for Digital Innovation Units".
Berente, N. 2020. "Agile Development as the Root Metaphor for Strategy in Digital Innovation Handbook
of Digital Innovation." Cheltenham, UK: Edward Elgar Publishing.
Bohnsack, R., Kurtz, H., and Hanelt, A. 2021. "Re-Examining Path Dependence in the Digital Age: The
Evolution of Connected Car Business Models," Research Policy (50:9).
Certo, S. T., Busenbark, J. R., Woo, H. s., and Semadeni, M. 2016. "Sample Selection Bias and Heckman
Models in Strategic Management Research," Strategic Management Journal (37:13), pp. 2639-2657.
Chanias, S., Myers, M., and Hess, T. 2019. "Digital Transformation Strategy Making in Pre-Digital
Organizations: The Case of a Financial Services Provider," The Journal of Strategic Information
Systems (28), pp. 17-33.
Custódio, C., Ferreira, M. A., and Matos, P. 2019. "Do General Managerial Skills Spur Innovation?,"
Management Science (65:2), pp. 459-476.
Dremel, C., Herterich, M., Wulf, J., Waizmann, J.-C., and Brenner, W. 2017. "How Audi Ag Established Big
Data Analytics in Its Digital Transformation," MIS Quarterly Executive (16), pp. 81-100.
Ellström, D., Holtstrom, J., Berg, E., and Johansson, C. 2021. "Dynamic Capabilities for Digital
Transformation," Journal of Strategy and Management.
Fichman, R., Dos Santos, B., and Zheng, Z. 2014. "Digital Innovation as a Fundamental and Powerful
Concept in the Information Systems Curriculum," MIS Quarterly, pp. 329-353.
Firk, S., Gehrke, Y., Hanelt, A., and Wolff, M. 2022. "Top Management Team Characteristics and Digital
Innovation: Exploring Digital Knowledge and Tmt Interfaces," Long Range Planning.
Firk, S., Hanelt, A., Oehmichen, J., and Wolff, M. 2021. "Chief Digital Officers: An Analysis of the Presence
of a Centralized Digital Transformation Role," Journal of Management Studies (58:7), pp. 1800-1831.
Fuchs, C., Barthel, P., Herberg, I., Berger, M., and Hess, T. 2019. "Characterizing Approaches to Digital
Transformation: Development of a Taxonomy of Digital Units".
Gassmann, O., and Becker, B. 2006. "Towards a Resource-Based View on Corporate Incubators,"
International Journal of Innovation Management, pp. 19-45.
Göbeler, L. S., D.; Hukal, P. 2020. "Initiating Ambidexterity through Digital Innovation Labs." In
Proceedings of the 28th European Conference on Information Systems.

Performance Implications of Digital Innovation Units
Forty-Third International Conference on Information Systems, Copenhagen 2022
16
Govindarajan, V., Rajgopal, S., and Srivastava, A. 2018. "Why Financial Statements Don’t Work for Digital
Companies," Harvard Business Review, pp. 2-6.
Gregory, R. W., Kaganer, E., Henfridsson, O., and Ruch, T. J. 2018. "It Consumerization and the
Transformation of It Governance," MIS Quarterly (42:4), pp. 1225-1253.
Hainmueller, J., and Xu, Y. 2013. "Ebalance: A Stata Package for Entropy Balancing," Journal of Statistical
Software (54:7).
Hanelt, A., Bohnsack, R., Marz, D., and Antunes Marante, C. 2021a. "A Systematic Review of the Literature
on Digital Transformation: Insights and Implications for Strategy and Organizational Change," Journal
of Management Studies.
Hanelt, A., Firk, S., Hildebrandt, B., and Kolbe, L. M. 2021b. "Digital M&A, Digital Innovation, and Firm
Performance: An Empirical Investigation," European Journal of Information Systems (30), pp. 3 - 26.
Haskamp, T., Breitenstein, A., and Lorson, A. 2021a. "A Management Control Systems Perspective on
Digital Innovation Units,".
Haskamp, T., Lorson, A., Paula, D. d., and Uebernickel, F. 2021b. "Bridging the Gap–an Analysis of
Requirements for Performance Measurement Systems in Digital Innovation Units," International
Conference on Wirtschaftsinformatik: Springer, pp. 587-605.
Haskamp, T., Marx, C., Dremel, C., and Uebernickel, F. 2021c. "Understanding Inertia in Digital
Transformation: A Literature Review and Multilevel Research Framework," Proceedings on the
International Conference on Information Systems (ICIS).
