International Journal of Information Management 59 (2021) 102335

Contents lists available at ScienceDirect

International Journal of Information Management

journal homepage: www.elsevier.com/locate/ijinfomgt

Research Article

Conceptualising value creation in data-driven services: The case of

vehicle data
Christian Kaiser a, *, Alexander Stocker a, Gianluigi Viscusi b, Michael Fellmann c,
Alexander Richter d
a
Virtual Vehicle Research GmbH, Inffeldgasse 21a, 8010, Graz, Austria
b
Imperial College Business School, South Kensington Campus, London SW7 2AZ, United Kingdom
c
University of Rostock, Albert-Einstein-Straße 22, Rostock, Germany
d
Wellington School of Business and Government, Rutherford House, 23 Lambton Quay, Wellington, New Zealand

A R T I C L E I N F O A B S T R A C T

Keywords: The creation of data-driven services generates new value streams, leading to the emergence of new actors and
Value creation ultimately to new market configurations. In the automotive industry, the data generated by vehicles during use
Data-driven services paves the way for new types of data-driven services. Based on interviews with eleven prominent experts of the
Automotive
Central European automotive industry, we identify key actors in establishing vehicle data-driven services and
Data sharing
Conceptual model
their data sharing relationships. We illustrate both in a conceptual multi-actor model for value creation in vehicle
data-driven services and evaluate it in the context of six real-life cases. Our study adopts an ecosystem
perspective and marks an important step towards the systematic design of a conceptual multi-actor model for
vehicle data-driven value creation that can help to guide next research endeavours in data-driven service
development.

1. Introduction and motivation other road users, vehicle manufacturers or service developers. Con­
nected vehicles enable the possibility to develop data-driven services
The ongoing transition towards a digitalised world also affects pri­ such as remote vehicle diagnostics or interactive trip analytics (Kuschel,
marily physical industries (Hanelt, Piccinini, Gregory, Hildebrandt, & 2008; Papatheocharous, Frecon, Kaiser, Festl, & Stocker, 2018). Thus,
Kolbe, 2015). Due to its long tradition in catering to a basic human need the digital transformation offers new players outside the automotive
– mobility, the automotive domain stands out in particular (Piccinini, sector the opportunity to enter this traditionally closed ecosystem
Hanelt, Gregory, & Kolbe, 2015). Traditionally, businesses within the (Athanasopoulou, Bouwman, Nikayin, & de Reuver, 2016). Among
automotive domain were geared towards offering goods (e.g. selling those, we find major companies like Tesla, Google, or Apple (Wittmann,
manufactured vehicles as the main product) and product-related ser­ 2017) and start-ups like vin.li and Zendrive.com who create data-driven
vices (e.g. selling spare parts and conducting maintenance work). services related to digital asset tracking, vehicle health or driving safety
However, digitalisation has led the automotive industry to think (Stocker, Kaiser, & Fellmann, 2017). Yet, it remains a challenge for those
differently, as vehicles become increasingly connected and capture a lot start-ups to translate their technical innovations into commercially
of data about themselves and their environment (Swan, 2013, 2015). successful product offerings.
This captured vehicle data eventually paves the way for new types of Despite these disruptive changes caused by digitalisation, the core
data-driven services (Bridgelall, Lu, Tolliver, & Xu, 2018; De Winter, industrial product of the automotive industry, the vehicle, cannot be
Dodou, Happee, & Eisma, 2019; Pillmann, Wietfeld, Zarcula, Raugust, & digitised entirely (Piccinini et al., 2015). Instead, it will be com­
Alonso, 2017; Pütz, Murphy, Mullins, & O’Malley, 2019). plemented by both traditional and data-driven services (Kaiser, Festl,
Consequently, vehicles are increasingly becoming part of an auto­ Pucher, Fellmann, & Stocker, 2019). Declining revenues from vehicle
motive ecosystem that includes not only drivers and passengers but also sales can be compensated by additional income from the monetarization

* Corresponding author.
E-mail addresses: [email protected] (C. Kaiser), [email protected] (A. Stocker), [email protected] (G. Viscusi), michael.fellmann@uni-
rostock.de (M. Fellmann), [email protected] (A. Richter).

https://doi.org/10.1016/j.ijinfomgt.2021.102335
Received 19 May 2020; Received in revised form 16 February 2021; Accepted 16 February 2021
Available online 13 March 2021
0268-4012/© 2021 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
C. Kaiser et al. International Journal of Information Management 59 (2021) 102335

of vehicle data (Bertoncello et al., 2016; Davenport, Pacheco, & Priest­ The theoretical gap addressed in our paper is the lack of conceptual
ley, 2020; Seiberth & Gründinger, 2018). However, it remains no less of models that can unravel the underlying value chain (actors and data
a challenge for incumbent companies in the automotive industry to fully sharing relationships) when establishing vehicle data-driven services.
embrace such digital innovation (Svahn, Mathiassen, & Lindgren, 2017). We thereby address the calls of researchers (Parvinen, Pöyry, Gus­
Data-driven services are services that support customers’ decision- tafsson, Laitila, & Rossi, 2020) to closely examine data-driven value
making processes by providing data and analytics to create value for creation and ecosystems (Wamba, Akter, Edwards, Chopin, & Gnanzou,
the customer (Schüritz, Farrell, Wixom, & Satzger, 2019). The provision 2015). Our designed conceptual model aims to link actors with specific
of data-driven services is often, but not necessarily, accompanied by a steps within the data value chain. Thereby, it seeks to help organisations
physical product equipped with sensors for digital connection to other that choose to become part of the automotive ecosystem to better un­
products and information systems-IS (Beverungen, Lüttenberg, & Wolf, derstand their role, relationships, and opportunities for data-driven
2018; Tomiyama, Lutters, Stark, & Abramovici, 2019). Although this service provision better. Thus, our model supports the design of
digital transformation in the automotive domain is underway (Kuhnert, vehicle data-driven services by introducing the most relevant actors, and
Stürmer, & Koste, 2018; McKinsey, 2016), little is known about the most data flows, ultimately leading to data-driven value creation. As of now,
relevant actors and their data sharing relationships to deliver the model preserves the different perspectives of key actors while
value-added services based on exploiting vehicle data. Especially in the addressing the research gap that key concepts and relations regarding
advent of big data, it is even more important than ever to understand the data-driven value creation are nowadays insufficiently explored, as we
characteristics of data-based or data-driven value creation (Lim et al., observe it in the research on vehicle data-driven services. As to these
2018; Schüritz et al., 2019). arguments, they are grounded in our research within the large-scale
We put our focus on the automotive domain as their industrial-age research project EVOLVE funded by the European Commission in the
core product cannot be digitized completely (Piccini et al., 2015). Horizon 2020 framework programme involving 19 key partners from 11
Furthermore, automotive is one of the most important industries related European countries from the automotive, big data, cloud and
to non-digital artefacts (i.e. vehicles) (Henfridsson, Mathiassen, & high-performance computing worlds aiming at better exploitation of big
Svahn, 2009). It is, however, worth noting that the automotive sector data.
has begun to experiment with vehicle telematics solutions and con­ The remainder of this paper is structured as follows: Section 2 pre­
nected car initiatives since a few years (Svahn et al., 2017). sents the framework that guided our research approach and paper
The primary goal of our research is to investigate ways through structure. Section 3 embraces the theoretical foundations of this paper.
which (small and big1) vehicle data can spawn new data-driven services After this, in Section 4, we elaborate on the data collection including
and to provide a framework to structure and evaluate vehicle data- results from eleven expert interviews and their sketching activities.
driven value creation. We argue that improved knowledge about key These views are unified and serve as the basis for our conceptual model
actors and their data sharing relationships will contribute to a better presented in Section 5. We provide the results of the final evaluation of
understanding of vehicle data-driven value creation. Accordingly, a our model in Section 6, and discuss our findings in Section 7 before we
fundamental starting point for our research is to map those actors that summarise and conclude our paper in Section 8.
will have a crucial role in data sharing and then design how data sharing
relationships can connect them. Thus, our paper addresses the following 2. Research framework and paper structure
three research questions:
We address the lack of conceptual models that can unravel the un­
• Which actors play a key role in vehicle data-driven service generation? derlying vehicle data value chain (actors and data sharing relationships)
• How do data sharing relationships connect those actors to enable value in establishing data-driven services. Our research framework is guided
creation? by the design-science paradigm (Hevner, March, Park, & Ram, 2004),
• How can a conceptual model illustrate both actors and their data sharing with its three research cycles (Hevner, 2007): relevance cycle, design
relationships? cycle, and rigor cycle (Fig. 1). Design-science extends "the boundaries of
human and organizational capabilities by creating new and innovative
The identification of the most relevant ecosystem actors provides the artifacts" (Hevner et al., 2004). In our case, the innovative artifact is the
foundation for better understanding their data sharing relationships and conceptual model for value creation in vehicle data-driven services. This
interdependencies, allowing us to design a conceptual model of data- research framework allowed us to obtain the different perspectives of
driven value creation. Conceptual models are abstract representations key stakeholders (researchers, users, clients, sponsors, and practitioners)
of some subject matter, which serve to promote communication and while studying complex problems.
common understanding between stakeholders, thereby improving the As part of the relevance cycle (Section 3), we conducted a literature
prospects for successful information system development and use (Wand review of well-regarded scientific electronic databases extended through
& Weber, 1993). Conceptual models are mostly of a graphic nature and backward and forward search regarding our application context (value
usually contain a visual arrangement of modelling constructs in the form creation in vehicle data-driven services) and theoretical lens (ecosys­
of graphical symbols and text (Bera, Soffer, & Parsons, 2019). Besides tems). Existing theory on the value of data-driven services and data
supporting communication, they contribute to a better understanding of value chains was used as theoretical input within the design phase.
a particular domain and provide input for the information systems In the design cycle, we first took a participatory approach to build a
design process (Wand & Weber, 2002). practice-based, conceptual model, capturing the individual views of
eleven experts from the automotive domain on value creation in vehicle
data driven services. We complemented our interviewing approach with
1
simple graphical design activities, letting experts draw sketches on key
Although big data and big data analytics are definitely important, we would actors and their data sharing relationships (Section 4). We consolidated
like to emphasize that we do not focus on research on the adoption on big data
the individual expert views in a conceptual model of value creation in
analytics. Many vehicle data-driven services are based on "small data" (using
vehicle data-driven services, applied conceptual modelling (Wand &
only a few data points of a single signal): For instance, services that can detect
safety critical situations inside a vehicle and forward this information to Weber, 2002) inspired by the concept of data value chains (e.g. Curry,
operational organizations like emergency services do not rely on big data an­ 2016; Faroukhi, El Alaoui, Gahi, & Amine, 2020; Latif, Saeed, Hoefler,
alytics. In many cases, a few data points from a few signals are sufficient to Stocker, & Wagner, 2009) and presented our artifact: a unified con­
generate a data-driven service with high added value as for instance mentioned ceptual model for the data value creation process consisting of three
by the interviewed data marketplace provider. parts, (i) actors involved, (ii) key ecosystem actors and (iii) data sharing

2
C. Kaiser et al. International Journal of Information Management 59 (2021) 102335

Fig. 1. Research Approach and Paper Structure.

