Integrating structural ambidexterity, radical innovation, and business model innovation in the context of large organisations

At present, global and social factors are pushing firms towards radical innovation. Because of their important resource endowments, large firms are expected to provide major solutions to such global problems. However, such organisations are normally proficient at providing significant improvements rather than putting forward radical developments (McDermott and O'Connor, 2002; Blindenbach-Driessen and van den Ende, 2014). The literature on ambidexterity tackles the issue on how to achieve incremental and radical innovations through the implementation of twodifferent types of learning: exploitation and exploration. The first being more useful for incremental innovations refers to the use and development of existing knowledge. The latter being more connected to radical innovations seeks to generate alternative new knowledgequestioning actual standards, procedures and technologies (March, 1991; Jansen et al., 2006)Combining these two distinct types of learning has been found to be of paramount importance for firms' performance in the long term (Raisch and Birkinshaw, 2008). However, when it comes to implementing ambidexterity, large firms struggle when combining exploitation andexploration. These two types of learning are very different in nature and require divergent processes and capabilities (Lubatkin et al., 2006). Previous studies on ambidexterity propose three main ways of achieving it: contextual ambidexterity, sequential ambidexterity and structuralambidexterity (Gibson and Birkinshaw, 2004). Contextual ambidexterity refers to achieving exploration and exploitation at the same time and in the same place; this is a complex process and seems to suit only very creative organisations. Sequential ambidexterity consists offocusing on exploitation during a period and on exploration on a subsequent period and seems to be more adequate for small firms with a limited resource endowment. Structural ambidexterity is mostly implemented by large firms that can afford undertaking exploitative andexplorative learning processes in different places and under different organisational conditions at the same time. Structural ambidexterity has been found to be an excellent solution for large firms because exploration activities do not normally fit well within large and bureaucraticorganisations (Kelley et al., 2009). However, the implementation of structural ambidexterity remains a problematic issue (Donada et al.,2021). In this paper, we examine a thorough case study on how Crédit Agricole, a large commercial bank, set up in 2018 a startup studio named LaFabrique to encourage and give support to exploration projects out of their massive organisational structure. During the last five years La Fabrique has put forward nine startups; eight of them still exist. These startups are consistently based on the search of new business modelsbased on radical fintech innovations that have been proved to be valuable either for Crédit Agricole internal systems or for third firms. Data was collected from June 2018 to early 2023 in the form of non-participant observation and semi-structured interviews. Our presencewas concentrated on the first year and a half (one day per week) during which we had many formal and informal discussions and participated in meetings. We kept a diary to report our observations. In the second period, we updated data through interviews. A total of 23 peopleparticipated in the semi-structured interviews from the beginning, 5 of them several times.In this paper, we put forward three important theoretical contributions. First, we develop further the concept of structural ambidexterity by proposing a startup studio structural ambidexterity model that explains in detail how to implement exploitation and exploration in distinctstructures. The exploration structure would consist of a startup studio that creates a portfolio of startups on a continuous basis, each startup representing a distinctive new business model based on a radical fintech innovation. These radical fintech innovations should therefore beclearly valuable for the mainstream organisation or for a third organisation willing to pay a price for it in the fintech market. Second, we propose a business model innovation portfolio approach to sense, seize and reconfigure (Teece, 2007) structural ambidexterity in largeorganisations. Third, we extend our understanding on radical innovation and business models by integrating both concepts and proposing that radical innovations that create, deliver and capture value could be considered as radical business model innovation. This distinctionemphasises the role of value (creation, delivery and capture) when examining a radical innovation by distinguishing between radical invention (novelty), radical innovation (use) and radical business model innovation (value creation, delivery and capture).