Hausman, J. A. 1978. "Specification Tests in Econometrics," Econometrica: Journal of the econometric
society, pp. 1251-1271.
Hellmich, J., Raabe, J.-P., and Schirmer, I. 2021. "Towards a Foundational and Extensional Dynamic
Capability Perspective on Digital Innovation Units,".
Ho, J., Tian, F., Wu, A., and Xu, S. X. 2017. "Seeking Value through Deviation? Economic Impacts of It
Overinvestment and Underinvestment," Information Systems Research (28:4), pp. 850-862.
Holotiuk, F. 2020. "The Organizational Design of Digital Innovation Labs: Enabling Ambidexterity to
Develop Digital Innovation," in Wi2020 Zentrale Tracks. GITO Verlag, pp. 1019-1034.
Holotiuk, F., and Beimborn, D. 2019. "Temporal Ambidexterity : How Digital Innovation Labs Connect
Exploration and Exploitation for Digital Innovation," AIS Electronic Library (AISeL).
Huang, J. C., Henfridsson, O., Liu, M. J., and Newell, S. 2017. "Growing on Steroids: Rapidly Scaling the
User Base of Digital Ventures through Digital Innovation," MIS Q. (41:1), pp. 301-314.
Hund, A., Wagner, H.-T., Beimborn, D., and Weitzel, T. 2021. "Digital Innovation: Review and Novel
Perspective," The Journal of Strategic Information Systems (30:4), p. 101695.
Hylving, L., and Schultze, U. 2020. "Accomplishing the Layered Modular Architecture in Digital
Innovation: The Case of the Car's Driver Information Module," The Journal of Strategic Information
Systems.
Hylving, L., and Selander, L. 2012. "Under the Pressure of Openness: Exploring Digital Innovation in User
Interface Design".
Iacus, S. M., King, G., and Porro, G. 2012. "Causal Inference without Balance Checking: Coarsened Exact
Matching," Political analysis (20:1), pp. 1-24.
Jöhnk, J., Ollig, P., Rövekamp, P., and Oesterle, S. 2022. "Managing the Complexity of Digital
Transformation—How Multiple Concurrent Initiatives Foster Hybrid Ambidexterity," Electronic
Markets.
Karimi, J., and Walter, Z. 2015. "The Role of Dynamic Capabilities in Responding to Digital Disruption: A
Factor-Based Study of the Newspaper Industry," Journal of Management Information Systems.
Kelestyn, B., Henfridsson, O., and Nandhakumar, J. 2017. "Scaling the User Base of Digital Ventures
through Generative Pattern Replication: The Case of Ridesharing,".
Kohli, R., and Melville, N. P. 2019. "Digital Innovation: A Review and Synthesis," Information Systems
Journal (29:1), pp. 200-223.
Lau, F., Lindemann, L. S., Münch, L.-T., Sindemann, T., and Wiegad, M. 2021. "Corporations on the Track
of Start-Ups 2021". 5th Edition. Hamburg: Infront Consulting & Management.
Lennox, C. S., Francis, J. R., and Wang, Z. 2012. "Selection Models in Accounting Research," The
accounting review (87:2), pp. 589-616.
Li, J., Li, M., Wang, X., and Thatcher, J. B. 2021a. "Strategic Directions for Ai: The Role of Cios and Boards
of Directors," MIS Quarterly (45:3).
Li, K., Mai, F., Shen, R., and Yan, X. 2021b. "Measuring Corporate Culture Using Machine Learning," The
Review of Financial Studies (34:7), pp. 3265-3315.

Performance Implications of Digital Innovation Units
Forty-Third International Conference on Information Systems, Copenhagen 2022
17
Limaj, E., Bernroider, E. W., and Choudrie, J. 2016. "The Impact of Social Information System Governance,
Utilization, and Capabilities on Absorptive Capacity and Innovation," Information & Management.
Lucas, H. C., and Goh, J. M. 2009. "Disruptive Technology: How Kodak Missed the Digital Photography
Revolution," The Journal of Strategic Information Systems (18:1), pp. 46-55.
Mayer, S., Haskamp, T., and De Paula, D. 2021. "Measuring What Counts: An Exploratory Study About the
Key Challenges of Measuring Design Thinking Activities in Digital Innovation Units,".
Mithas, S., Tafti, A., Bardhan, I., and Goh, J. M. 2012. "Information Technology and Firm Profitability:
Mechanisms and Empirical Evidence," MIS Quarterly, pp. 205-224.