relationships between the actors. We established proof-of concept before businesses and society (Bharadwaj, El Sawy, Pavlou, & Venkatraman,
we proceeded to evaluating our model with real-life cases (Sections 4 2013; Chesbrough & Spohrer, 2006; Lim et al., 2018; Lusch & Nambisan,
and 5). 2015; Spohrer & Maglio, 2008). Among the different available defini­
In the rigour cycle (Sections 5 and 6), we performed conceptual tions, we adopt here the concise summary provided by Spohrer and
model evaluations by applying our model to in total six real-life appli­ Maglio (2008, p. 241) defining service as "pay for performance in which
cation cases enabled by vehicle data and established its proof-of-value. value is coproduced by client and provider". This is true, for example,
After each design, we conducted an evaluation of the model that when considering information intensive services (IIS) where the "value
resulted in model revision. This resulted either in a change in model is created primarily via information interactions, rather than physical
actors, a change in data-sharing relationships, or a change in both, while and interpersonal interactions, between the customer and the provider"
the general structural design of the model remained unchanged. Our (Lim et al., 2018, p.121). Moreover, these services rely on the data that
paper only includes the sixth evaluation of our model within a real- generate the information driving the activities, making them valuable
world case, the development of a data-driven service for road surface for the final customer (Azkan, Iggena, Gür, Möller, & Otto, 2020; Kumar
quality detection to underpin its practical applicability. et al., 2013; Maass, Parsons, Purao, Storey, & Woo, 2018). Conse­
quently, value creation based on data should take into account the data
3. Theoretical foundations value chain as well as key factors, such as for example the data, the data
source, data collection, data analysis, information delivery, information
This section places our research in the context of the relevant existing on the user, the value in information use, and the provider network (Lim
literature. First, we illustrate the concept of data-driven services in the et al., 2018, p.122).
automotive domain and the rationale behind it. Second, we take a look Taking these issues into account, the role of data and information
at the literature on ecosystems which we use as a theoretical lens for our value (Attard & Brennan, 2018; Batini, Castelli, Viscusi, Cappiello, &
study. Francalanci, 2018; Brennan, Attard, Petkov, Nagle, & Helfert, 2019) is a
Table 1 shows how research on data-driven services is steadily central challenge in the competitive scenarios emerging from digital­
increasing. In total, we identified 222 papers published since 2011 in isation, in particular for understanding what concerns the evaluation of
established scientific electronic databases as AISeL, ScienceDirect, Sco­ the information capacity suitable to allow companies to the create and
pus, IEEE Xplore and ACM DL (Falagas, Pitsouni, Malietzis, & Pappas, capture value by digital assets and data-driven services (Batini et al.,
2008; Gusenbauer & Haddaway, 2020). More than 36 % of these papers 2018). Furthermore, according to Dedrick (2010) researchers have
were published in 2019. framed the impacts of the IT on environment as first-order (impacts of
We included articles that used the following terms: "value creation" ICT hardware during the product lifecycle), second-order (impacts of
or "value", and "data-driven services", or "data-based services", and ICT on other processes such as transportation or industrial production),
"automotive", or "vehicle", or "car", or "mobility". We added further pa­ and third-order effects (changes in lifestyles and economic structures).
pers on the value of data-driven services in general and on vehicle data- The latter are relevant when considering the increased use of the me­
driven services in particular by applying backwards and forward search. dia’s transformative potential of ’green’ IS on the demand side,
In what follows, we thematically discuss the main concepts from a encouraging practices such as, e.g., carpooling and ridesharing appli­
representative sample of 48 of the retrieved papers on the value of cations coupled with the Internet of Things (Malhotra, Melville, &
(vehicle) data-driven services. Watson, 2013).
Moreover, scholars from computer science and IS have also ques­
tioned, which business models could be suitable to capture the value of
3.1. Value of data-driven services data-driven services (Lim et al., 2018; Schüritz & Satzger, 2016;
Schüritz, Seebacher, & Dorner, 2017; Zolnowski, Anke, & Gudat, 2017;
In the last two decades, the service sector has seen an unprecedented Zolnowski, Christiansen, & Gudat, 2016). These contributions comple­
development, also due to the expansion of the application of Information ment the questions advanced in the field of technology management
and Communication Technologies – ICTs (Berkley & Gupta, 1994; Rai & (Hartmann, Mohamed, Niels, & Andy, 2016; Sorescu, 2017) about the
Sambamurthy, 2006) and the subsequent digital transformation of role of data-driven services in business model innovation. Additionally,
IS scholars have investigated the antecedent factors of value creation in
Table 1 connection with the big data analytics phenomenon (Günther, Rezazade
Search results (all fields) in AISeL, ScienceDirect, Scopus, IEEE and ACM (2011- Mehrizi, Huysman, & Feldberg, 2017; Mikalef, Pappas, Krogstie, &
2019). Pavlou, 2020; Surbakti, Wang, Indulska, & Sadiq, 2020; Wiener, Saun­
Year\ Database AISeL ScienceDirect SCOPUS IEEE Xplore ACM DL ders, & Marabelli, 2020). Central questions concern the big data ana­
2011− 2013 1 3 3 2 1
lytics capabilities that companies require to a) enhance organisational
2014− 2016 8 15 7 5 4 performance (Akter, Wamba, Gunasekaran, Dubey, & Childe, 2016;
2017− 2019 54 56 35 16 12 Wamba et al., 2017), b) create business value (Conboy, Dennehy, &

3
C. Kaiser et al. International Journal of Information Management 59 (2021) 102335

O’Connor, 2020; Grover, Chiang, Liang, & Zhang, 2018; Wamba et al., CarData" (BMW, 2020; Daimler, 2020a, 2020b; Volkswagen, 2018).
2015), and c) enable service innovation (Lehrer, Wieneke, vom Brocke, However, there are also approaches for vehicle data collection and use in
Jung, & Seidel, 2018), as well as d) which barriers may prevent their data-driven services that bypass vehicle manufacturers. These are the
adoption. ones pursued by tech start-ups such as dash.by, vin.li, or pace.car who
For example, Dremel, Herterich, Wulf, Waizmann, and Brenner bring their own solutions into vehicles (to create a gateway to sensor
(2017) discuss in a case study of AUDI how traditional manufacturing data) and thereby compete with the activities of vehicle manufacturers
organizations can introduce big data analytics and master related in vehicle data collection (Stocker et al., 2017).
organizational transformations. Dremel, Herterich, Wulf, and Vom Also, emerging data marketplaces, such as caruso-dataplace.com,
Brocke (2020) identify big data analytics (BDA) actualization mecha­ high-mobility.com or otonomo.io, provide another approach to
nisms from a revelatory case of a vehicle manufacturer. Akter et al. leverage vehicle data (Pillmann et al., 2017). Data marketplaces are
(2016) aim at improving the organizational performance of a company digital platforms on which data products are traded, acting as neutral
through big data analytics and proposes a hierarchical model. Grover intermediaries, and allowing others to sell their data products (Spie­
et al. (2018) explore the success of big data analytics projects with kermann, 2019). The aim of vehicle data marketplaces is to make
respect to creating strategic business value, i.e., by addressing available vehicle data collected by different brands of connected vehi­
intra-organizational aspects. Lehrer et al. (2018) propose a theoretical cles, vehicle manufacturers, fleet operators and other data providers to
model of big data analytics service innovation developed from multiple interested data-driven service developers directly or indirectly through a
cases from insurance, banking, telecommunications, and e-commerce single point of access.
that have all implemented big data analytics. Mikalef, Pappas, Krogstie,
and Giannakos (2017) recommend that more attention should be paid to 3.3. Theoretical lens: the ecosystem concept
the organizational changes that big data analytics brings and how big
data analytics should be adopted strategically. Mocker and Fonstad In general, an ecosystem describes the relationships and interactions
(2017) discuss AUDI’s challenges towards the sharing economy and how between living organisms and their environment (Briscoe & De Wilde,
AUDI has transformed its organizational structure, processes, and ar­ 2006; Schulze, Beck, & Müller-Hohenstein, 2005). To differentiate an
chitecture. Svahn et al. (2017) address, how incumbent firms embrace artificial ecosystem from a natural one, some authors add further attri­
digital innovation proposing the Volvo case study and identifying four butes to the term to qualify it, e.g. software ecosystem, business
concerns, but focusing on the perspective of the vehicle manufacturer, ecosystem or digital service ecosystem (Immonen, Ovaska, & Kalaoja,
only. Wamba et al. (2015) emphasize a lack of empirical research to 2015). However, a commonly agreed definition does not yet exist.
assess the potential of big data and provides both a literature review and Considering the field of strategy as relevant for the focus of this
case studies to present an interpretive framework to analyze the research on the automotive industry, an early definition has been pro­
different perspectives of big data as well as a taxonomy to better un­ vided by Teece (2007, p. 1325), who considers an ecosystem as "the
derstand the role of big data in value creation. Wamba et al. (2017) community of organisations, institutions, and individuals that impact
propose a big data analytics capability model, extend previous research the enterprise and the enterprise’s customers and supplies" including
by examining the direct effects of big data analytics on firm perfor­ "complementors, suppliers, regulatory authorities, standard-setting
mance. Woźniak, Valton, and Fjeld (2015) introduce a practical example bodies, the judiciary, and educational and research institutions".
for big data value creation from Volvo and share the story of building a Focusing on modularity and coordination for different types of com­
big data service for the automotive industry in a case study. These papers plementarities (in production vs. in consumption), Jacobides, Cennamo,
focus heavily on the big data analytics phenomenon and the strategic and Gawer (2018) have proposed a consolidated perspective on the
and organizational capabilities required to create value from big data ecosystem concept. They define it as "a set of actors with varying degrees
analytics. They all take a single-actor (i.e., micro) perspective. of multilateral, non-generic complementarities that are not fully hier­
archically controlled" (p. 2264). Furthermore, Adner (2016, p. 40)
3.2. Data-driven services based on vehicle data define an ecosystem as "the alignment structure of the multilateral set of
partners that need to interact in order for a focal value proposition to
The recent advances in computing infrastructure, including the materialise". Considering ’partners’ in the automotive industry, while
Internet of Things (IoTs) as well as the data generation/processing ca­ Original Equipment Manufacturer (OEM) traditionally exerted a strong
pabilities in products, have boosted the development of data-driven influence on ecosystems, this configuration is currently challenged by
services. However, those who generate and collect the data are not the digitalisation characterising new breeds of quantified vehicles and
necessarily those who develop and provide data-driven services. The new actors on the market (Stocker et al., 2017).
systematic use of the data generated in connection with vehicle use Nischak, Hanelt, and Kolbe (2017) emphasise that three components
happens in practice within complex actor-networks and ecosystems. are essential elements of digital business ecosystems: value exchange
Vehicle data paves the way to enabling novel data-driven services (innovation, information, products/services), resources (digital and
(Stocker et al., 2017) and with the current increase in connected vehicles non-digital) and actors (organisations, individuals, societies). This
this data can finally be exploited. Connected vehicles are equipped with definition can be adapted and specialised for digital automotive eco­
hardware and software to connect them to the cloud, collect data from systems. Similar to a digital business ecosystem, a digital automotive
sensors (e.g., the vehicles speed, acceleration, and steering wheel angle ecosystem contains actors that in this case are original equipment
at a certain time), and send these data to the vehicle manufacturers’ manufacturers (OEM), data intermediaries or data service providers, for
servers; this allows obtaining insights on, e.g., driving pattern analysis example. These actors have access to resources, such as data and
or estimated time of arrival in the case of fleets. Thus, an ecosystem for infrastructure, for generating, transmitting and storing data. Leveraging
such services emerges (Dhungana et al., 2016; Venkataram, 2019). these resources, the actors participate in value exchanges by providing
In general, vehicle manufacturers seek to leverage the value of the or consuming data.
data collected through their vehicles to better meet customer needs Nevertheless, research on digital automotive ecosystems is still
(Kaiser, Stocker, Festl, Lechner, & Fellmann, 2018; Stocker et al., 2017). limited. Particularly in connection with vehicle data and the process of
According to Gissler (2015), all new passenger vehicles sold in 2025 will creating data-driven services, the literature repeatedly refers to data-
be connected, forcing vehicle manufacturers to define their role and based business ecosystems (Curry, 2016; Kitsios, Papachristos, &
determine where they can best benefit from connectivity. Volkswagen, Kamariotou, 2017; Nachira, Dini, & Nicolai, 2007). For instance,
Daimler and BMW all recently announced major investments in Immonen, Palviainen, and Ovaska (2014) outline the open data
data-driven services like "Volkswagen We", "Mercedes me" or "BMW ecosystem from a business viewpoint and define ecosystem actors such

