Data and Digital Platforms in Industry: Implication for enterprises strategies and governance

Butollo, Florian; Schneidemesser, Lea Research Report Data and Digital Platforms in Industry: Implication for enterprises strategies and governance Weizenbaum Series, No. 19 Provided in Cooperation with: WZB Berlin Social Science Center Suggested Citation: Butollo, Florian; Schneidemesser, Lea (2021) : Data and Digital Platforms in Industry: Implication for enterprises strategies and governance, Weizenbaum Series, No. 19, Weizenbaum Institute for the Networked Society - The German Internet Institute, Berlin, https://doi.org/10.34669/wi.ws/19 This Version is available at: https://hdl.handle.net/10419/251213 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. http://creativecommons.org/licenses/by/4.0/ Florian Butollo and Lea Schneidemesser Data and Digital Platforms in Industry Implication for enterprises strategies and governance OCTOBER 2021 WORKING PAPER 19WEIZENBAUM SERIES \ \ \ Data and Digital Platforms in Industry Florian Butollo* \ [email protected] Lea Schneidemesser* \ [email protected] * Weizenbaum Institute for the Networked Society and Berlin Social Science Center, Berlin, Germany ISSN 2748-5587 \ DOI 10.34669/WI.WS/19 EDITORS: The Managing Board members of the Weizenbaum-Institut e.V. Prof. Dr. Christoph Neuberger Prof. Dr. Sascha Friesike Prof. Dr. Martin Krzywdzinski Dr. Karin-Irene Eiermann Hardenbergstraße 32 \ 10623 Berlin \ Tel.: +49 30 700141-001 [email protected] \ www.weizenbaum-institut.de TYPESETTING: Roland Toth, M.A. COPYRIGHT: This series is available open access and is licensed under Creative Commons Attribution 4.0 (CC-BY_4.0): https://creativecommons.org/licenses/by/4.0/ WEIZENBAUM INSTITUTE: The Weizenbaum Institute for the Networked Society – The German Internet Institute is a joint project funded by the Federal Ministry of Education and Research (BMBF). It conducts interdisciplinary and basic research on the changes in society caused by digitalisation and develops options for shaping politics, business and civil society. This work has been funded by the Federal Ministry of Education and Research of Germany (BMBF) (grant no.: 16DII121, 16DII122, 16DII123, 16DII124, 16DII125, 16DII126, 16DII127, 16DII128 – “Deutsches Internet-Institut”). Abstract This article explores the position of industrial internet platforms (IIP) in manufacturing value chains. We develop an understanding of the role of data in global value chains (GVCs), referring to literature on intangible assets and theories on platform business models. We use data from a qualitative empirical study based on 45 interviews on platforms active on the German market to answer (1) whether there are tendencies of oligopolization that lead to an accumulation of power on the side of the platforms, and (2) whether it is the platforms that capture most of the gains derived from higher productivity or lower transaction costs. The analysis shows that platforms mainly act as service providers and/or intermediaries that support manufacturing companies in reaping benefits from data. While the relationship between platforms and manufacturers currently corresponds to a symbiosis, a stronger power imbalance could evolve in the future since processes of oligopolization are likely. Table of Contents 1 Introduction 5 2 Intangible assets and value distribution in GVCs 6 3 Intangibles and data 7 4 Platforms as agents in value chains: analytical cornerstones for the empirical analysis 8 4.1 Platforms and their functions 9 4.2 Platform business models 10 5 Research design and methods 11 6 Production-centred platforms: infrastructure oligopolies or service providers? 12 6.1 Value proposition: facilitating the use of data for enhanced productivity 12 6.2 Platform architectures: balancing relationships to customers and complementors 13 6.3 Revenue models: participating from productivity gains through services provision 15 6.4 Discussion: constraints to oligopolization 16 DATA AND DIGITAl PlATFoRMS IN INDUSTRY 7 Distribution-centred platforms: introducing e-commerce to industry 17 7.1 Value Proposition: reducing transaction costs 17 7.2 Platform Architectures: curating the manufacturing network 17 7.3 Revenue model: fees on transactions 19 7.4 Discussion: towards e-commerce oligopolies in industry 19 8 Conclusion 20 References 21 Endnotes 24 DATA AND DIGITAl PlATFoRMS IN INDUSTRY 1 Introduction New digital technologies are about to transform the economy as we know it. By combining steep increases in computing power, the abundance of data from all sorts of transactions and new methods of analysing and learning from such data (Brynjolfsson & McAfee, 2014), equipment manufacturers and software providers are able to offer a broad variety of ‘new digital technologies’ (Sturgeon, 2019) which promise to enhance the capacities of their (indust rial) customers. Partly this is about new technological artefacts. There is significant technological progress in the fields of collaborative robotics, modularized automation lines, digital assistance systems, 3D printers and other types of material equipment. In this contribution, however, we depart from the hypothesis that the more fundamental changes for industrial organization rest on developments that one cannot touch or see: the recursive processes of data generation, analysis and usage that increasingly shape business models and enterprise organization in global value chains (GVCs). The term ‘industrial internet’ describes such new possibilities for process rationalization and business model innovation related to the analysis of data in the industrial context. Such options concern the optimization of processes (e.g. production scheduling, maintenance, quality control), the improvement of products by making use of life-cycle data (i.e. connected car, smart home, etc.), and the databased match making in business-to-business (B2B) transactions. Digital platforms (henceforth: “Industrial Internetplatforms”, IIP) are important facilitators of such approaches. Just like in the field of the consumer-oriented internet, they take on the position as infrastructures of digitized transactions and enable manufacturers to take advantage of software applications to analyse industry-related data (Acatech, 2015; BDI, 2019; Graff, Krenz, & Kronenwett, 2018). The effects of such transformations on GVCs and the specific roles that IIP take on within them are virtually unknown. Most of the research on the digitalization of manufacturing has focused on technological artefacts like robots or digital assistance systems and their implications on the shopfloor (Briken, Chillas, Krzywdzinski, & Marks, 2017; Ford, 2016; Hirsch-Kreinsen, 2016). Debates on the impact of digital platforms on economic