Artificial intelligence (AI) is considered one of the most strategic technologies for the future of Germany. Last year, Angela Merkel, therefore, announced the launch of the “Future Package, providing an additional two billion euros for AI ecosystems of international appeal, attractive conditions for talent, and state-of-the-art computing infrastructures.” Efficient allocation of this funding, however, requires a clear understanding of the AI application domains in which private and public German actors are active. 

New research at the WHU Innovation Ecosystem Hub provides novel insights into the German AI ecosystem. For his master thesis at the WHU Otto Beisheim School of Management, Lukas Schürmann applied state-of-the-art machine learning approaches (i.e., topic modeling) on a database of nearly 21000 global AI startups to identify 50 different AI application domains. Subsequently, these 50 application domains were used to map the AI activities of both German startups and German research institutes.

Position of Germany on a global level

Relying on data from Crunchbase, the study managed to identify 474 German AI-startups, bringing Germany to 7th place in the global AI startup ranking. Berlin is clearly the most important hotspot for AI startups within Germany. With approximately 200 AI-startups, Berlin is the 9th largest ecosystem globally and the 3rd largest ecosystem in Europe. With around 50 AI-startups, Munich and Hamburg can also be considered as strong AI hotspots within Germany.

Comparing AI activities of German AI-startups and German research institutes

The trained topic model was subsequently used to get a deeper understanding of the application domains on which German AI startups focus. The results point to “Process Automation” as the most popular topic for German AI-startups. 9% of the German AI-startups actively work on this topic.  “Decision Support” (6% of German AI-startups) and “Voice Assistant” (5% of German AI-startups) also emerged as core topics for the German AI startup ecosystem.

The trained topic model was also applied on a database called “The Global InnovationGraph” from MAPEGY GmbH, including information inter alia on more than 2000 public AI-related research projects. In 591 of them, German institutes participated. Analyzing these data demonstrated that the AI topics, which are most popular among German research institutes, are quite different from the topics that dominate the AI activities of German startups. For German research institutes, significant activity is observed on the topics of “Healthtech”, “Human Behavior Robotics”, and “Risk Management”. Whereas 9% of German AI startups work on the topic of Process Automation, this topic is only relevant for 2% of the publicly funded research projects in which German institutes are active.

Based on these observations, Prof. Faems, coordinator of the WHU Innovation Ecosystem Hub and supervisor of the master thesis project, concludes that Germany has strong potential to become a significant global AI player. However, he also emphasizes the need for better alignment between the AI activities of different actors. On the one hand, the analyses point to topics in which German research institutes seem to be very active but have not yet reached the attention focus of German startups. On the other hand, it illuminates domains that are of importance for German startups, which could get more attention and funding at German research institutes.

Fake news is often described as a fundamental disruption that is threatening the foundation of democratic societies. However, we know much less about how fake news is affecting business activities. In his master thesis at the WHU Otto Beisheim School of management, Christoph Busse conducted a systematic analysis of how fake news is affecting the business models of traditional media companies. To do so, he interviewed several media experts and relied on secondary data.

The surprising conclusion of his research is that traditional media organizations do not see fake news as a threat for their business model. Instead, they see it as a huge opportunity. Whereas providing reliable and credible news was historically seen as a necessary condition in the media industry, it is now something that consumers explicitly value. In other words, whereas credible news used to be a commodity, it has now become a valuable asset that allows traditional media companies to differentiate themselves from alternative news providers.

Busse also identified several strategies that traditional media companies use to leverage their reputation as reliable provider of credible news. Whereas media companies were used to focus on trying to be first in generating news, they increasingly focus attention on analyzing and framing the news that has been generated by others. Moreover, in order to sustain their reputation, they heavily invest in (i) improving transparency of news generation, (ii) involving external experts in the analysis of news and (iii) developing new technological tools and platforms to verify news. To do so, traditional media companies increasingly rely on collaborative approaches where they even work together with competitors to create the necessary scale advantages.

In the past decade, media companies have experienced multiple digital disruptions that have fundamentally changed and threatened their business models and even survival. However, this research project concludes that the increasing manifestation of fake news actually implies an important opportunity for established media organizations to fight back and regain market share.