4
C. Kaiser et al. International Journal of Information Management 59 (2021) 102335

as application users, data and service providers, application developers Involved experts had on average more than 16 years professional
and infrastructure providers along with their role in the data-based work experience (cf. Table 2) and included large industries (e.g. auto­
ecosystem. Also, in many cases the authors refer more to technical motive manufacturers), small and medium enterprises (e.g. data mar­
ecosystems (e.g. Kolbe, Kubler, Robert, Le Traon, & Zaslavsky, 2017; ketplaces, suppliers, and data-driven start-ups), public authorities and
Gerloff & Cleophas, 2017; Kuschel, 2008; Martínez de Aragón, Alon­ automotive research organisations. Due to the reputation of the experts,
so-Zarate, & Laya, 2018). In these technology-oriented perspectives, an it sometimes took months before an appointment was possible. In­
analysis of the business relations enabled through the digitalisation of terviews lasted between 60− 90 min and were divided into several parts:
the vehicle and the feasibility of new data-driven services is largely
missing. • Part 1: We covered the experts’ background, professional experience,
Researchers focusing on the exploration of actors and relationships and attitude towards using data-driven applications.
between actors have often taken a different perspective, e.g. describing • Part 2: We asked them to describe vehicle data-driven services they
automotive engineering as an automotive ecosystem of interacting or­ knew and have already used to judge their experience better.
ganisations (Knauss & Damian, 2014), or presenting a strategically • Part 3: We showed experts an existing ecosystem model from the
motivated approach to discover business models in traditional industries media domain built by Gordijn, Petit, and Wieringa (2006) and asked
and apply them to the mobility sector without empirically substantiating them to attempt to sketch their view on vehicle-data driven value
their findings (Remane, Hildebrandt, Hanelt, & Kolbe, 2016). Re­ creation, which we assumed to be a cognitively challenging task. To
searchers have also used data from automotive investment and part­ guide experts, we asked them to start their personal design process
nering activities to better understand the ecosystem: Riasanow, Galic, by first naming relevant actors before designing data-sharing re­
and Böhm (2017) have used data from crunchbase.com to derive roles, lationships. Finally, we asked them to describe the changes they
design the automotive value network, and discuss the model with five expect in the digital automotive service ecosystem over the next 5
experts. Nischak and Hanelt (2019) have used data about alliances, joint years.
ventures, mergers and acquisitions along with network visualisation
techniques for a longitudinal analysis of the automotive ecosystem. We have conducted in total eleven expert meetings, four of them face
Although vehicle data paves the way to ecosystem-building activities, to face with experts who were using pen and paper to sketch their views
none of the reviewed articles contains a focus on vehicle data-based (experts 2, 3, 8, and 9). The seven remaining meetings were conducted
ecosystems. online, using a video conferencing service with screen sharing enabled.
For the virtual meetings, we prepared a special online document for
4. An expert perspective on the value of vehicle data-driven ecosystem design in which the experts had to list the relevant actors
services before linking them with data sharing relationships. In total, eight ex­
perts gave their consent to have their voices recorded during the
The literature review showed that actors and their data sharing re­ meeting, while the remaining three experts refused recording, due to
lationships were only marginally considered with regard to the devel­ strict automotive confidentiality policies.
opment of services based on the data generated by connected vehicles. In the following section, we present and discuss the various
Also, the majority of the reviewed contributions do not address the ecosystem models designed by the experts, with a detailed example from
specifics of vehicle data-driven ecosystems, which we aim to elicit by expert No. 3 and a summary of all other experts.
conducting interviews with eleven automotive domain experts with an
average professional experience of more than 16 years, all of them being
opinion leaders for the Central European market.

4.1. Data collection

Table 2
From May 2018 to December 2018, we have conducted in-depth Information on the background of the experts involved in the design process.
interviews with experts from the Central European automotive in­ Expert Organisation Expert profile Work
dustry. More specifically, we combine two instruments, capturing No. Exp.
automotive experts’ general views on value creation in vehicle data- 1 Public authority Responsible for a metadata service 23 yrs.
driven services by conducting semi-structured interviews, and then for accessing vehicle data
aiming towards gaining a deeper understanding of the data-driven value 2 Automotive research Research manager dealing with 25 yrs.
vehicle data and data-driven services
creation process through experts’ graphical models of actors and data 3 Automotive research Senior data scientist involved in 9 yrs.
sharing relationships. Two of the authors conducted the interviews and vehicle data analytics projects
the fieldwork, while the other three authors acted as critical and re­ 4 Automotive research Senior researcher involved in 5 yrs.
flexive actors (Gioia, Corley, & Hamilton, 2013) during the monthly projects with vehicle manufacturers
that deal with data-driven services
online meetings for discussing the material added to the emerging
5 Provider of data- Senior manager of a provider o f 14 yrs.
corpus of interviews, memos, and archival documents. driven service vehicle data services
According to findings of Scholte, van Teeffelen, and Verburg (2015) 6 Public authority Representative in international 26 yrs.
from ecosystem research, expert-based approaches hold the potential committees in charge of a vehicle
that experts can be asked to express their own opinions and values data provision service
7 Data marketplace Leader of a national research project 21 yrs.
starting with what they find important, while in-depth (unstructured or provider on data marketplaces
semi-structured) interviews can be used to gain a deeper understanding 8 Provider of data- Senior consultant involved in 7 yrs.
on ecosystems. Interviews have been used in the past by previous related driven services development of vehicle data-driven
research (e.g. Beverungen, Müller, Matzner, Mendling, & Vom Brocke, services
9 Automotive and Owner and managing director of an 24 yrs.
2019; Riasanow et al., 2017) to conceptualise service (eco)systems.
software automotive engineering service
However, we argue that conducting interviews alone may not be suffi­ engineering company
cient to gain a deep understanding of the complex data-sharing re­ 10 Provider of data- Senior developer of vehicle data- 5 yrs.
lationships of identified actors. Therefore, we complemented our driven services driven services
interviewing approach with simple graphical design activities to let 11 Vehicle Head of a data-driven service 19 yrs.
manufacturer department
experts visualise the value creation process from their perspective.

5
C. Kaiser et al. International Journal of Information Management 59 (2021) 102335

4.2. Case vignette [...] They would also make this data available to the decision-makers, which
would be classic loop-data." The cooperation of actors outside the closed
We present the output of one expert interview as a case vignette. This automotive ecosystems will only slowly take shape. There is still a lack
is a representative example to illustrate that all experts have a particular of cooperation at national and European level, which would benefit both
context from their field of expertise, but an excellent overview of the policy makers and traffic managers. "Decision-makers, infrastructure
automotive and mobility sector in general. providers and traffic management - which is often the same institution - have
Expert 3, Frank, doctor of technical mathematics, has more than 9 to join forces and network at least at European level in order to achieve a
years of experience as a senior data scientist for an industrial research critical mass in order to represent the interests of data intermediaries, who
company. He was jointly responsible for the development and operation currently have a very high power."
of a data-driven service based on Floating Car Data (FCD), which is used In summary, from expert 3’s view point, the main actors for the
by a traffic control center to provide information for road users, traffic design of data-driven services are data intermediaries (10 relationships),
planners, and state governments. Therefore, Frank has a particular view road users (7 relationships), traffic management (4 relationships) and
on the sharing of vehicle data, which is characterized by his own infrastructure providers (4 relationships). Among decision makers,
working context. traffic management, road operator and traffic planning, four actors have
Stakeholders relevant to Frank are decision-makers, infrastructure been identified who are related to or usually funded by national au­
providers, vehicle manufacturers, suppliers of vehicle manufacturers, thorities reflecting the research background of the expert.
road users, data intermediaries, traffic news offices and traffic man­
agement. Frank identified several data-sharing relationships between
4.3. Summary of the individual design processes of the remaining experts
these actors and presented them as connecting lines on the drawing
board (cf. Fig. 2). As a data scientist, Frank began designing the
This sub-section summarises the results of the individual design ac­
ecosystem around ’data’ that is the basic ingredient for data-driven
tivities of the remaining experts. Since their sketches are space-
services: "The problem that I have is that this data is separated. Data
consuming, we present only the sketches of the experts No. 5 and 6.
packets go from the data intermediary to the road user, and that does not
We are aware that each expert argues from his or her own perspective,
necessarily have to be the same data that the road user sends somewhere else."
also depending on the organization in which the expert is employed, so
During this design process Frank also starts to explain and interpret what
there are discrepancies in the interview statements. The aim of the
he has achieved so far in the ecosystem model, related to different actors,
empirical data collection, however, was to gain an overview as complete
their needs and relationships in the ecosystem: "Infrastructure providers
as possible of the actors and data exchange relationships.
would like to [get data from vehicle manufacturers], but they don’t get it [the
Expert 1 was responsible for a metadata service to access vehicle
data]." Therefore, infrastructure providers, as service users, appear to be
sensor data and identified the vehicle (driver) as the main actor trans­
actors that would benefit greatly from data on vehicle movements and
mitting generated vehicle data via a telecommunication provider to
would even start collecting such data by using stational roadside units to
either the OEM, a private- or a public data platform provider. A meta­
detect passing vehicles, e.g. to measure and predict traffic flows. How­
data provider, an actor in which he is personally involved, could provide
ever, their willingness to pay other actors for vehicle data is still ques­
the interface to a service provider or road operator to search and auto­
tionable: "This is still in the making, that infrastructure providers really pay
mate access to vehicle data.
data intermediaries for data", and adds, "INRIX, TomTom, or HERE – these
Expert 2 understands the ecosystem as a network of relationships
are the classic [data intermediary] players."
between actors around a data marketplace. Data collection is mentioned
Data intermediaries emerge as new players who are beginning to
several times and seems to be an unresolved problem, as expert 2 is
establish a powerful position within the ecosystem. "These are institutions
uncertain who is currently deciding on data sharing: The expert assumes
that penetrate the market from outside and deal a lot with data. They are
that data could be shared with a service provider without the knowledge
rather atypical. What is now very immanent in this system is that someone
of vehicle users. A total of 11 actors were included in the ecosystem
enters the traffic data market that actually has nothing to do with it origi­
sketch, 10 of which were connected with data sharing relationships. The
nally." Other new players are about to enter the market for data-driven
eleventh actor, the vehicle user, "does not receive any data, but actually
service generation and are seeking cooperation with existing players.
only the services". Main actors are OEM (7 links), vehicle owner (5 links),
Some actors seem to have developed their own practices to gather data for
service provider (3 links), vehicle (3 links) and data marketplace (3
decision making, e.g. traffic planners are used to collect their own data
links).
manually, instead of cooperating with other players: "… traffic planning is
Expert 4 was involved in several large-scale projects with vehicle
still a point, but now they are still outside. These would already be relevant,
manufacturers, mentions six actors and adds that there are literally data
but they now usually collect the data using standard methods." Infrastructure
sharing relationships between them all. The expert adds that a user can
providers started to make their data available to traffic planning and
generally pay for services either "with data or with money". Furthermore,
management: "Vehicle measuring stations on the motorways belong to the
the expert argues that the data marketplace will be a "closed platform [of
infrastructure provider who makes the data available to traffic management.
OEMs]", as the "access to useful vehicle data is too critical to be open",

Fig. 2. Expert 3 hand drawing and digitized vehicle data-driven service sketch.