organization, on the contrary, have focused on the role of large tech companies of the consumer-oriented internet that have disrupted the field of media, communication and retail (Dolata, 2015; Kenney & Zysman, 2016; Staab, 2019; Zuboff, 2015). The industrial internet and platforms as its infrastructural backbone are still at an early stage of implementation as they become only meaningful with their diffusion in enterprises, which has just begun to take off. Correspondingly, empirical research that traces the possible outcomes of the platformization of industries on GVCs is scarce. First contributions have outlined possible trajectories with regard to the opportunities for industrial upgrading of suppliers (Humphrey, 2018; Sturgeon, 2019), the competition between tech companies and manufacturers in the field (Lechowski & Krzywdzinski, 2019; Ziegler, 2020), and possible effects on the governance of industries (Lüthje, 2019; Thun & Sturgeon, 2019, Author). Our contribution adds to the emerging literature on the subject by focusing on the relationship between industrial companies and IIP. By means of a qualitative study on the business practices of IIP in Germany, we aim to answer the question of whether IIP as economic agents will assume an equally powerful position in the industrial field as their peers in the consumer-oriented internet. More specifically we ask: (1) Do IIP emerge as agents that capture most of the gains from higher productivity and/or lower transaction costs, or does the relationship between IIP and industrial customers amount to a symbiosis with shared benefits? (2) Are there tendencies of oligopolization in the field of IIP? DATA AND DIGITAl PlATFoRMS IN INDUSTRY \ 5 To answer these questions, we first develop a theoretical understanding of the role of IIP in GVCs by discussing the relationship between ‘intangible assets’, data, and platforms (sections 2 and 3). We then operationalize these insights and introduce the subject-matter and the methods of our investigation, focusing on two types of platforms: productioncentred and distribution-centred platforms (sections 4 and 5). In section 6 and 7 the empirical material is presented with a focus on platform business models and the variables that define their position in GVCs. In the final section we conclude that due to significant differences between business models in the consumer-oriented and the industrial internet, the position of IIP rather resembles one of strategically important service providers and/ or intermediaries that participate in the value creation networks of digitalized manufacturing than that of an oligopoly that expands its reach on cost of manufacturers. However, tendencies of an oligopolization could evolve in the future, especially in the field of distribution-centred platforms. 2 Intangible assets and value distribution in GVCs The strategic role of IIP is linked to the increasing significance of data in fragmented production networks. The application of the Internet of Things (IoT) as a means of generating and connecting data from industrial processes radically enhances the volumes and accuracy of up-to-date (or even real-time) data (Brynjolfsson & McAfee, 2014; Sturgeon, 2019). Artificial Intelligence provides new possibilities to make economic use of this data by detecting patterns, making predictions and improving processes based on the sheer amount of available data and distributed computing power. Even though the economic significance of data, often dubbed the ‘new oil’, is widely recognized, their role for inter-firm relations in GVCs is not theoretically explored sufficiently with few exceptions (Foster & Graham, 2017; Sturgeon, 2019). The role of knowledge-intensive production factors described as ‘intangible assets’, however, lies at the core of theory building on GVCs (Durand & Milberg, 2020; Kaplinsky, 2020; Mudambi, 2008). In what follows, we will first review the existing insights on intangibles as they were taken up in GVC theory and then discuss the role of data in this context, which we interpret as an increasingly important resource for the production of intangibles. The term ‘intangibles’ refers to intellectual or knowledge assets (Lev, 2001). These can comprise of legally defensible titles such as patents, copyrights and brands but also consist of organizational structures, inter-organizational relationships and human creativity (Mudambi, 2008). It has been empirically shown that intangible assets, in spite of some inherent problems regarding their monetarization, are generating an increasing share of returns, roughly a third of all production factors (Alsamawi et al., 2020; Mudambi, 2008). According to Haskel and Westlake (2017) the measurable impact of intangibles is only partially represented in its de-facto impact on business models and competition. In a knowledge-intensive ‘capitalism without capital’, the generation of rents through the capture and monetarization of intangibles plays an ever more prominent role. Crucially, intangibles are allocated unevenly in disintegrated value chains. Intangible assets tend to be concentrated in activities that are allocated prior or after the actual manufacturing process, i.e. in R&D or design activities on the one hand and in marketing, advertising and after-sales services on the other (Mudambi, 2008). This polarization is often explained in alignment to Vernon’s product life cycle model: pure-play manufacturing activities can easily be replicated (especially by firms in emerging DATA AND DIGITAl PlATFoRMS IN INDUSTRY \ 6 economies). They hence become ‘commoditized’, i.e. easily exchanged by other suppliers in off-theshelf transactions, and are exposed to price pressures. Preand post-production activities, on the contrary, are more difficult to copy and often include a service dimension that is customized according to users’ preferences (Kaplinsky, 2020; Mudambi, 2008). While empirical studies on some industries confirmed this pattern (e.g. Ali-Yrkkö, Rouvinen, Seppälä, & YläAnttila, 2011; Timmer, Erumban, Los, Stehrer, & de Vries, 2014), the equation of low-value added activities with manufacturing is oversimplified. Especially in innovation-intensive producer-driven commodity chains (Gereffi, 1994), value creation crucially depends on the permanent adjustment of processes in recursive innovation processes that are partly related to practical shop floor knowledge (Herrigel & Zeitlin, 2010; Nahm & Steinfeld, 2014). 