One of the most pressing problems facing modern society is the need to feed the growing world population with increasingly limited resources. Moreover, the recent Corona crisis has highlighted the huge ethical and safety issues regarding traditional mass production of meat in Germany. Cultivated meat, implying the use of animal cells to produce meat, has the potential to mitigate this societal challenge while simultaneously minimizing negative externalities such as global health threats and environmental destruction. In their joint thesis project at the WHU Otto Beisheim School of Management, Gilda Lukacs and Julia Schimanietz conducted an in-depth examination of the entrepreneurial ecosystems for cultivated meat in US and Europe. To do so, they collected both quantitative and qualitative data from cultivated meat startups, investors, and research institutions.

The researchers conclude that Germany has potential to become an entrepreneurial hotspot for cultivated meat. Several German startups (i.e., Planetary Foods, Innocent Meat, Alife Foods, Peace of Meat) are working on cultivated meat products. Moreover, Germany has a strong backbone of life sciences industry giants such as Bayer and Merck and meat industry leaders like Westfleisch. Fostering collaboration between these existing incumbents and the emerging startups could spur the emergence of a strong ecosystem for cultivated meat, which could compete with other ecosystems that are emerging in regions such as California and Israel.

The researchers also point to the crucial importance of NGOs in creating a German hotspot for cultivated meat. Often, NGOs are seen as actors that are hampering technological revolution in the food industry. However, Lukacs and Schimanietz conclude that, in this particular setting, NGOs can function as ecosystem brokers that connect start-ups with universities and venture capitalists. Moreover, NGOs can play an important role in facilitating adoption of cultivated meat by the general public.

Finally, the researchers also reflect on the role of the government to stimulate cultivated meat in Germany. They emphasize that the government should establish a clear regulatory framework for the complete value chain of cultivated meat processes, ranging from the initial cell extraction to the final product, which companies can sell to end-consumers. Such stable regulatory framework does not only reduce uncertainty for startups, but will increase the attractiveness of Germany for a wide variety of clean meat investors.

In sum, Germany has the potential to become a leading player in revolutionizing the meat industry. However, this will only be possible when all involved actors, which often positioned themselves as adversaries in the past, become close collaborators in the future.

In this report, Willem Posthouwer and Dries Faems jointly describe how companies can manage the fuzzy front-end of open innovation. Based on the long-standing managerial experience of the first author at FrieslandCampina, we provide an in-depth description of the model that FrieslandCampina has applied to systematically manage this pre-investment stage of open innovation trajectories. The model is characterized by (i) a triple helix of joint collaboration, business, and development, (ii) which is steered by five relational parameters, (iii) in order to realize a choreographed transition from agreement on action, to agreement on intention, and finally agreement on execution.

New research from the WHU Innovation Ecosystem Hub provides insights into the factors that drive the transfer of knowledge in inter-firm alliances. Prior research has solely looked at the type and quantity of technologies a firm can combine to create new innovations and has ignored the fact that not all firms have the same capability to combine these technologies to create new innovations. In the fuel cell industry, for instance, the fact that a firm has expertise in the domains of hydrogen storage and automotive vehicles does not necessarily imply that it is able to combine these technologies to create hydrogen storage tanks that can function inside a car.

Examining alliances in the fuel cell industry, this research project, developed in collaboration with researchers from WHU and the University of Groningen, focuses on the capability of alliance partners to combine technologies in unique ways to address this research gap. This research shows that when alliances partners have the capability to combine technologies in ways that no other firm in the industry can do, this increases knowledge transfer opportunities within an alliance. After all, when such partners are involved in an alliance, greater learning opportunities are present in it. At the same time, however, these learning opportunities might not always be translated into new innovations: partnering with firms whose combinatory capabilities are excessively unique can lead to lower knowledge transfer rates because of the complexity involved with understanding such a partner.