6
C. Kaiser et al. International Journal of Information Management 59 (2021) 102335

meaning that the information could be exploited to launch a cyber- interview but did not sketch them explicitly. The expert also mentioned
attack on vehicles. the EU as an external influencing factor.
Expert 5 (cf. Fig. 3) is a senior manager of a vehicle service provider, Expert 11 is head of service development at a vehicle manufacturer
designs an in-depth model and presents the vehicle as a central actor that and outlined three different actors in two different scenarios depicting
passes on vehicle data to seven actors. The expert mentions an external the dominant role of vehicle manufacturers in the ecosystem. In the first
influence through national and European regulations to positively in­ scenario, where a customer uses a service from the OEM, the vehicle user
fluence OEMs to provide access to vehicle data for innovative service allows data access, "the consumer has the right to say no", and pays the
developments of other actors. The expert concluded by saying that "trust OEM for the service, which provides the technical infrastructure such as
is the key to the whole ecosystem". the mobile connection installed in the vehicle. The OEM in turn provides
Expert 6 (cf. Fig. 3) sketches a data value chain from the vehicle vehicle data to a contractually bound service provider and provides the
driver via a data enricher, to a service provider who provides a service to vehicle user with the developed service. In the second scenario, the
the OEM, who in turn provides a service to any service user such as vehicle user buys a service from a third-party service provider and thus
workshops, statistic services, enablers such as data marketplaces, in­ grants the service provider access to vehicle data. Due to strict European
surance companies, or public authorities. data protection legislation the vehicle user can "already decide, which
Expert 7 is involved in the development of a data marketplace and parties can be granted access to the data". The service provider, in turn,
has sketched a data flow from the OEMs to a data market provider that uses the OEM’s technical infrastructure, such as the mobile connection
makes the data available to potential buyers and service developers. The installed in the vehicle, and pays the OEM for its use, which the expert
expert describes "while the OEM servers will host the data, the data markets underlines by the statement that "vehicles are equipped with more expen­
will only do the contracting and data access and will be able to mesh data sive technology to enable data sharing".
[from different data providers]" and predicts that "data markets will succeed The statements made by the eleven experts clearly show the influ­
and take hold [in the ecosystem]. Many of them are just beginning, and some ence of their own work on the designation of key actors and data sharing
successful ones will survive". relationships. Experts working in the classic automotive industry (e.g.
Expert 8 is a consultant who sketches the model based on his own experts 9, 10 and 11) see the vehicle manufacturer in a dominant role in
experiences in developing data-driven services with SMEs and public the data-driven service ecosystem, while scientific actors and those
authorities. The main actor is "data", which can be interpreted as a data working in service development take a more differentiated view on the
platform or portal, but automotive (as data supplier) and infrastructure ecosystem.
providers (who receive data from three other actors) play an important
role, too. The actor ‘automotive’ (a synonym for car/vehicle manufac­ 5. A conceptual model for value creation in vehicle data-driven
turers) "also retrieves the data for own services, which is probably the main services
application for car manufacturers". Service providers, IT infrastructure
providers and academic research are all relevant players in service 5.1. Design process
provision, with access to the data remaining the key element.
Expert 9 is the managing director of an automotive company and did We have used two data sources, interview statements and expert
not sketch direct data-sharing relationships but mentioned eleven ac­ sketches, to derive key actors and their data sharing relationships. We
tors. He sees OEMs in a stronger position, which is suggested by the carefully examined the transcribed interviews and the individual con­
statement that "start-ups will disappear when larger players [such as OEMs] ceptual models sketched by experts and extracted terms that had been
enter the [service] market". He doubts that external players will enter the used to describe the different actors. We ended up with a list of 90 terms,
value chain between the data source and the data enricher, because some of which were mentioned more than once, and were finally able to
"data should not simply be passed on to external parties, [..], CAN data must identify 64 different actors. As experts tend to use different terms, levels
be interpreted correctly". He argues that "the balance of power between and descriptions for the same type of actors (e.g. ’OEM’, ’automotive
technology companies vs. OEMs vs. public authorities will be crucial [for the manufacturer’, and ’vehicle manufacturer’) we have renamed some
future of the data-driven service ecosystem], and a balanced situation would actors in order to create a consistent terminology for our conceptual
be best" for all stakeholders. model. Another challenge was the distinction between specific actors,
Expert 10, who is employed by a service provider, mentioned eight such as cloud service providers, and providers of data-driven services to
ecosystem actors. Vehicle data flows logically through the gateway end-users. As a result, one group of actors was referred to as ’provider of
provided by a gateway provider to a data platform provided by a hosting cloud computing services’, while another one was referred to as ’pro­
provider, to a service provider, and then to customers and fleet opera­ vider of data-driven services. We then categorised individual actors into
tors. The expert mentioned several data-sharing relationships during the groups and placed terms such as ’AI provider’, ’cloud provider’, and

Fig. 3. Digitized vehicle data-driven service sketches of experts 5 (left) and 6 (right).

7
C. Kaiser et al. International Journal of Information Management 59 (2021) 102335

’database provider’ into the actor group ’platform provider’. After the alone, but also use contextual data such as weather data or aggregated
ninth expert interview, it became apparent that no previously unknown traffic data that obviously cannot be provided by the vehicle user alone,
actors were named by the experts who could not be classified into the we added the actor role ‘contextual data provider’ in a second design
actor groups described below. Following the principle of theoretical iteration to the model to refer to actors providing other data as part of
saturation (Saunders et al., 2018), we therefore judged our sample to be the value creation process. Furthermore, we have learned from several
complete. As a result, the cleaned set included 25 actor groups. In order cases that the main beneficiary of vehicle-data driven services can be the
to make the actor groups more tangible for our model design, we vehicle user, e.g. by offering services such as intelligent parking while
identified the six highest ranked actors named by the experts and clas­ driving. However, vehicle data can also lead to services whose benefi­
sified them into groups (Table 3). ciaries go beyond the vehicle user, e.g. by providing a dynamic map of
We built our first conceptual model on the expert interview state­ traffic density to urban traffic managers. Hence, we have added the actor
ments and their model sketches. We thereby carefully examined tran­ role ‘other consumers’ in a third iteration of our model. For space rea­
scribed interviews and their sketched models and extracted terms sons, we will only show the final evaluation of our model with an
(actors, actor roles, types of data sharing relationships) to create a exemplary real-life application case in Section 6 of our paper.
consistent terminology for the conceptual model (cf. Table 3). We then
designed the first conceptual model of a unified model using only the 5.2. Conceptual model description
main actor groups from the consolidated actor group list and upon the
reviewed literature, thus establishing proof-of-concept (Nunamaker, The presented conceptual model is a result of iteratively designing a
Briggs, Derrick, & Schwabe, 2015). In a further design step, we linked conceptual model. In our design process, we performed six iterations of
actors with data supply and data consumption activities to outline the the model, to evaluate it against the views of the interviewed automotive
data transformation process. The process from data supply to data use is experts and against several real-life application cases of value creation in
often referred to in the scientific literature as the data value chain data-driven services. Existing theory on value of data-driven services (cf.
(Curry, 2016; Kaiser et al., 2019; Latif et al., 2009; Miller & Mork, 2013). Section 3.1) and data-driven value chains (e.g. Curry, 2016; Faroukhi
Our first design of a conceptual model was inspired by structuring ap­ et al., 2020; Latif et al., 2009) was used as theoretical input within this
proaches that linked actors with data transformation steps (Latif et al., design phase. Our structural design of the model was informed by Latif
2009). et al. (2009), referring to entities that can act as ecosystem roles con­
After each design, we conducted an evaluation of the model that nected by data sharing relationships, i.e. consuming or providing data.
resulted in a model revision, either in a change in model actors, a change Fig. 4 shows the metamodel of our conceptual model. It outlines that
in data-sharing relationships, or a change in both, while the general each participating entity (i.e. organisations or persons) can act in one or
structural design of the model remained unchanged. We carried out a more actor roles, thereby either providing data to the data-driven value
total of six such iterations of the model in order to evaluate it against the creation process, consuming value-added data, or doing both (if more
individual views of the engaged automotive experts and against several than one role is taken by the same ecosystem entity).
real-life use cases of value creation in data-driven services. In doing so, Participating entities can be individuals, organisations or organisa­
we follow the suggestions of design researchers such as Gregor and tional units that can take on one or more of the following actor roles:
Hevner (2013) to use case studies as a technique for conceptual model vehicle users as primary data providers, contextual data providers of­
evaluation. With regard to Sonnenberg and vom Brocke (2011) and fering additional data for service design, vehicle manufacturers that can
Venable, Pries-Heje, and Baskerville (2016), our evaluation can be seen exploit access to vehicle sensors, gateway providers collecting vehicle
as an ex-post evaluation, while we referred to the evaluation criteria data with their own equipment, data marketplace or portal providers
model completeness, fidelity with real-world phenomena, internal allowing access to data via their application programming interfaces
consistency, level of detail and robustness as published by March and (APIs), data-driven service providers, and finally vehicle users as well as
Smith (1995). In our ex-post evaluations, we also demonstrated the other consumers. We will now take a closer look into these ecosystem
usefulness of the model to describe value creation in data-driven ser­ actor roles and illustrate their data sharing relationships.
vices, establishing proof of value (Nunamaker et al., 2015). A vehicle user is a professional or private actor that decides to provide
Using our designed conceptual model to describe value creation vehicle data (i.e. data generated while vehicle operation by sensors and
referring to concrete vehicle-data driven services led to several im­ electronic control units) to be used in data-driven services in any format
provements of the model. We have already included the vehicle user as and in any level of aggregation to the related vehicle manufacturer
an essential element in the first conceptual model. However, as we found directly, or to other actors via a gateway provider indirectly. Vehicle users
that many vehicle data-driven services are not enabled by vehicle data must give their consent to the sharing of vehicle data to other ecosystem
actors.
A contextual data provider is any organisation that has additional
Table 3 contextual data that is relevant to the provision of data-driven services
Top ranked actors (left) and actor groups (right). and is willing to share this data for service development. Examples of
Top ranked actors from expert interviews (N = 64) Quantity contextual data providers are companies that can provide geodata,
OEM 7 weather data, traffic data or map data, but also governmental actors that
Service provider 6 publish open data.
Infrastructure provider 3 A vehicle manufacturer is an actor that develops, manufactures, and
Public authority 3
maintains vehicles as its main industrial product. Vehicle manufacturers
Road operator 3
Vehicle 3
have equipped vehicles with advanced sensors that collect and process a
[58 further actors] 1− 2 wealth of data to ensure the driving function, optimize the vehicle’s
internal functions and facilitate safety. Most vehicle manufacturers have
Top ranked actor groups (N = 25) Quantity
equipped their latest vehicles already with telematics software and
Vehicle manufacturer 10
connectivity to allow use of the data generated in data-driven services.
Data marketplace 9
Vehicle data service provider 9 Various types of vehicle dynamics data such as vehicle speed, acceler­
Vehicle user 7 ation, rotation, position as well as other data such as information on fuel,
Consumer 6 battery, service, and window status, wheel rotation, or steering wheel
Platform provider 5 angle can be provided at different sampling rates.
[19 further actor groups] 1− 4
A gateway provider is an actor that collects either raw or processed