3 Intangibles and data The question of whether or not a firm can develop intangibles touches a great variety of questions from the general characteristics of a region’s innovation system, the innovative capabilities of a firm, the conditions for technology transfer to the availability of a suitably trained workforce and the specific company cultures. While some of these factors rely on the general institutional and political context in which GVCs are embedded and some remain the domain of proper lab-level basic innovation, others rely on incremental improvements of products and processes based on information that is gathered from customers or shop floor experiences (Herrigel, 2018; Herrigel & Zeitlin, 2010). This requires feedback loops from customers’ user experience to product developers (product innovation) or from shop floor performance to process design (process innovation). As Michael Porter and Victor Miller (1985) argue, ‘[e]very value activity has both a physical and an information-processing component. The physical component includes all the physical tasks required to perform the activity. The information-processing component encompasses the steps required to capture, manipulate, and channel the data necessary to perform the activity.’ The information-processing component can be used to manipulate and improve the physical component. The history of industrial organization to a significant degree revolves around the question of how to make use of information derived from manufacturing processes and markets (Baukrowitz et al., 2006). Taylorist scientific management, for instance, rested on a detailed mapping of the work process by taking the time of each production step manually and using this data to comprehensively redesign the workflow. The organizational revolution of lean production in the 1990s increased flexibilization by improving the way information was transmitted along the supply chain based on Kanban and Kaizen techniques (Womack, Jones, & Roos, 1990). As supply chains disintegrated and became more complex, rationalization became a matter of ‘systemic rationalization’ of the supply chain (Altmann, Deiß, Döhl, & Sauer, 1986), resulting in the rise of supply chain management as a separate management discipline and systematic supply chain monitoring as one of its major instruments. All of these processes were accompanied by the intensification of ‘codification, standardization and monitoring of the workflow’ (Durand & Milberg, 2020, p. 408). The growing need for the coordination of processes in complex value chains and the possibilities to use software to facilitate the monitoring and recursive adaptation of processes gave rise to industrial information systems, in particular systems for supply chain management, Enterprise Resource Planning (ERP) and Manufacturing Execution (MES). Such software facilitated the adjustment of production DATA AND DIGITAl PlATFoRMS IN INDUSTRY \ 7 processes to market demand based on productionrelated data. In addition, social media data and data on B2C transactions also began to play an important role in detecting consumers’ preferences and developing appropriate marketing and product design strategies. Brand building, a prerequisite for rent generation in consumer industries, increasingly relied on market intelligence, i.e. data on consumer behaviour (Pfeiffer, 2021; Rikap, 2020). Digital data thus has played an ever increasing role for firms’ abilities of product design and (dynamic) process innovation. [1] The technological progress towards the IoT, i.e. the ability to generate high-resolution data from real-life processes and to connect this data from different devices at a unitary data layer, enhances the possibility to support key enterprise functions through data-based intangibles (Ziegler, 2020, pp. 27–52). Progress in Machine Learning but also more traditional methods of data analysis can help to utilize large data sets in order to detect patterns, predict future developments and integrate automated decision making in management functions. IIP are needed to integrate and use this data in order to use software applications and to improve the matchmaking between industrial customers and suppliers. 4 Platforms as agents in value chains: analytical cornerstones for the empirical analysis As there’s an enhanced importance of data that can be utilized in order to generate value, the question of how it can be used and who benefits from it becomes paramount. For this end, firms need to rely on a cloud infrastructure and on platform solutions that can connect different sets of data and integrate software applications to analyse it. This provides industrial customers with advantages of enhanced productivity and/or reduced transaction costs, but it also puts the owners of cloud and platform services in a potentially powerful position, particularly if IIP owners can acquire and monetarize customers’ data. Such strategies have been a cornerstone of platform business models in the consumer-oriented internet, where platforms sell user data for advertising purposes (Srnicek, 2016; Zuboff, 2015). However, it can be expected that tighter requirements for the secrecy of the data by industrial customers constitutes a limitation to replicate such strategies in the field of the industrial internet. Thus the conditions under which platform business models can expand and the potential effects on the relationship between platforms and manufacturers need to be investigated in order to arrive at a concrete analysis of power relations in this emerging field. A related question concerns the oligopolization of platforms. In the consumer-oriented internet, digital platforms in the field of e-commerce (Amazon), social media (Facebook), and web services (Google) soon reached a market-dominating position. Their success rests on the creation of ecosystems that offer customers attractive options through network effects and other distinct features of platformbased business models (Abdelkafi, Raasch, Roth, & Srinivasan, 2019; Cusumano, Gawer, & Yoffie, 2019; Dolata, 2015). As platforms in the industrial realm replicate some of the strategies of their peers in the consumer-oriented internet, similar processes of oligopolization might emerge. Based on these considerations, we pursue the following research questions in our empirical study: (1) Are there tendencies of oligopolization in the field of IIP? DATA AND DIGITAl PlATFoRMS IN INDUSTRY \ 8 One thing is we develop the infrastructure in the first place. […]. But then of course there are many third parties out there who then develop compatible solutions. And here we are doing our best to promote the largest possible ecosystem, because that’s what will decide success or failure at the end of the day. Nobody at [platform name] can do that himor herself. [...] (PC1) This need for compatibility also pushes the platforms towards openness with regard to their core software elements. This is described by Ziegler (2020, pp. 247–254) in his case study of a IIP run by an industrial company that progressively opened up its software development until fully turning Open Source in order to maximise its compatibility and integration with SaaS complementors. The need to attract a vibrant ecosystem of SaaS providers around the platforms’ core affects the bargaining position of platforms. Ideally, they can offer access to a large number of potential customers by integrating SaaS elements into their platform ecosystem. Unlike with the app stores Google Play or Apple App Store in the consumer-oriented internet that monopolize the distribution of smart phone apps, software developers can distribute their apps through different channels and quit the relationship with a platform altogether in case their interests are not met. The platforms need to negotiate and cooperate with their complementors on an equal footing, at least as long as the platform landscape remains fragmented and no single platform emerges as a dominant channel through which software applications are distributed. 