Nambisan, S., Lyytinen, K., Majchrzak, A., and Song, M. 2017. "Digital Innovation Management:
Reinventing Innovation Management Research in a Digital World," MIS Quarterly (41).
Pan, Y., Huang, P., and Gopal, A. 2016. "Board Independence and Firm Performance in the It Industry: The
Moderating Role of New Entry Threats," MIS Quarterly.
Piccinini, E., Hanelt, A., Gregory, R., and Kolbe, L. 2015. "Transforming Industrial Business: The Impact of
Digital Transformation on Automotive Organizations," In: 43rd Hawaii International Conference on
System Sciences.
Raabe, J.-P., Drews, P., Horlach, B., and Schirmer, I. 2021. "Towards an Intra-and Interorganizational
Perspective: Objectives and Areas of Activity of Digital Innovation Units".
Raabe, J.-P., Horlach, B., Drews, P., and Schirmer, I. 2020a. "Digital Innovation Units: Exploring Types,
Linking Mechanisms and Evolution Strategies in Bimodal It Setups".
Raabe, J.-P., Horlach, B., Schirmer, I., and Drews, P. 2020b. "‘Forewarned Is Forearmed’: Overcoming
Multifaceted Challenges of Digital Innovation Units".
Schumm, M., and Hanelt, A. 2021. "Transformational Dynamics - Systemizing the Co-Evolution of
Organizational Forms and Information Systems,". Proceedings on the International Conference on
Information Systems (ICIS).
Shipman, J. E., Swanquist, Q. T., and Whited, R. L. 2017. "Propensity Score Matching in Accounting
Research," The Accounting Review (92:1), pp. 213-244.
Singh, A., and Hess, T. 2020. "How Chief Digital Officers Promote the Digital Transformation of Their
Companies," Strategic Information Management. Routledge, pp. 202-220.
Skog, D. A., Wimelius, H., and Sandberg, J. 2018. "Digital Disruption," Business & Information Systems
Engineering (60:5), pp. 431-437.
Smith, P., and Beretta, M. 2021. "The Gordian Knot of Practicing Digital Transformation: Coping with
Emergent Paradoxes in Ambidextrous Organizing Structures," Journal of Product Innovation
Management (38:1), pp. 166-191.
Steininger, D., Mikalef, P., Pateli, A., and Guinea, A. 2022. "Dynamic Capabilities in Information Systems
Research: A Critical Review, Synthesis of Current Knowledge, and Recommendations for Future
Research," Journal of the Association for Information Systems (23), pp. 447-490.
Svahn, F., Mathiassen, L., and Lindgren, R. 2017. "Embracing Digital Innovation in Incumbent Firms: How
Volvo Cars Managed Competing Concerns," MIS Quarterly (41), pp. 239-253.
Teece, D. J. 2007. "Explicating Dynamic Capabilities: The Nature and Microfoundations of (Sustainable)
Enterprise Performance," Strategic management journal (28:13), pp. 1319-1350.
Teece, D. J. 2018. "Business Models and Dynamic Capabilities," Long range planning (51:1), pp. 40-49.
Teece, D. J., Pisano, G., and Shuen, A. 1997. "Dynamic Capabilities and Strategic Management," Strategic
management journal (18:7), pp. 509-533.
Van Peteghem, M., Joshi, A., Mithas, S., Bollen, L., and De Haes, S. 2019. "Board It Competence and Firm
Performance," Fortieth International Conference on Information Systems.
Vial, G. 2019. "Understanding Digital Transformation: A Review and a Research Agenda," The Journal of
Strategic Information Systems (28).
Warner, K. S., and Wäger, M. 2019. "Building Dynamic Capabilities for Digital Transformation: An Ongoing
Process of Strategic Renewal," Long range planning (52:3), pp. 326-349.
Yoo, Y., Boland, R. J., Lyytinen, K., and Majchrzak, A. 2012. "Organizing for Innovation in the Digitized
World," Organization Science (23:5), pp. 1398-1408.
Yoo, Y., Henfridsson, O., and Lyytinen, K. 2010. "The New Organizing Logic of Digital Innovation: An
Agenda for Information Systems Research," Information Systems Research (21), pp. 724-735.
Zapadka, P., Hanelt, A., and Firk, S. 2022. "Digital at the Edge–Antecedents and Performance Effects of
Boundary Resource Deployment," The Journal of Strategic Information Systems (31:1), p. 101708.