8
C. Kaiser et al. International Journal of Information Management 59 (2021) 102335

Fig. 4. Metamodel of our conceptual model: Participating entities, digital ecosystem actor roles and types of data in the data-driven value creation process.

vehicle data or other contextual data such as weather data or map data measured and collected directly from vehicle sensors without any kind
for the development of data-driven services. Gateway providers may of pre-processing) or as processed vehicle/contextual data (i.e. including
collect vehicle data through deploying a data capturing device con­ some kind of data cleaning, transformation, resampling and conversion
nected to the vehicle’s on-board diagnostics interface (OBD), to the into a data format that is better suitable for service development).
controller area network bus (CAN), or through the use of a dedicated Service-specific vehicle/third party data is provided by a data market­
independent sensor and connectivity device. Besides, vehicle dynamics place, platform or portal provider that has been transformed from raw or
data can also be collected by the gateway provider using a special sensor processed vehicle data into a form that can be used by data-driven ser­
kit that is not connected to the vehicle’s bus systems or a mobile vice developers within data-driven services. Finally, consumable service
application installed on a smartphone which captures data from data is provided by data-driven service developers to vehicle users and
smartphone sensors while the vehicle is moving. third parties within provided applications (services), creating value for
A data marketplace, platform or portal provider is an actor that receives the end-users.
data from various vehicle manufacturers, contextual data providers, The conceptual model, as shown in Fig. 5, outlines individual actors
and/or gateway providers and performs data harmonisation, trans­ and their steps in vehicle data-driven value creation. The value concept
formation, and storage activities, either with the distinct purpose of we used in the model is added value for the data consumer. From the
selling service-specific vehicle data and/or service-relevant data from perspective of end-users, consumable service data is the most valuable
third parties (marketplace, platform) to enable the development of data- data. Therefore, end users may be willing to provide monetary or non-
driven services, or to provide such data for free (portal). Data market monetary consideration for this type of data.
place providers may provide data to the developers of data-driven ser­
vices who only need to integrate once via their APIs, instead of having to 6. Evaluation
enter into many different relationships with OEMs and other data sup­
pliers, while at the same time having to deal with diverse (and changing) We evaluated our conceptual model (the artifact) ex-post by
data formats. applying it to six real-life cases such as designing a data-driven service for
A provider of data-driven services is an actor that consumes service- road surface quality detection, to identify actors and data sharing re­
specific (vehicle) data from a data marketplace or data portal provider lationships as shown in Fig. 6. After each design, we conducted an ex-
and provides consumable service data to a service user, which in turn post evaluation (Sonnenberg & vom Brocke, 2011; Venable et al.,
may be either a vehicle user or another consumer, i.e. any other type of 2016) of the model that resulted in a model revision, either in a change
end-user or organisation wishing to consume a data-driven service of model actors, a change of data-sharing relationships, or a change of
enabled by vehicle data and probably enriched with other contextual both, while the general structural design of the model remained un­
relevant data. Vehicle data service providers ultimately offer data- changed. The presented case is the sixth and final evaluation of the
driven services, such as road surface quality detection, harsh driving model and based on real experiences of two authors working on the
detection, or predictive maintenance. project mentioned in the introduction. After this last evaluation, the
Finally, both vehicle users and other consumers may be consumers of model remained stable.
data-driven services offered by data-driven service providers. As the A data-driven service for road surface quality detection can be envisaged
final actor in the data-driven value chain, these professional or private by the municipality of a city, responsible for a road network (e.g. the
actors are end users and main beneficiaries of the entire data trans­ City of Vienna with a road network of 3.000 km). The municipality
formation process. Examples of data-driven services are road surface operates a vehicle fleet and has an infrastructure management depart­
quality detection (consumed by municipalities or a road infrastructure ment which orchestrates road maintenance work. Thus the municipality
managers) or harsh driving detection (consumed drivers to improve acts as vehicle user (collecting fleet operation data) and other consumer
their driving style or by insurance companies to provide a ’pay as you (consuming road surface quality data) in this case (orange background
drive’ insurance that calculates the insurance premium based on the colour in Fig. 6).
driving style). Our model indicates drivers from the municipality fleet as data cre­
Actors provide and consume different types of data within the data ators who can opt in to deliver service-relevant data such as vehicle
value creation process. First, in order to comply with data protection speed, acceleration, rotation and position. Furthermore, map data from
regulations such as the General Data Protection Regulation (GDPR) in a contextual data provider (Map provider, orange font colour) must be
Europe, vehicle users should grant access to the data their vehicle gen­ used for georeferencing detected road surface quality from recorded
erates before vehicle data may be used in services. Contextual data rele­ vehicle data. Vehicle data can be captured directly by vehicle manu­
vant for the development of a particular data-driven service, such as facturers in case they already operate vehicles in this city that can
weather data, traffic data or data on accident hotspots, are provided by capture and transmit those data, which requires special contracts with
providers of contextual data for the data value creation process. This selected manufacturers. Vehicle data can also be collected via gateway
data can be provided as raw vehicle/contextual data (e.g. as data that is providers that provide devices for installation in vehicles equipped with

9
C. Kaiser et al. International Journal of Information Management 59 (2021) 102335

Fig. 5. A conceptual model for value creation in vehicle data-driven services.

Fig. 6. Actors and data sharing relationships in the design of a data-driven service for road surface quality detection.

10
C. Kaiser et al. International Journal of Information Management 59 (2021) 102335

the necessary sensors, which presupposes that a sufficient number of significant contribution. While previous research in (big) data has
gateways are installed in vehicles moving in the area where road surface shown a clear focus on data users (Wiener et al., 2020), we also
quality is to be measured. A municipality may already operate own fleets emphasize the importance of data providers and intermediaries and
with different vehicle brands, consisting of all employees’ business ve­ their interactions in a multi-actor model, thus extending the perspective
hicles. In this case, the municipality may form a business relationship to the ecosystem where the value creation is enacted. Consequently, we
with a gateway provider (OBD data provider, orange font colour) to emphasize that data-driven value creation in the automotive ecosystem
support the data collection by equipping all vehicles in the fleet inde­ must be achieved through collaboration among various stakeholders,
pendent of their brand with gateways. Vehicle operation data being thus contributing to the debate on realizing value from (big) data (cf.
collected could be prepared for further processing and then either be Günther et al., 2017) by stressing a multi-actor perspective.
made available to the provider of the data-driven service (a company in Several researchers in the field of information systems have also been
charge of developing the road surface quality detection service) by a engaged in the study of value creation from big data because big data is a
platform/portal provider or a data marketplace ideally in combination comparatively new phenomenon and the organizational implications of
with corresponding map data from a map provider. As the provider of big data are of great interest to them (e.g., Akter et al., 2016; Dremel
the data-driven service already cooperates with the cloud provider AWS et al., 2017, 2020; Grover et al., 2018; Lehrer et al., 2018; Mikalef et al.,
(orange font colour), AWS is also chosen as platform provider. Finally, a 2017; Svahn et al., 2017; Wamba et al., 2015, 2017). However, they
contracted software development company (orange font colour) is focus on the impact of big data analytics on the level of an individual
responsible for developing the road surface quality detection service and organization (e.g., on organizational performance, strategic business
takes vehicle data and map data required for service provision, applies a value, strategic use, organizational change, or required organizational
data processing approach, extracts events indicating a particular road capabilities) and exclude the network and ecosystem perspectives for
surface quality such as potholes from identified deviations of the pro­ creating data-driven services. They focus on an intra-organizational (i.
cessed vertical acceleration and pitch together with their positions e., micro) perspective, whereas we want to look at value creation in a
(vehicle data), and visualises the geographic position of identified pot­ multi-actor ecosystem (i.e., macro) perspective. While their research
holes in a web dashboard (using map data) to create value for the service specifically targets the big data phenomenon, we want to emphasize that
consumer, the infrastructure management department of the munici­ value can also be created from services enabled by the exchange of small
pality (orange background colour). In addition, the software develop­ data between actors.
ment company prepares a table with prioritised repair lists and Several studies investigate how vehicle usage data can lead to novel
interaction possibilities to investigate the worsening or improvement of services, such as location-based services for carsharing vehicles (Wag­
road surface quality for the infrastructure management department. ner, Willing, Brandt, & Neumann, 2015), predictive maintenance of
This real-life application case illustrates the complexity of devel­ connected vehicles (Gerloff & Cleophas, 2017), or eco-feedback on
oping a road surface quality detection service and shows the usefulness driving behaviour (Bätz, Gimpel, Heger, & Wöhl, 2020). Yet, these
of our model for better understanding the roles of concrete ecosystem studies focus rather on data analytics approaches to exploit vehicle data
actors and their data sharing relationships in the development of a than on the data ecosystem perspective. Also considering the state of the
vehicle data-driven service. art gaps discussed in previous sections, we argue that our proposed
conceptual model would allow relevant actors to be identified and
7. Discussion mapped in order to eventually achieve periods of stability and change
(Nischak et al., 2017, p. 17) and the interactions that ultimately lead to
Vehicles are increasingly equipped with advanced sensors to ensure the envelopment (Eisenmann, Parker, & Van Alstyne, 2011) of other
driving functionality, optimise the vehicle’s functions, and facilitate emerging digital business ecosystems. Furthermore, our model indicates
safety and comfort through increased automation such as providing choices for how the value chain can evolve and, above all, which other
adaptive driving assistance systems (Stocker et al., 2017). Moreover, actors are needed, because the development of a data-driven service and
most vehicle manufacturers have additionally equipped their latest ve­ the selection of suitable actors is a decision-making task.
hicles with advanced software and connectivity to make use of the data Our model shows that actors are involved in a multi-party data value
generated and to provide additional services to drivers. The data creation process to ultimately provide sustainable data-driven services
generated during vehicle use can enable new types of data-driven ser­ to service customers such as vehicle drivers and therefore contributes to
vices addressing many interesting use cases (if drivers opt in to vehicle a better understanding of vehicle data-driven value creation in general.
data sharing) that go far beyond supporting the operation of vehicles, Based on our interviews with experts, all of whom have a connection to
especially through intelligent linking of vehicle sensor data with other vehicle data-driven value creation and some of whom are developing
contextual data such as weather data or data on the traffic situation. This these vehicle data-driven services themselves, we have learned that the
raises the important question of which ecosystem actors can and will successful development and provision of data-driven services in the
contribute to use cases that can only be implemented if data is shared automotive domain and thus the successful monetisation of vehicle
between multiple actors. operation data will require new partnerships between individual
ecosystem actors, as no actor will bear the service development risk
7.1. Implications for theory alone. We argue that our conceptual model provides a solid under­
standing of the ecosystem actors and their role in data sharing and in the
The theoretical gap addressed in our paper is the lack of conceptual creation of data-driven services, thus supporting strategic decisions, e.g.,
models that can unravel the underlying value chain (actors and data in terms of partnerships and sourcing. In doing so, we are contributing to
sharing relationships) when establishing vehicle data-driven services. In research on data monetization, responding to the call by Parvinen et al.
this paper, we have therefore presented a novel conceptual model that (2020) for a better understanding of the role of data aggregators and
includes multiple actors and their data sharing relationships (i.e. in refiners in data monetization, how they create value and how different
terms of a data value chain) that are relevant for vehicle data-driven parties can capture it.
value creation. As such, our multi-actor model shows data and infor­ We have developed our model empirically, drawing on the knowl­
mation flows as a series of data sharing and data transformation steps edge of automotive domain experts who have an average of more than
that are needed to finally generate value and useful insights to service 16 years of professional experience in the mobility industry. Laying
consumers, establishing proof-of-concept (Nunamaker et al., 2015). emphasis on actors that have a stake in data generation and sharing, we
Following Baskerville, Baiyere, Gregor, Hevner, and Rossi (2018) we differ methodically from the approaches of other researchers who study
present a novel and useful conceptual model and thus generate a ecosystems in the mobility domain, including Riasanow et al. (2017)