6.3 Revenue models: participating from productivity gains through services provision The integration of generic software elements into industrial processes requires a great amount of adaptation and specification. Often this also involves the installation of infrastructure (sensors, hardware hubs, edge computing devices and the like). All interviewed representatives of production-centred platforms stressed the high requirements on domain-specific knowledge. Standardized and generic data-analysis tools have to be adjusted to the requirements of each specific domain including the fine-tuning between the software and the specific machinery or equipment it is integrated with. A generic visual recognition tool, for instance, needs to be adjusted, according to whether it is integrated into public transport vehicles or production machinery (IP1). The revenue model of all surveyed production-centred platforms therefore aims at the sale (or leasing) of SaaS applications in combination with consulting services that concern the implementation of IoTprojects. Production-centred platforms therefore typically combine the function of a software distributor with consulting services: they provide assistance with adjustments between production equipment and software elements to customers that lack the specific knowledge of how to capitalize on the data generated through new digital devices. Most platforms (PC2, PC4, PC5) offer off-theshelf monthly subscription plans to get access to more or less comprehensive platform functions. These can be complemented by more customized packages of apps and services. However, platforms can find additional sources of revenue as well, as the case of Siemens Mindsphere demonstrates. On the Mindsphere platform also complementors have to subscribe to access the platforms developer tools. Furthermore, Siemens as a company with a background in automation equipment, offers hardware components – the Mind Connect Elements – that facilitate a frictionless integration of devices by any manufacturer with the Mindsphere platform. (Siemens, 2017, pp. 8–9) The service-centred character of the platforms’ business models means that the revenues of platforms are highly dependent on the success of their customers. There are strong reservations among manufacturers against paying for services and equipment without having any security about the concrete economic gains that can be achieved. DATA AND DIGITAl PlATFoRMS IN INDUSTRY \ 15 Some platform representatives reported that their customers enter the business relationship with a very pragmatic stance and that many were still waiting for proofs of concrete benefits (PC2a, PC2b, PC4a). The platforms’ business models thus depend on the fate of their customers: only if productivity gains actually materialize, they will be able to benefit from such progress. 6.4 Discussion: constraints to oligopolization Production-centred IIP are acquiring a position in industrial GVCs that is of increasing importance: the function of enabling corporate customers to take advantage of data in order to improve processes and raise productivity. Unlike their peers in the consumer-oriented internet, however, the digital platforms of the industrial realm face stronger obstacles to acquire a dominant position which would enhance their bargaining power visà-vis other network participants. Especially the constraints for unleashing same-side network effects, that are rooted in the unwillingness of corporate customers to share their data, makes an easy road towards market dominance in ‘winnertakes-it-all’ markets impossible. It also is a barrier to revenue models that focus on the secondary usage of data, one of the main sources of revenue in the consumer-oriented internet. Several interview partners nonetheless expected a consolidation of the market segment to a handful of large players. According to this perspective, only some platforms will manage to build vibrant software ecosystems while attracting a large number of industrial customers. One representative of a prominent platform assumed that the field will be fragmented along industry boundaries, whereas one platform might turn out as the main beneficiary in each segment (PC2a). However, due to the need to ensure interoperability with a heterogeneous hardware landscape and to attract SaaS complementors to the platform, production-centred platforms have to balance their monetary self-interest with the need to cooperate with complementors and customers on an equal footing. Or, in the words of one interviewee: “A platform is not by itself relevant. It is relevant in combination with its apps” (PC2b). Crucially, the success of production-centred platforms depends on the capacity of their customers to raise productivity based on the provision of SaaS elements. The primary objective of productioncentred platforms is not to grow on cost of their customers, but to acquire a share of the prospected productivity. In a long-term perspective, however, the strongest production-centred platforms could gain bargaining power vis-à-vis other market participants in case they evolve towards oligopolies that could implement de facto standards. Even if the platforms’ ecosystems remained open, such a scenario would increase switching costs for customers and thus create lock-in effects. DATA AND DIGITAl PlATFoRMS IN INDUSTRY \ 16 7 Distribution-centred platforms: introducing e-commerce to industry 7.1 Value Proposition: reducing transaction costs Distribution-centred platforms function as B2Bmarketplaces by positioning themselves as matchmakers between manufacturers of supply products and their industrial customers. The platforms do not operate any manufacturing facilities by themselves but instead cooperate with networks of manufacturers. For the three platforms we studied these partner-networks range between 240–6,000 companies. One industry in which the distribution-centred platform model is expanding fast is mechanical component manufacturing. In Europe and the US, platform owners, usually start-ups that combine IT capabilities with a good knowledge of their target industries, act as intermediaries between mechanical component manufacturers and their customers.