These findings are important since they indicate that firms should apply more nuanced criteria when evaluating the learning potential of alliance partners. Moreover, they demonstrate that firms that are capable of creating unique combinations between technologies can benefit from this in two ways: (1) they are able to more effectively absorb knowledge from their partners and (2) they are better able to protect their internal resources from other firms. This information is important for alliance managers that seek to maximize the amount of knowledge they can access from their partners, while ensuring that their own knowledge pool remains sufficiently protected.

This summary was first published on https://www.hhs.se/en/about-us/news/hoi-news/2020/unique-technological-combinations-drive-knowledge-transfer-in-inter-firm-alliances/.

WHU study provides new insights into how multinational companies can maximize the technological benefits of collaborating with external partners.

Multinational companies realize that collaboration with external partners is vital for developing new technologies and innovations. These companies therefore increasingly embrace an open innovation strategy, building a portfolio of alliances with other companies. In the pharmaceutical industry, for instance, we can see that companies such as GSK, Johnson & Johnson and Pfizer have initiated multiple alliances with biotech firms to push their pipeline of drug development activities.

Why are some multinational companies more successful in such open innovation strategies than others? This research question was addressed by Prof. Dr. Dries Faems, who currently holds the Chair Entrepreneurship, Innovation and Technological Transformation at the WHU Otto Beisheim School of Management. Together with Brenda Bos, Florian Noseleit (University of Groningen) and Bart Leten (Katholieke Universiteit Leuven), he examined the alliance activities of 2,258 R&D subsidiaries belonging to 118 multinational firms in the pharmaceutical industry.

The results of this study are published in the Journal of Management. In this study, Faems and his co-authors surprisingly find that the internal network structure of multinational companies is an important factor in explaining the success of open innovation. Multinational companies consist of multiple subsidiaries, which each can have alliances with external partners. The study shows that subsidiaries can not only learn valuable knowledge from their own alliance partners, but can also receive important technological insights from the alliance partners of other subsidiaries. Moreover, the study finds that the extent to which subsidiaries can benefit from the alliances of sister subsidiaries depends on characteristics such as headquarters proximity, knowledge overlap, and size of the subsidiary’s own alliance network.

Based on these findings, Faems highlights the need for advanced knowledge management tools within multinational companies to optimize the returns on open innovation investments: “I often see that, within multinationals, subsidiaries have limited understanding of what other subsidiaries are doing in terms of open innovation. Our results highlight that such lack of knowledge is a missed opportunity. When companies are able to implement knowledge systems, providing information on the alliance activities of other R&D subsidiaries, they could foster additional knowledge flows that can help subsidiaries to excel in the generation of new knowledge and innovations.”

Companies increasingly rely on close collaboration with external partners for innovation. For such R&D partnerships to work, trust is highly needed - but often violated and contested. In fact, conflict and distrust can turn a successful partnership into a painful trajectory towards unsuccessful divorce.

 Following a R&D partnership over several years and conducting interviews with the most important stakeholders, dr. Anna Brattström at Lund University, Professor Dries Faems at WHU Otto Beisheim School of Management and Professor Magnus Mähring at the Stockholm School of Economics have conducted an in-depth study of what happens when trust is broken. The study, which was recently published in Organization Studies, reveals novel insights into how to turn around a conflicted R&D partnership and repair trust.

 When firms enter a conflict, it is common that “All-Hands Meetings” are organized, in which all relevant stakeholders of both partners are involved. Findings from this study question this all-in approach. The authors observed that, when all stakeholders were involved in crisis meetings, these meetings actually contributed to further spreading of distrust among people, who initially were not involved in the conflict. At the same time, they observed that explicitly shielding-off employees from the conflict helped stopping distrust from spreading in a relationship and provided a local platform on which trust could be locally repaired. The authors therefore conclude that, when a relational crisis emerges in partnerships, the best option is to quickly confine the conflict and shield-off individuals, who have not yet been personally exposed to the transgression. These shielded individuals can subsequently make a valuable contribution to repairing the broken trust in the conflicted zones of the partnership. In this way, firms can turn a conflicted relationship into a more productive one. 

Brattström, A., Faems, D.; Mahring, M. (2019) From trust convergence to trust divergence: Trust development in conflictual interorganizational relationships. Organization Studies, 40 (11): 1685–1711