11
C. Kaiser et al. International Journal of Information Management 59 (2021) 102335

using crunchbase.com data to visualise the current automotive others via its own datacentre.
ecosystem in a generic value network, Remane et al. (2016) focusing on Start-ups interested in producing data-driven services may realise
the identification of business model types of start-ups, or Kolbe et al. that they can also turn to data marketplaces that have already signed
(2017) creating an IoT framework and focusing on semantic contracts with vehicle manufacturers and do not need to negotiate
interoperability. individually with each manufacturer to access the necessary data. The
Our background is in the field of data-driven service development in provision of vehicle data to data marketplaces can also lead to new ways
the automotive domain, and we stress that our conceptual model is for vehicle manufacturers to monetise vehicle data, namely when others
inspired by research on data-driven value creation published by Curry use it to develop services that generate value independent of their core
(2016), Miller and Mork (2013), or Latif et al. (2009). Our concept of product, the vehicle. Those who wish to design data-driven services can
connecting automotive ecosystem actors with data sharing and enrich­ better identify the key players in the ecosystem they need to deal with,
ment processes is new. We understand our model as a descriptive tool and those who want to be part of the service delivery process can better
that shows the process towards providing a data-driven service from understand who they need to work with. Since one of the first decisions
both an actor and a data perspective. Furthermore, we believe that our for organisations seeking to monetize vehicle data is to figure out, where
presented research is also helpful in better describing and classifying to play in the value chain (Hood, Hoda, & Robinson, 2019), we consider
existing data-driven services. Our model can support ecosystem actors to the knowledge contained in our model to be a significant contribution.
better recognise and understand their interdependencies with other
actors or even to understand what interdependencies exist at all. 7.3. Limitations
It is worth mentioning that actors within the ecosystem for vehicle
data-driven value creation are different from the classical actors within In our concept phase, we tried to generalise the expert’s individual
the vehicle supply chain. For instance, although vehicle manufacturers mental models on data-driven value creation in order to eliminate in­
(OEMs) are heavily dependent on original equipment suppliers in the dividual perspectives as much as possible. Furthermore, we have
supply chain, these Tier-1 (module or system suppliers) and Tier-2 involved eleven experts from Central Europe in the data collection, who
(component suppliers) are not specifically addressed in our model. also work together with specific players in the automotive ecosystem
However, they have an indirect relevance within the creation of vehicle- and thus contribute their own views. All interviewed experts are opinion
data driven services: First, they can supply the vehicle telematics device leaders for the Central European market (the location of some of the
to the vehicle manufacturer, which enables data acquisition and data largest vehicle manufacturers in the world), and therefore we believe
transfer to the manufacturer’s backend servers. However, suppliers do that the interviewed experts represent an impressive amount of knowl­
not have a direct role within the process "from data to service", as they edge. The interviews and individual sketching activities of the experts
do not have direct access to the vehicle data transmitted by their sup­ showed that there was a consensus on many important patterns (i.e., on
plied telematics units to the vehicle manufacturer. Second, suppliers the roles of the actors and their data sharing relationships). This seems to
may act as service developers providing not only hardware but also data- show that our sample is appropriate for our research purpose. It is also
driven services to vehicle manufacturers. If suppliers choose to do so, worth noting that two of the authors have been working in the auto­
they are included in the model in the actor role "Provider of data-driven motive sector for eight years each. Their contextual bias is mitigated by
services". We have deliberately avoided an actor role "supplier" in the closely involving the other three authors in the research process in order
model. For example, Tier-1 Robert Bosch GmbH not only designs vehicle to adopt an external and critical perspective, and by reflecting the results
telematics devices but also offers data-driven services, such as road of the design process with them, so that "the higher-level perspective
condition-based services (Bosch, 2020a) or Connected Horizon (Bosch, necessary for informed theorising" is maintained (Gioia et al., 2013, p.
2020b). The development and provision of both services can be well 5). Finally, we have evaluated the model in total six times ex-post by
described by the use of our model, and both cases served within the applying it to real-life cases, establishing proof-of-value (Nunamaker
conceptual model evaluation process. Third, suppliers can act as users of et al., 2015). Furthermore, we established proof-of-use by successfully
a data-driven service, and in this case, are included in the model as applying the model in a research proposal that was granted with
"other consumers". A prominent example case is the provision of a funding.
data-driven service for ECU health, that is made available to suppliers.
This service can help suppliers to monitor the functionality of ECUs they 8. Conclusion
have designed and delivered to vehicle manufacturers and that are
installed in the vehicle by the OEM. Suppliers can also take advantage of In this article we adopt an ecosystem (i.e., macro) perspective and
driving style recognition or environmental condition monitoring ser­ propose a novel conceptual, multi-actor model for value creation in
vices that will both help them to improve their ECU designs as well. vehicle data-driven services consisting of ecosystem actors and their
data sharing relationships, establishing proof-of-concept. We thereby
7.2. Implications for practice illustrate how different key actors such as vehicle users, manufacturers,
data marketplaces, and service providers have to engage in data sharing
In addition, we see several implications of our work for business relationships to create value from vehicle data (i.e., data that is collected
practice. Based on a specific role of an ecosystem actor, we have shown by the vehicle’s sensors) and other relevant contextual data. We evalu­
in the evaluation that our conceptual model is useful to practitioners to ated our model ex-post by applying it to six real-life application cases,
better understand their own position in the ecosystem. such as the development of a vehicle data-driven service for road surface
For example, a manager responsible for digitalisation can identify quality detection, which we also present in our paper, establishing
which actors are relevant to provide data-driven services. In addition, proof-of-value.
service developers may recognise the special role of a vehicle user, The theoretical gap addressed in our paper is the lack of conceptual
without whose consent to the provision of collected data the develop­ multi-actor models that can unravel the underlying value chain (actors
ment of a data-driven service will not be possible. Vehicle manufacturers and data sharing relationships) when establishing (vehicle) data-driven
may be able to better communicate their own position in the value chain services and consider an ecosystem perspective. Many of the researchers
as the one who can technically store, interpret, and forward generated cited have focused on the perspective of a single organization, with an
vehicle data. The manufacturer may recognise that a scaling provision of emphasis on deciphering the phenomenon of big data analytics and its
certain data-based services will only be possible if other actors are implications at the intra-organizational (i.e., micro) level. As our eval­
granted access to the vehicle’s bus information systems or if the uation has shown the conceptual model contributes to a better under­
manufacturer stores, transmits and makes vehicle data available to standing of the (data-driven) value creation logic and reveals critical