[5] The manufacturing services offered by complementors include laser, plasma and waterjet cutting, CNC turning, milling and drilling, and 3D printing for different kinds of metals and synthetic materials. The manufacturers offer their production capacities via the platform to customers in industries such as machine tools, aerospace, robotics, automotive and medical devices. The platforms value proposition is very straight forward: they reduce transaction costs for their customers. They do so by reducing the time traditionally invested in finding and auditing suppliers and by simplifying interactions through a platform as digital interface, thus automating and standardizing the handling of orders, payment and contracts. One representative claimed that his platform reduces the time traditionally needed to fulfil an order by 50% (DC 1). Furthermore, the network of complementors allows for more flexibility especially at delivering time-sensitive orders. The disruptions to GVCs during the outbreak of Covid19 in the beginning of 2020 accordingly have given these platforms a boost, as lead firms had to reorganize their supply chains and often looked for short-term replacements for failed transactions. 7.2 Platform Architectures: curating the manufacturing network Distribution-centred platforms curate their network of manufacturing partners through rules of access and continuous performance-evaluation. As we will show, they can take advantage of cross-side network effects and some of the more advanced platforms utilize data to improve their services. This has significant consequences for manufacturers. To become part of a platform’s network of manufacturing partners, firms have to provide detailed information on the production processes they offer and they have to pass a trial phase during which orders are closely monitored and evaluated along dimensions such as product quality, punctuality and fast communication. The mechanisms of monitoring and evaluation are institutionalized and become permanent over the course of the cooperation and they affect the likelihood to receive orders in the future. The factors that determine whether a particular manufacturer is chosen, however, remain opaque for manufacturers (DC2c, DC3). To create and expand their network, distributioncentred platforms make use of cross-side network effects. The platforms attractiveness to customers mainly rests on the range and flexibility of the manufacturing services offered. Both are largely rooted in the size and diversity of the – in some DATA AND DIGITAl PlATFoRMS IN INDUSTRY \ 17 cases global – network itself rather than in the flexibility of individual manufacturing partners (Butollo & Schneidemesser, 2021). Seen from the complementors’ side, a platform is most relevant if it has a solid base of industrial buyers which generates a steady flow of orders. Therefore, the expansion of the platform needs to achieve a good balance towards both user groups. The three distribution-centred platforms we studied utilize the data they gather from transactions with customers and manufacturing partners in order to improve their services, but the extent to which they do so varies. One platform curates its network of manufacturers manually and allocates orders via email and telephone (DC2). Other platforms are particularly active in developing software elements such as instant pricing tools that automate aspects of the business relationship (in this case by calculating a binding price for a certain product). Such AI-based tools record the properties of a technical drawing, requested materials, required processing techniques, lead time and batch size and compare them with automatically-generated benchmarks from its vast database. One platform representative claimed that their self-learning algorithm has already analysed more than one million CAD-files to automatically calculate prizes (DC3). Another interviewee expressed the vision to establish a ‘universally agreed price for manufacturing that reflects supply and demand in the global market place’ based on the recorded data (DC1). Such approaches could tilt the bargaining relationship in favour of the platforms and result in enhanced competitive pressures for the manufacturing partners because the conditions become non-negotiable. This is particularly the case when platforms encompass firms from low-wage and high-wage countries that are put into direct competition with each other. Manufacturers then effectively are benchmarked against the globally most efficient, fastest and cheapest participants in the market. The same accounts for delivery lead times. The manager of one distribution-centred platform explains: One of the things that we’re doing is kind of levelling standards across the globe. So, for example […] normally, the lead time in Europe is four to six weeks, while in America and Asia one to two weeks is normal. […] And so, when we talk to European CNC suppliers, we tell them like, our standard auction for customers is two weeks. (DC1) However, according to our interviews with platforms’ manufacturing partners, such pricing strategies are not necessarily experienced as a race to the bottom. Some of the manufacturers of our sample report no differences in revenues between orders that were transmitted through the platforms and orders that were received directly from customers without interference of platforms (com1DC, com2DC). Yet others reported that revenues on some orders processed through the platforms are significantly lower (com3DC, com4DC). Another aspect that concerns the power relation between platforms and complementors is the relationship to customers. Manufacturing partners are cut off from direct interactions with industrial customers as it is the platforms that mediate these interactions, which are careful not to enable direct interactions between the two sides of their ecosystem. Direct relationships of manufacturers to customers, however, not only provide a certain stability in the business relationships, but also provide opportunities for more lucrative consulting activities that concern the preproduction processes. As a direct interaction between manufacturing partners and customers is obstructed through the platforms, such services are monopolized by them while manufacturers only perform core manufacturing tasks, a potential functional impoverishment for the involved firms. Conversely, one platform in our sample was particularly ambitious in developing such preproduction services by utilizing the vast amount of data on machine-part-designs and CAD-drawings that is uploaded to the platforms by their customers for automated testing of the manufacturability of designs or lead-time calculation (DC3). This DATA AND DIGITAl PlATFoRMS IN INDUSTRY \ 18 shows how – unlike the case of production-centred platforms – secondary use of data (CAD-designs as well as log-data) is exercised and provides a range of possibilities for distribution-centred platforms to extend their abilities to capture value from intangible functions. 