12
C. Kaiser et al. International Journal of Information Management 59 (2021) 102335

actors and their data sharing activities that ultimately lead to created Attard, J., & Brennan, R. (2018). Challenges in value-driven data governance. On the
move to meaningful internet systems (pp. 546–554). Cham: Springer.
value.
Azkan, C., Iggena, L., Gür, I., Möller, F., & Otto, B. (2020). A taxonomy for data-driven
While we designed our conceptual model as a high-level model to services in manufacturing industries. PACIS, 184.
reduce the complexity of the whole automotive ecosystem and focus on Baskerville, R., Baiyere, A., Gregor, S., Hevner, A., & Rossi, M. (2018). Design science
vehicle data provision and use, we are aware that our model cannot research contributions: Finding a balance between artifact and theory. Journal of the
Association for Information Systems, 19(5), 358–376.
represent and explain all relevant aspects, value flows and power re­ Batini, C., Castelli, M., Viscusi, G., Cappiello, C., & Francalanci, C. (2018). Digital
lations. We have focused on the data value chain and have therefore only information asset evaluation: A case study in manufacturing. SIGMIS Database, 49
included the most important actors in terms of data sharing. Neverthe­ (3), 19–33. https://doi.org/10.1145/3242734.3242737.
Bätz, A., Gimpel, H., Heger, S., & Wöhl, M. (2020). Driving Sustainably–The influence of
less, we see numerous practical implications as our model could be used IoT-based eco-feedback on driving behavior. Proceedings of the 53rd Hawaii
as a governance and/or creativity tool to influence data sharing regu­ International Conference on System Sciences.
lation (e.g., to better understand the dominant role of the OEM in Bera, P., Soffer, P., & Parsons, J. (2019). Using eye tracking to expose cognitive processes
in understanding conceptual models. MIS Quarterly, 43(4), 1105–1126.
enabling vehicle data-driven services) or even for the design of data- Berkley, B. J., & Gupta, A. (1994). Improving service quality with information
driven services outside the automotive domain. In addition, at an aca­ technology. International Journal of Information Management, 14(2), 109–121.
demic level, we see our research as a first contribution to the systematic https://doi.org/10.1016/0268-4012(94)90030-2.
Bertoncello, M., Camplone, G., Gao, P., Kaas, H. W., Mohr, D., Möller, T., et al. (2016).
design of a multi-actor model for vehicle data-driven value creation in Monetizing car data—New service business opportunities to create new customer benefits.
the automotive sector that can help to guide next research endeavours in McKinsey & Company.
data-driven service development. Beverungen, D., Lüttenberg, H., & Wolf, V. (2018). Recombinant service systems
engineering. Business & Information Systems Engineering, 60(5), 377–391.
Finally, we expect our paper to have further implications on research
Beverungen, D., Müller, O., Matzner, M., Mendling, J., & Vom Brocke, J. (2019).
such as becoming a structuring tool to design, compare and/or analyse Conceptualizing smart service systems. Electronic Markets, 29(1), 7–18.
cases of data-driven service development, or simply help future re­ Bharadwaj, A., El Sawy, O., Pavlou, P., & Venkatraman, N. (2013). Digital business
searchers to better understand potentials and pitfalls in the development strategy: Toward a next generation of insights. MIS Quarterly, 37(2), 471–482.
BMW. (2020). BMW and MINI CarData. https://www.bmwgroup.com/en/innovation/
of data-driven services. We even believe that the presented model is technologies-and-mobility/cardata.html.
transferable to other domains where non-digital artefacts are the core Bosch. (2020a). Predictive road condition services. https://www.bosch-mobility-solutions.
product that generate data during use (such as the aircraft industry), com/en/products-and-services/passenger-cars-and-light-commercial-vehicles
/automated-driving/predictive-road-condition-services/.
although proving this claim would go beyond the scope of this paper. Bosch. (2020b). Connected horizon. https://www.bosch-mobility-solutions.com/en/prod
Going after this claim, however, may spur future research endeavours. ucts-and-services/passenger-cars-and-light-commercial-vehicles/connectivity-sol
utions/connected-horizon/.
Brennan, R., Attard, J., Petkov, P., Nagle, T., & Helfert, M. (2019). Exploring data value
CRediT authorship contribution statement assessment: A survey method and investigation of the perceived relative importance
of data value dimensions. Proceedings of the 21st International Conference on Enterprise
Christian Kaiser: Conceptualization, Methodology, Formal analysis, Information Systems - Volume 1: ICEIS, 200–207. https://doi.org/10.5220/
0007723402000207.
Investigation, Data curation, Writing - original draft, Writing - review & Bridgelall, R., Lu, P., Tolliver, D. D., & Xu, T. (2018). Mining connected vehicle data for
editing, Visualization, Project administration. Alexander Stocker: beneficial patterns in Dubai taxi operations. Journal of Advanced Transportation,
Conceptualization, Methodology, Formal analysis, Investigation, Data 2018.
Briscoe, G., & De Wilde, P. (2006). Digital ecosystems: Evolving service-oriented
curation, Writing - original draft, Writing - review & editing, Visuali­
architectures. Proceedings of Conference on Bio Inspired Models of Network, Information
zation, Supervision, Funding acquisition. Gianluigi Viscusi: Writing - and Computing Systems.
original draft, Writing - review & editing. Michael Fellmann: Meth­ Chesbrough, H., & Spohrer, J. (2006). A research manifesto for services science.
Communications of the ACM, 49(7), 35–40.
odology, Writing - original draft, Writing - review & editing, Visuali­
Conboy, K., Dennehy, D., & O’Connor, M. (2020). ‘Big time’: An examination of temporal
zation, Supervision. Alexander Richter: Conceptualization, complexity and business value in analytics. Information & Management, 57(1),
Methodology, Formal analysis, Writing - review & editing, Supervision. 103077. https://doi.org/10.1016/j.im.2018.05.010.
Curry, E. (2016). The big data value chain: Definitions, concepts, and theoretical
approaches. New horizons for a data-driven economy (pp. 29–37). Cham: Springer.
Declaration of Competing Interest Daimler. (2020a). Digitalisation. https://www.daimler.com/innovation/digitalisation/.
Daimler. (2020b). So far – and yet so near. https://www.daimler.com/innovation/case/
connectivity/connected-with-the-vehicle.html.
The authors report no declarations of interest.
Davenport, J., Pacheco, P., & Priestley, A. (2020). Market trends: Monetizing connected and
autonomous vehicle data. Gartner Information Technology Research.
Acknowledgements De Winter, J. C. F., Dodou, D., Happee, R., & Eisma, Y. B. (2019). Will vehicle data be
shared to address the how, where, and who of traffic accidents? European Journal of
Futures Research, 7(1), 1–9.
The EVOLVE project (www.evolve-h2020.eu) has received funding Dedrick, J. (2010). Green IS: Concepts and issues for information systems research.
from the European Union’s Horizon 2020 research and innovation Communications of the Association for Information Systems, 27, 173–184.
program under Grant Agreement No 825061. The document reflects Dhungana, D., Engelbrecht, G., Parreira, J. X., Schuster, A., Tobler, R., & Valerio, D.
(2016). Data-driven ecosystems in smart cities: A living example from Seestadt
only the author’s views and the Commission is not responsible for any Aspern. In 2016 IEEE 3rd World Forum on Internet of Things (WF-IoT) (pp. 82–87).
use that may be made of information contained therein. Dremel, C., Herterich, M., Wulf, J., Waizmann, J.-C., & Brenner, W. (2017). How AUDI
This work has been partially supported by the project SCALINGS AG established big data analytics in its digital transformation. MIS Quarterly
Executive, 16(2), 81–100.
(Scaling Up Co-creation: Avenues and Limits for Integrating Society in Dremel, C., Herterich, M. M., Wulf, J., & Vom Brocke, J. (2020). Actualizing big data
Science and Innovation), which has received funding from the European analytics affordances: A revelatory case study. Information & Management, 57(1),
Union’s Horizon 2020 research and innovation programme under Grant Article 103121.
Eisenmann, T., Parker, G., & Van Alstyne, M. (2011). Platform envelopment. Strategic
Agreement No 788359. Management Journal, 32(12), 1270–1285.
Falagas, M. E., Pitsouni, E. I., Malietzis, G. A., & Pappas, G. (2008). Comparison of
References PubMed, Scopus, web of science, and Google scholar: Strengths and weaknesses. The
FASEB journal, 22(2), 338–342. https://doi.org/10.1096/fj.07-9492LSF.
Faroukhi, A. Z., El Alaoui, I., Gahi, Y., & Amine, A. (2020). Big data monetization
Adner, R. (2016). Ecosystem as structure: An actionable construct for strategy. Journal of
throughout Big Data Value Chain: A comprehensive review. Journal of Big Data, 7(1),
Management, 43(1), 39–58.
1–22.
Akter, S., Wamba, S. F., Gunasekaran, A., Dubey, R., & Childe, S. J. (2016). How to
Gerloff, C., & Cleophas, C. (2017). Excavating the treasure of IoT data: An architecture to
improve firm performance using big data analytics capability and business strategy
empower rapid data analytics for predictive maintenance of connected vehicles.
alignment? International Journal of Production Economics, 182, 113–131.
Proceedings of the Thirty-Eight International Conference on Information Systems (ICIS
Athanasopoulou, A., Bouwman, H., Nikayin, F., & de Reuver, M. (2016). The disruptive
2017).
impact of digitalisation on the automotive ecosystem: A research agenda on business
models, platforms and consumer issues. Proceedings of 29th Bled eConference,
597–604.