7.3 Revenue model: fees on transactions Distribution-centred platforms in the mechanical component industry earn a commission on every order that is handled. They charge for the matchmaking with a suitable supplier and for certain pre-production services. The platforms provide an attractive option for industrial customers to source components as they provide access to a flexible network of producers at very competitive prices. This offer is especially attractive in low volumes/ high mix industries in which customers often have the need to order very specific components. In such fields distribution-centred platforms have become relevant supplements to regular supplychain management practices that rely on direct transactions between producers and customers. 7.4 Discussion: towards e-commerce oligopolies in industry Distribution-centred platforms enable industrial customers to source components more efficiently. This mainly implies a special kind of supplier governance through rules and evaluations set by the platform, the reliance on cross-side network effects and the charging of commissions on each transaction. As with production-centred platforms, there is a symbiosis between platforms and industrial companies that are involved in the platform ecosystems: manufacturing partners receive steady orders while the platforms benefit from their matchmaking activities. For manufacturing partners this is a two-edged sword: on the one hand they can benefit from additional sales channels and easy access to new customers, on the other hand they might lose opportunities in pre-production functions and become subjected to enhanced competitive pressure. While effects on manufacturing partners thus are ambivalent, distribution-centred platforms prospectively will emerge as the main beneficiaries in the process, capturing large shares of revenues from the savings in transaction costs. If distributioncentred business models turn out to be scalable and diversify across industries, an oligopolization that resembles developments in B2C e-commerce could be possible with similar implications: a growing dependency of sellers and customers. However, neither is it clear that market developments automatically result in oligopolistic structures (i.e. a fragmented landscape of specialized platforms could be an alternative scenario), nor can it be taken for granted that manufacturing partners (particularly those in low-wage countries) will suffer from the heightened competitive pressures. After all, many small-scale manufacturers choose to participate in the manufacturing networks of distribution-centred platforms because they benefit from the option of flexibly accepting orders that often are supplementary to their regular customer relationships (com3DC, com4DC). This way, they can indirectly access a market that lies beyond their (often regionally confined) reach and improve the utilization of their production capacities. DATA AND DIGITAl PlATFoRMS IN INDUSTRY \ 19 8 Conclusion Our discussion of the role of IIP in GVCs departed from reflections on the enhanced role of data in industrial processes and the observation that data often constitutes the raw material for ‘intangible assets’ that constitute an important variable for the distribution of revenues across firms. In the field of the consumer-oriented internet, digital platforms emerged as infrastructures to take advantage of new data-based business models. They acquired an extraordinarily strong economic role as oligopolies of the digital economy and challenged traditional companies in the sectors they are active in (e.g. retail, media, communication). Our article thus discussed the question of whether there could be similar tendencies at work in the realm of the industrial internet by empirically analysing two prominent platform types, production-centred and distribution-centred platforms. The results of our investigation help to better understand the position that IIP acquire in GVCs. The analysed business models do not mainly aim at the usurpation of industrial data with the goal of monopolizing intangible assets, i.e. those resources that are paramount to capture value. Rather, they act as service providers and/or intermediaries that support manufacturing companies in reaping benefits from data, i.e. raising productivity of manufacturing processes or lowering transaction costs through efficient matchmaking. The success of both types of platform business models essentially depends on the capacity of their customers to generate revenue to which they contribute capabilities and of which they demand a share. Hence the issue at stake is not whether industrial platforms will outcompete or replace industrial companies, but whether they emerge as strong service providers that maximise their revenues vis-à-vis traditional manufacturers. Their ability to do so decisively depends on the platforms’ ability to acquire power based on network effects. In this sense the trajectories of both platform types are different: Production-centred platforms cannot harvest same-side network effects as long as there remain obstacles to the sharing of industrial data. Instead, they need to curate a diverse network of complementors in order to create cross-side network effects, which is only possible if they keep their ecosystems open. Cross-side network effects could result in oligopolization as platforms mature, but there is also the option of a fragmented landscape of more specialized platforms that operate in the niches of their expertise (Sturgeon, 2019, p. 15). Distribution-centred platforms, on the contrary, show many similarities with regular e-commerce platforms with potentially strong cross-side network effects. This could enhance their power vis-à-vis industrial complementors and thus their leeway for charging for transaction services. What is more, these platforms do record the data from transaction processes, which implies an information asymmetry vis-à-vis their complementors (Staab, 2019) that allows them to improve and expand their match making qualities and their pre-production services. In both surveyed fields, oligopolization eventually might preponderate. This mainly means that IIP will stabilize their position in GVCs. As in the consumer-oriented internet this means that they might replace traditional contenders in the field. In the case of production-centred platforms this mainly accounts for non-platform software distributers (not manufacturers). In the field of distribution-centred platforms this not only accounts for traditional trade intermediaries, but also for single manufacturers that pursue the goal of fast on-demand production of customized products. On a different path that harvests the flexibility of the network, distributioncentred platforms thus could deliver, what the engineering-heavy strategy of Industry 4.0 promises (Butollo & Schneidemesser, 2021). DATA AND DIGITAl PlATFoRMS IN INDUSTRY \ 20 References Abdelkafi, N., Raasch, C., Roth, A., & Srinivasan, R. (2019). Multi-sided platforms. Electronic Markets, 29(4), 553–559. https://doi.org/10/ ghgn3g Acatech. (2015). Smart Service Welt: Recommendations for the Strategic Initiative Web-based Services for Businesses. Final report. Berlin. Retrieved from https:// en.acatech.de/publication/recommendationsfor-the-strategic-initiative-web-based-servicesfor-businesses-final-report-of-the-smart-serviceworking-group/ Ali-Yrkkö, J., Rouvinen, P., Seppälä, T., & YläAnttila, P. (2011). Who Captures Value in Global Supply Chains? Case Nokia N95 Smartphone. Journal of Industry, Competition and Trade, 11(3), 263–278. https://doi.org/10/csx58r Alsamawi, A., Cadestin, C., Jaax, A., Guilhoto, J., Miroudot, S., & Zurcher, C. (2020). Returns to intangible capital in global value chains: New evidence on trends and policy determinants. OECD Trade Policy Papers, (240). https:// www.ilo.org/public/english/revue/ Altmann, N., Deiß, M., Döhl, V., & Sauer, D. (1986). Ein „Neuer Rationalisierungstyp‘— Neue Anforderungen an Die Industriesoziologie. Soziale Welt, 37(2/3), 191–207. Baukrowitz, A., Berker, T., Boes, A., Pfeiffer, S., Schmiede, R., & Will-Zocholl, M. (2006). Informatisierung der Arbeit—Gesellschaft im Umbruch (1st ed.). Berlin: edition sigma. BDI. (2019). Deutsche digitale B2B Plattformen. Retrieved from https://bdi.eu/publikation/news/ deutsche-digitale-b2b-plattformen/ Briken, K., Chillas, S., Krzywdzinski, M., & Marks, A. (Eds.). (2017). The new digital workplace: How new technologies revolutionise work. London: Palgrave Macmillan. Brynjolfsson, E., & McAfee, A. (2014). The second machine age: Work, progress, and prosperity in a time of brilliant technologies (First Edition). New York: W. W. Norton & Company. Butollo, F., & Schneidemesser, L. (2021). Beyond “Industry 4.0”: B2B factory networks as an alternative path towards the digital transformation of manufacturing and work. International Labour Review, ilr.12211. https:// doi.org/10.1111/ilr.12211 Chesbrough, H. W. (2003). Open Innovation: The New Imperative for Creating and Profiting from Technology (First Trade Paper Edition). Boston, Mass: Harvard Business Review Press. Cusumano, M. A., Gawer, A., & Yoffie, D. B. (2019). The business of platforms: Strategy in the age of digital competition, innovation, and power (First edition). New York, NY: Harper Business, an imprint of HarperCollinsPublishers. Dolata, U. (2015). Volatile Monopole. Konzentration, Konkurrenz und Innovationsstrategien der Internetkonzerne. Berliner Journal für Soziologie, 24(4), 505–529. https://doi.org/10/ghdhqx Durand, C., & Milberg, W. (2020). Intellectual monopoly in global value chains. Review of International Political Economy, 27(2), 404– 429. https://doi.org/10/ghmjzw Fleisch, E., Weinberger, M., & Wortmann, F. (2014). Geschäftsmodelle im Internet der Dinge. BOSCH, Universität St. Gallen. Retrieved from BOSCH, Universität St. Gallen website: https:// www.iot-lab.ch/wp-content/uploads/2014/09/ GM-im-IOT_Bosch-Lab-White-Paper.pdf Ford, M. (2016). Rise of the Robots: Technology and the Threat of a Jobless Future (Reprint). New York: Basic Books. DATA AND DIGITAl PlATFoRMS IN INDUSTRY \ 21 Foster, C., & Graham, M. (2017). Reconsidering the role of the digital in global production networks. Global Networks, 17(1), 68–88. https://doi.org/10.1111/glob.12142 Gawer, A., & Cusumano, M. A. (2014). Industry Platforms and Ecosystem Innovation. Journal of Product Innovation Management, 31(3), 417– 433. https://doi.org/10.1111/jpim.12105 Gereffi, G. (1994). The Organization of BuyerDriven Global Commodity Chains: How U.S. Retailers Shape Overseas Production Networks. In G. Gereffi & M. Korzeniewicz (Eds.), Commodity Chains and Global Capitalism (pp. 95–122). Westport: Praeger. Graff, J., Krenz, W., & Kronenwett, D. (2018). IIoT Platforms: Sources of Profit or Inflated Hype? How Machinery Companies Should Target Their Investments to Maximize Benefits. (p. 9). Oliver Wyman. Retrieved from Oliver Wyman website: https://www. oliverwyman.de/content/dam/oliver-wyman/v2/ publications/2018/november/perspectives-onmanufacturing-industries-cover-story.pdf Haskel, J., & Westlake, S. (2017). Capitalism without Capital: Rise of Intangible Economy. Princeton, New Jersey: Princeton Univers. Press. Herrigel, G. (2018). Experimentalist systems in manufacturing multinationals: German automobile and machinery examples. Critical Perspectives on International Business, 14(4), 362–382. https://doi.org/10.1108/ cpoib-10-2017-0063 Herrigel, G., & Zeitlin, J. (2010). Inter‐ Firm Relations in Global Manufacturing: Disintegrated Production and Its Globalization. In G. Morgan, J. L. Campbell, C. Crouch, O. K. Pedersen, & R. Whitley (Eds.), The Oxford Handbook of Comparative Institutional Analysis. Oxford: Oxford University Press. Retrieved from https:// www.oxfordhandbooks.com/view/10.1093/ oxfordhb/9780199233762.001.0001/oxfordhb9780199233762-e-19 Hirsch-Kreinsen, H. (2016). Digitization of industrial work: Development paths and prospects. Journal for Labour Market Research, 49(1), 1–14. https://doi.org/10/gg3kfw Humphrey, J. (2018). Value chain governance in the age of platforms. IDE Discussion Papers 714, Sussex. IW & DBI. (2021). Datenwirtschaft in Deutschland. Wo stehen die Unternehmen in der Datennutzung und was sind ihre größten Hemmnisse? Kaplinsky, R. (2020). Rents and inequality in global value chains. In S. Ponte, G. Gereffi, & G. Raj-Reichert (Eds.), Handbook on Global Value Chains (pp. 153–168). Cheltenham and Northhampton: Edward Elgar Publishing. Kenney, M., Rouvinen, P., Seppälä, T., & Zysman, J. (2019). Platforms and Industrial Change. Industry and Innovation, 26, 871–879. https:// doi.org/10/ggnbsx Kenney, M., & Zysman, J. (2016). The Rise of the Platform Economy. Issues in Science and Technology, 32, 61–69. Lechowski, G., & Krzywdzinski, M. (2019). Governing “digital transformation” through