13
C. Kaiser et al. International Journal of Information Management 59 (2021) 102335

Gioia, D. A., Corley, K. G., & Hamilton, A. L. (2013). Seeking qualitative rigor in Malhotra, A., Melville, N. P., & Watson, R. T. (2013). Spurring impactful research on
inductive research: Notes on the Gioia methodology. Organisational Research information systems for environmental sustainability. MIS Quarterly, 37(4),
Methods, 16(1), 15–31. 1265–1274.
Gissler, A. (2015). Connected vehicle: Succeeding with a disruptive technology. Accenture. March, S. T., & Smith, G. F. (1995). Design and natural science research on information
http://www.automotivebusiness.com.br/abinteligencia/pdf/Accenture-digital- technology. Decision Support Systems, 15(4), 251–266, 1995 Dec 1.
Connected-Vehicle.pdf. Martínez de Aragón, B., Alonso-Zarate, J., & Laya, A. (2018). How connectivity is
Gordijn, J., Petit, M., & Wieringa, R. (2006). Understanding business strategies of transforming the automotive ecosystem. Internet Technology Letters, (1). e14.
networked value constellations using goal- and value modeling. Proceedings of 14th McKinsey. (2016). Automotive revolution - perspective towards 2030. https://www.
IEEE International Requirements Engineering Conference, 129–138. mckinsey.com/~/media/mckinsey/industries/high tech/our insights/disruptive
Gregor, S., & Hevner, A. R. (2013). Positioning and presenting design research for trends that will transform the auto industry/auto 2030 report jan 2016.ashx
maximum impact. MIS Quarterly, 37(2), 337–355. June 2013. Accessed: May 16, 2020.
Grover, V., Chiang, R. H. L., Liang, T.-P., & Zhang, D. (2018). Creating strategic business Mikalef, P., Pappas, I. O., Krogstie, J., & Pavlou, P. A. (2020). Big data and business
value from big data analytics: A research framework. Journal of Management analytics: A research agenda for realizing business value. Information & Management,
Information Systems, 35(2), 388–423. https://doi.org/10.1080/ 57(1), 103237. https://doi.org/10.1016/j.im.2019.103237.
07421222.2018.1451951. Mikalef, P., Pappas, I. O., Krogstie, J., & Giannakos, M. (2017). Big data analytics
Günther, W. A., Rezazade Mehrizi, M. H., Huysman, M., & Feldberg, F. (2017). Debating capabilities: A systematic literature review and research agenda. Information Systems
big data: A literature review on realizing value from big data. The Journal of Strategic and E-Business Management, 1–32. https://doi.org/10.1007/s10257-017-0362-y).
Information Systems, 26(3), 191–209. Miller, H. G., & Mork, P. (2013). From data to decisions: A value chain for big data. IT
Gusenbauer, M., & Haddaway, N. R. (2020). Which academic search systems are suitable Professional, (1), 57–59.
for systematic reviews or meta-analyses? Evaluating retrieval qualities of Google Mocker, M., & Fonstad, N. O. (2017). How AUDI AG is driving toward the sharing
Scholar, PubMed, and 26 other resources. Research Synthesis Methods, 11(2), economy. MIS Quarterly Executive, 16(4), 279–293.
181–217. Nachira, F., Dini, P., & Nicolai, A. (2007). A network of digital business ecosystems for
Hanelt, A., Piccinini, E., Gregory, R. W., Hildebrandt, B., & Kolbe, L. M. (2015). Digital Europe: Roots, processes and perspectives. Bruxelles: European Commission.
transformation of primarily physical industries – Exploring the impact of digital Introductory Paper, 106.
trends on business models of automobile manufacturers. Proceedings of Nischak, F., & Hanelt, A. (2019). Ecosystem change in the era of digital innovation–A
Wirtschaftsinformatik, 1313–1327, 2015. longitudinal analysis and visualization of the automotive ecosystem. Proceedings of
Hartmann, M. P., Mohamed, Z., Niels, F., & Andy, N. (2016). Capturing value from big International Conference on Information Systems (ICIS 2019).
data – A taxonomy of data-driven business models used by start-up firms. Nischak, F., Hanelt, A., & Kolbe, L. M. (2017). Unraveling the interaction of information
International Journal of Operations & Production Management, 36(10), 1382–1406. systems and ecosystems – A comprehensive classification of literature. Proceedings of
https://doi.org/10.1108/IJOPM-02-2014-0098. the Thirty Eight International Conference on Information Systems (ICIS 2017).
Henfridsson, O., Mathiassen, L., & Svahn, F. (2009). Reconfiguring modularity: Closing Nunamaker, J. F., Briggs, R. O., Derrick, D. C., & Schwabe, G. (2015). The last research
capability gaps in digital innovation. All Sprouts Content. 277. mile: Achieving both rigor and relevance in information systems research. Journal of
Hevner, A. R. (2007). A three cycle view of design science research. Scandinavian Journal Management Information Systems, 32(3), 10–47.
of Information Systems, 19(2), 4, 2007 Jan 1. Papatheocharous, E., Frecon, E., Kaiser, C., Festl, A., & Stocker, A. (2018). Towards a
Hevner, A. R., March, S. T., Park, J., & Ram, S. (2004). Design science in information generic IoT platform for data-driven vehicle services. In 2018 IEEE International
systems research. MIS Quarterly, (2004), 75–105. Conference on Vehicular Electronics and Safety (ICVES) (pp. 1–6).
Hood, J., Hoda, O., & Robinson, R. (2019). Monetizing data in the age of connected vehicles. Parvinen, P., Pöyry, E., Gustafsson, R., Laitila, M., & Rossi, M. (2020). Advancing data
Deloitte Insights. https://www2.deloitte.com/us/en/insights/industry/automotive/ monetization and the creation of data-based business models. Communications of the
monetizing-data-connected-vehicles.html. association for information systems (p. 47). https://doi.org/10.17705/1CAIS.04702.
Immonen, A., Ovaska, E., & Kalaoja, J. (2015). A service requirements engineering Piccinini, E., Hanelt, A., Gregory, R., & Kolbe, L. M. (2015). Transforming industrial
method for a digital service ecosystem. Service Oriented Computing and Applications, business: The impact of digital transformation on automotive organisations.
10(2), 151–172. Proceedings of Thirty Sixth International Conference on Information Systems (ICIS 2015).
Immonen, A., Palviainen, M., & Ovaska, E. (2014). Requirements of an open data based Pillmann, J., Wietfeld, C., Zarcula, A., Raugust, T., & Alonso, D. C. (2017). Novel
business ecosystem. IEEE Access, 2, 88–103. common vehicle information model (cvim) for future automotive vehicle big data
Jacobides, M. G., Cennamo, C., & Gawer, A. (2018). Towards a theory of ecosystems. marketplaces. In 2017 IEEE Intelligent Vehicles Symposium (IV) (pp. 1910–1915).
Strategic Management Journal, 39(8), 2255–2276. Pütz, F., Murphy, F., Mullins, M., & O’Malley, L. (2019). Connected automated vehicles
Kaiser, C., Festl, A., Pucher, G., Fellmann, M., & Stocker, A. (2019). The vehicle data and insurance: Analysing future market-structure from a business ecosystem
value chain as a lightweight model to describe digital vehicle services. 15th perspective. Technology in Society, 59, Article 101182.
International Conference on Web Information Systems and Technologies. Rai, A., & Sambamurthy, V. (2006). Editorial notes—The growth of interest in services
Kaiser, C., Stocker, A., Festl, A., Lechner, G., & Fellmann, M. (2018). A research agenda management: Opportunities for information systems scholars. Information Systems
for vehicle information systems. Proceedings of European Conference on Information Research, 17(4), 327–331.
Systems (ECIS 2018). Remane, G., Hildebrandt, B., Hanelt, A., & Kolbe, L. M. (2016). Discovering new digital
Kitsios, F., Papachristos, N., & Kamariotou, M. (2017). Business models for open data business model types-a study of technology startups from the mobility sector.
ecosystem: Challenges and motivations for entrepreneurship and innovation. In 2017 Proceedings of Pacific Asia Conference on Information Systems (PACIS 2016), 289.
IEEE 19th Conference on Business Informatics (CBI) (Vol. 1, pp. 398–407). Riasanow, T., Galic, G., & Böhm, M. (2017). Digital transformation in the automotive
Knauss, E., & Damian, D. E. (2014). Towards enabling Cross-organisational modeling in industry: Towards a generic value network. Proceedings of European Conference on
automotive ecosystems. Proceedings of MD2P2@ MoDELS, 38–47. Information Systems (ECIS 2017).
Kolbe, N., Kubler, S., Robert, J., Le Traon, Y., & Zaslavsky, A. (2017). Towards semantic Saunders, B., Sim, J., Kingstone, T., Baker, S., Waterfield, J., Bartlam, B., et al. (2018).
interoperability in an open IoT ecosystem for connected vehicle services. Proceedings Saturation in qualitative research: Exploring its conceptualization and
of 2017 Global Internet of Things Summit (GIoTS), 1–5. operationalization. Quality & Quantity, 52(4), 1893–1907, 2018 Jul 1.
Kuhnert, F., Stürmer, C., & Koste, A. (2018). Five trends transforming the automotive Scholte, S. S., van Teeffelen, A. J., & Verburg, P. H. (2015). Integrating socio-cultural
industry. Frankfurt, Germany: PricewaterhouseCoopers. perspectives into ecosystem service valuation: A review of concepts and methods.
Kumar, V., Chattaraman, V., Neghina, C., Skiera, B., Aksoy, L., Buoye, A., et al. (2013). Ecological Economics, 114, 67–78.
Data-driven services marketing in a connected world. Journal of Service Management, Schulze, E. D., Beck, E., & Müller-Hohenstein, K. (2005). Plant ecology. Berlin/
24(3), 330–352. https://doi.org/10.1108/09564231311327021. Heidelberg: Springer.
Kuschel, J. (2008). The vehicle ecosystem. In IFIP Working Conference on Open IT-Based Schüritz, R., & Satzger, G. (2016). Patterns of data-infused business model innovation. In
Innovation: Moving Towards Cooperative IT Transfer and Knowledge Diffusion (pp. Business Informatics (CBI), 2016 IEEE 18th Conference on (Vol. 1, pp. 133–142).
309–322). Schüritz, R., Farrell, K., Wixom, B. W., & Satzger, G. (2019). Value Co-creation in data-
Latif, A., Saeed, A. U., Hoefler, P., Stocker, A., & Wagner, C. (2009). The linked data driven services: Towards a deeper understanding of the joint sphere. Proceedings of
value chain: A lightweight model for business engineers. Proceedings of I- International Conference on Information Systems (ICIS 2019).
SEMANTICS, 568–575. Schüritz, R., Seebacher, S., & Dorner, R. (2017). Capturing value from data: Revenue
Lehrer, C., Wieneke, A., vom Brocke, J., Jung, R., & Seidel, S. (2018). How big data models for data-driven services. Proceedings of the 50th Hawaii International
analytics enables service innovation: Materiality, affordance, and the Conference on System Sciences.
individualization of service. Journal of Management Information Systems, 35(2), Seiberth, G., & Gründinger, W. (2018). Data-driven business models in connected cars,
424–460. mobility services & beyond. BVDW Research, 1, 18.
Lim, C., Kim, K. H., Kim, M. J., Heo, J. Y., Kim, K. J., & Maglio, P. P. (2018). From data to Sonnenberg, C., & vom Brocke, J. (2011). Evaluation patterns for design science research
value: A nine-factor framework for data-based value creation in information- artefacts, in practical aspects of design science (pp. 71–83). Springer Berlin Heidelberg.
intensive services. International Journal of Information Management, 39, 121–135. https://doi.org/10.1007/978-3-642-33681-2_7).
Lusch, R. F., & Nambisan, S. (2015). Service innovation: A service-dominant logic Sorescu, A. (2017). Data-driven business model innovation. Journal of Product Innovation
perspective. MIS Quarterly, 39(1), 155–175. Management, 34(5), 691–696. https://doi.org/10.1111/jpim.12398.
Maass, W., Parsons, J., Purao, S., Storey, V. C., & Woo, C. (2018). Data-driven meets Spiekermann, M. (2019). Data marketplaces: Trends and monetisation of data goods.
theory-driven research in the era of big data: Opportunities and challenges for Intereconomics, 54(4), 208–216.
information systems research. Journal of the Association for Information Systems, 19 Spohrer, J., & Maglio, P. P. (2008). The emergence of service science: Toward systematic
(12), 1. service innovations to accelerate Co-creation of value. Production and Operations
Management, 17(3), 238–246. https://doi.org/10.3401/poms.1080.0027.

14
C. Kaiser et al. International Journal of Information Management 59 (2021) 102335

Stocker, A., Kaiser, C., & Fellmann, M. (2017). Quantified vehicles. Business & Information Wamba, S. F., Akter, S., Edwards, A., Chopin, G., & Gnanzou, D. (2015). How ‘big data’
Systems Engineering, 59(2), 125–130. can make big impact: Findings from a systematic review and a longitudinal case
Surbakti, F. P. S., Wang, W., Indulska, M., & Sadiq, S. (2020). Factors influencing study. International Journal of Production Economics, 165, 234–246. https://doi.org/
effective use of big data: A research framework. Information & Management, 57(1), 10.1016/j.ijpe.2014.12.031.
103146. https://doi.org/10.1016/j.im.2019.02.001. Wamba, S. F., Gunasekaran, A., Akter, S., Ren, S. J., Dubey, R., & Childe, S. J. (2017). Big
Svahn, F., Mathiassen, L., & Lindgren, R. (2017). Embracing digital innovation in data analytics and firm performance: Effects of dynamic capabilities. Journal of
incumbent firms: How volvo cars managed competing concerns. MIS Quarterly, 41 Business Research, 70, 356–365. https://doi.org/10.1016/j.jbusres.2016.08.009.
(1), 2017 Mar 1. Wand, Y., & Weber, R. (1993). On the ontological expressiveness of information systems
Swan, M. (2013). The quantified self: Fundamental disruption in big data science and analysis and design grammars. Journal of Information Systems, (3), 217–237.
biological discovery. Big Data, 1(2), 85–99. Wand, Y., & Weber, R. (2002). Information systems and conceptual modeling: A research
Swan, M. (2015). Connected car: Quantified self becomes quantified car. Journal of agenda. Information Systems Research, 13(4), 363–376.
Sensor and Actuator Networks, 4(1), 2–29, 2015. Wiener, M., Saunders, C., & Marabelli, M. (2020). Big-data business models: A critical
Teece, D. J. (2007). Explicating dynamic capabilities: The nature and microfoundations literature review and multiperspective research framework. Journal of Information
of (sustainable) enterprise performance. Strategic Management Journal, 28(13), Technology, 35(1), 66–91. SAGE Publications Ltd.
1319–1350. Wittmann, J. (2017). Electrification and digitalization as disruptive trends: New
Tomiyama, T., Lutters, E., Stark, R., & Abramovici, M. (2019). Development capabilities perspectives for the automotive industry? In A. Khare, B. Stewart, & R. Schatz (Eds.),
for smart products. CIRP Annals, 68(2), 727–750. Phantom Ex Machina (pp. 137–159). Cham: Springer.
Venable, J., Pries-Heje, J., & Baskerville, R. (2016). FEDS: A framework for evaluation in Woźniak, P., Valton, R., & Fjeld, M. (2015). Volvo single view of vehicle: Building a big
design science research. European Journal of Information Systems, 25(1), 77–89. data service from scratch in the automotive industry. Proceedings of the 33rd Annual
https://doi.org/10.1057/ejis.2014.36). ACM Conference Extended Abstracts on Human Factors in Computing Systems, 671–678.
Venkataram, P. B. (2019). Connected vehicles-an ecosystem for services (No. 2019-28- https://doi.org/10.1145/2702613.2702972.
2447). SAE technical paper. Zolnowski, A., Anke, J., & Gudat, J. (2017). Towards a cost-benefit-analysis of data-
Volkswagen. (2018). Volkswagen develops the largest digital ecosystem in the automotive driven business models. In J. M. Leimeister, & W. Brenner (Eds.), In Proceedings of
industry. https://www.volkswagenag.com/en/news/stories/2018/08/volkswagen- 13th International Conference on Wirtschaftsinformatik (WI 2017) (pp. 181–195).
develops-the-largest-digital-ecosystem-in-the-automot.html. Universität St. Gallen.
Wagner, S., Willing, C., Brandt, T., & Neumann, D. (2015). Data analytics for location- Zolnowski, A., Christiansen, T., & Gudat, J. (2016). Business model transformation
based services: Enabling user-based relocation of Carsharing vehicles. Proceedings of patterns of data-driven innovations. In Proceedings of European Conference on
the Thirty Sixth International Conference on Information System (ICIS 2015). Information Systems (ECIS 2016).

15