transnational technological standards: How collective standard-setting efforts reshape the industrialautomation sector [draft]. Lev, B. (2001). Intangibles: Management, Measurement, and Reporting. Washington, D.C: Brookings Institution Press. Lüthje, B. (2019). Platform Capitalism ‘Made in China’? Intelligent Manufacturing, Taobao Villages, and the Restructuring of Work. Science , Technology and Society, 24(2), 199–217. Mayring, P. (2015). Qualitative Inhaltsanalyse: Grundlagen und Techniken (Neuausgabe, 12., aktualisierte). Weinheim Basel: Beltz. McAfee, A., & Brynjolfsson, E. (2017). Machine, Platform, Crowd: Harnessing Our Digital Future. New York: Norton & Company. Mudambi, R. (2008). Location, control and innovation in knowledge-intensive industries. Journal of Economic Geography, 8(5), 699– 725. https://doi.org/10/cmr288 DATA AND DIGITAl PlATFoRMS IN INDUSTRY \ 22 Nahm, J., & Steinfeld, E. S. (2014). Scale-up Nation: China’s Specialization in Innovative Manufacturing. World Development, 54, 288–300. https://doi.org/10.1016/j. worlddev.2013.09.003 Pfeiffer, S. (2021). Digitalisierung als Distributivkraft: Über das Neue am digitalen Kapitalismus. Bielefeld: transcript. Platform Industrie 4.0. (n.d.). What is Industrie 4.0? Retrieved from Platform Industrie 4.0 website: https://www.plattform-i40.de/PI40/ Navigation/EN/Industrie40/WhatIsIndustrie40/ what-is-industrie40.html Porter, M. E., & Millar, V. E. (1985, July 1). How Information Gives You Competitive Advantage. Harvard Business Review. Retrieved from https://hbr.org/1985/07/how-information-givesyou-competitive-advantage Rikap, C. (2020). Amazon: A story of accumulation through intellectual rentiership and predation. Competition & Change, 102452942093241. https://doi. org/10.1177/1024529420932418 Siemens. (2017). MindSphere. The cloudbased, open IoT operating system for digital transformation. Retrieved from https:// new.siemens.com/uk/en/products/software/ mindsphere/mindsphere-white-paper.html Srnicek, N. (2016). Platform Capitalism (1.). Cambridge, UK ; Malden, MA: Polity. Staab, P. (2019). Digitaler Kapitalismus: Markt und Herrschaft in der Ökonomie der Unknappheit (Originalausgabe). Frankfurt am Main: Suhrkamp Verlag. Sturgeon, T. J. (2019). Upgrading strategies for the digital economy. Global Strategy Journal, 1–24. https://doi.org/10/ggcnfh/gsj.1364 Thun, E., & Sturgeon, T. (2017). When Global Technology Meets Local Standards: Reassessing the China’s Mobile Telecom Policy in the Age of Platform Innovation. MIT Working Paper Series, 1–54. Thun, E., & Sturgeon, T. (2019). When Global Technology Meets Local Standards: Reassessing the China’s Mobile Telecom Policy in the Age of Platform Innovation. In T. Rawski & L. Brandt (Eds.), Policy, Regulation and Innovation in China’s Electricity and Telecom Industries. Cambridge University Press. Retrieved from http://eureka.sbs.ox.ac.uk/7168/ Timmer, M. P., Erumban, A. A., Los, B., Stehrer, R., & de Vries, G. J. (2014). Slicing Up Global Value Chains. Journal of Economic Perspectives, 28(2), 99–118. https://doi.org/10/gftcvn Timmers, P. (1998). Business Models for Electronic Markets. EM - Electronic Markets, 28(2), 3–8. Werling, M., Weber, P., & Lasi, H. (2020). Partizipation. Im Spannungsfeld von PlattformGiganten, staatlicher Datentreuhand und genossenschaftlich-kooperativen Ansätzen. (p. 6) [Arbeitsbericht]. Ferdinand-SteinbeisInstitut. Retrieved from Ferdinand-SteinbeisInstitut website: https://steinbeis-fsti.de/ wp-content/uploads/20-09-15-Arbeitsbericht_ Datengenossenschaften.pdf Womack, J. P., Jones, D. T., & Roos, D. (1990). The machine that changed the world: Based on the Massachusetts Institute of Technology 5-million dollar 5-year study on the future of the automobile. New York: Rawson Associates. Ziegler, A. (2020). Der Aufstieg des Internet der Dinge: Wie sich Industrieunternehmen zu TechUnternehmen entwickeln. Frankfurt am Main ; New York: Campus. Zuboff, S. (2015). Big other: Surveillance Capitalism and the Prospects of an Information Civilization. Journal of Information Technology, 30(1), 75–89. https://doi.org/10.1057/jit.2015.5 Zysman, J., Murray, J., Feldman, S., Nielsen, N., & Kushida, K. (2011). Services with Everything: The ICT-Enabled Digital Transformation of Services. SSRN Electronic Journal. https://doi.org/10.2139/ssrn.1863550 DATA AND DIGITAl PlATFoRMS IN INDUSTRY \ 23 Endnotes [1] What is more, product markets are characterized by an increasing amount of digital services that are based on data. This is most evident in the telecommunication sector where the physical smart phone, merely acts as a carrier for a broad range of apps that can process data from daily interactions recorded through mobile devices (Thun & Sturgeon, 2019). Similar logics of an IoT-driven servitization of the economy are at work in the fields of connected cars, smart homes, smart cities and many other industries. The ability to acquire and process data and to develop digital service applications to this end becomes an important factor that shapes competition in a broad range of product equipment (Zysman, Murray, Feldman, Nielsen, & Kushida, 2011). In the field of mechanical engineering this means that some firms strive to develop software applications related to the steering of manufacturing processes and digital platforms to integrate such applications (Author, forthcoming). [2] For instance, the primary strategic objective of an innovation platform is the growth of an ecosystem that comprises of diverse complementors that add applications, whereas transaction platforms, while also striving to expand the size of their reach, need to constantly improve their matchmaking techniques in order to reduce frictions in transactions (Cusumano, Gawer, & Yoffie, 2019). [3] We use the following labelling system for quoting interviews: Each interviewee group has an abbriviation (IP=infrastructure provider, PC=production-centred platform, DC=distribution-centred platform, comPC=complementor of production-centred platform, comDC=complementor of distribution-centred platform). Each platform/complementor is assigned a number. If we refer to more than one interview conducted with the same platform/complementor this is indicated by small letters (a,b,c,…) after the number. E.g. the code PC2b refers to the second interview we had with a representative of production-centred platform number 2 in our sample. [4] The original German-language quotations are translated by the authors. [5] In China a similar distribution-centred platform model can be observed in consumer goods manufacturing. There, the e-commerce company Alibaba (along with Pinduoduo and JD.com) is connecting consumer goods manufacturers and e-commerce retailers via a platform (Author). DATA AND DIGITAl PlATFoRMS IN INDUSTRY \ 24