Large and mature organisations, with their access to knowledge, capital and customers, are perfectly positioned to walk the road from invention to innovation; to turn promising breakthrough technologies and creative concepts into profitable and scalable business opportunities. However, these organisations rarely generate winds of creative destruction and instead start-ups disrupt them at an increasing pace (Anthony et al., 2018; Elsbach & Stigliani, 2018). Large and mature organisations struggle to innovate sustainably, in part because of their rigid organisational structures and processes that maintain the status quo (O’Reilly & Binns, 2019). To overcome this, organisations increasingly deploy ‘innovation hubs’. Innovation hubs are partially independent physical and managerial spaces intended as safe havens for exploratory activities. Examples of hubs are Xerox's’ PARC and Google X ‘the Moonshot Factory’. These are spaces where innovators find freedom to challenge the status quo and where there is space to consider alternatives, to experiment and to learn. Innovation hubs fuel the discussion of “what might be”.However, if organisations want to transform their business, they need to go beyond generating thought-provoking concepts. They need to implement promising concepts and integrate them with the rest of the organisation. Scholars call this gap that exists between concept generation and implementation the ‘Valley of Death’ (from heron: VoD) (Markham et al., 2010). It is crucial that organisations resolve issues related to the VoD if they want to reap the benefits of innovation. However, innovation implementation is a relatively under-examined field (Baer, 2012).Innovation implementations scholars predominantly focus on the proposed concepts. Questions arise, such as are the ideas ‘good’ enough? Are they ‘radical’? Do they serve an actual need? Alternatively, the innovator becomes the focal point of the study. There are stories (in both popular and academic writing) in which one well connected, head strong champion heroically shepherds an innovative concept into realisation, in resistance to incumbent forces. But it is risky for organisations to bet their future survival on the presence, capabilities and ultimately, success of lone champions who succeed despite organisational circumstances, not because of them (Dougherty & Hardy, 1996). Especially since failing to implement innovations often stems from factors beyond the control of champions8(Goepel et al., 2012). Thus, in this research, I take an approach to explore what organisational conditions help innovators to mitigate the VoD and achieve implementation.As a designer, I particularly focus on the relationship between design practices in innovation and the VoD. The Design Council states that design practices can mitigate the VoD (Kolarz et al., 2015). Others suggest they may actually aggravate the issue (Carlgren et al., 2016a). Recently, scholars have noted that designers need to consider implementation issues if they want to contribute to resolving organisational and society-level challenges (Dorst, 2019b; Norman & Stappers, 2015). In this thesis, I consider different conceptualisations of design in an innovation context (as problem solving and as inquiry) and shed light on the role of design in mitigating the VoD.Research DesignI performed this study using an action research approach (Reason & Bradbury, 2008a) in collaboration with a large heritage airline ‘FlyCo’ (kept anonymous for privacy reasons). FlyCo finds itself in a competitive landscape. Weighted down by large labour forces, considerable and long-term capital investments, and legacy management structures, FlyCo faces a battle to remain airborne while competing with both low-cost entrants (e.g., EasyJet) and high-quality ‘Gulf’ behemoths (e.g., Emirates). It operates in (for safety and security), a highly regulated and increasingly commoditised industry, which makes achieving innovation difficult yet rewarding. In response, FlyCo started an ambitious ‘architectural transformation’ (Safrudin et al., 2014) in which ‘design thinking’ was a central pillar to deliver a more customer-centred and cost-efficient service. This transformation required that FlyCo adjust its organisation to implement innovation projects more effectively. This situation provided a solid launching pad for this study. The research objectives, combined with the needs of FlyCo, informed the following main research question:How can design catalyse innovation implementation at a service organisation?Over a 14-month period, I embedded as an action researcher at ‘FlyCo’. I engaged employees from different levels of FlyCo to conduct actions as part of reflective, collaborative research cycles. The research contained three action research cycles (ARCs). Each ARC was performed in collaboration9with a distinct set of stakeholders and with different research aims. In the first ARC, my efforts focussed on building a network and an understanding of FlyCo and the VoD phenomenon. In ARC 2, the focus moved towards investigating conditions that contribute to a VoD with a focus on the role of design practices. In ARC 3, the focus again shifted towards how design interventions in organisational context could contribute to implementation success. Over the research period, I became increasingly immersed in FlyCo as my role shifted from being an outsider to obtaining increasingly influential positions (I became an interim manager in ARC 3 for example), which provided an opportunity to gather a rich dataset.During the embedded period, I employed multiple data gathering methods. Predominantly, I took part in- and observed corporate activities, resulting in 231 temporal observations (events). I captured observations and reflections in field notes, resulting in 426 pages of notes and drawings. Additionally, I gathered internal documents (such as strategies, project proposals, training manuals). Finally, 48 interviews were conducted at multiple intervals during the study. Of these interviews, 17 were semi-structured, audio-recorded, and transcribed, whilst 31 were conversational and recorded via hand-written notes. I initially analysed the data using a visual mapping strategy. Subsequently, a thematic analysis was performed using NVivo software. A breakdown between identified themes and existing literature finally informed a narrative analysis strategy. Together, this data collection and analysis strategy helped to observe nuances in FlyCo's innovation and implementation processes that can evade detection by other ‘outside-in’ research designs.InsightsThe data inform four sets of insights. Extant research on innovation implementation has focussed on product/manufacturing organisations (with historically large R&D departments) that aim to reach additional customers through new/improved products. In this context, managers and scholars noticed that R&D output did not reach controlled stage-gate New Product Development (NPD) processes. But innovation hubs are also increasingly popular at service organisations (Blindenbach-Driessen & Van Den Ende, 2014), which have different (and less structured) innovation processes. The first set of insights describes an exploration and re-conceptualisation of the VoD phenomenon in a service organisation context. I identify three10organisational unit types that contribute to innovation: exploration hubs, support partners and operational units. In this context, the metaphor of a singular ‘valley’ between two contributing units appears erroneous, as implementation challenges exceed the dichotomous relationship between design and production.A deeper investigation into the mechanism that drives the VoD shaped the second set of insights, which highlights the role of institutions, specifically organisational logics. At FlyCo, a constellation of three organisational logics and the absence of a recombination strategy fosters an environment inhibiting resource pooling between organisational units. The three logics inform conflicts on three issues: innovation priorities, innovation processes and problem frames. As logics guide legitimacy judgement, conflicts between logics lead to a Not-Invented-Here attitude from gatekeepers towards concepts from ‘foreign’ logics. Consequently, champions can’t gather the resources needed for implementation and their concepts end in a VoD.The third set of insights describes how 10 barriers contribute to the VoD. I identify four barriers related to organisation properties of FlyCo. A complex and siloed organisation, the absence of a shared service vision, decentralised innovation portfolio management, and a competing internal innovation marketplace stimulate a VoD. Two barriers describe project characteristics related to the VoD: founding problem frames in an inferior domain and proposed solutions with a weak fit with the existing service system. Two process-related barriers highlight how engaging stakeholders late in the innovation process and inadequate communication of project decisions contributes to a VoD. Finally, two barriers describe how the organisational set-up of an exploration hub contributes to a VoD: when there is no ‘Shadow of the Future’ and when hubs have limited access to resources, they struggle to mitigate the VoD.The fourth set of insights explores the relation of design practices with innovation implementation. When viewed as a problem-solving approach, I exhibit how design practices contribute to mitigating implementation issues by fostering more holistic concepts and an innovation process with engaged and aligned stakeholders. However, as an inquiry process, design practices contribute to a VoD when projects are reframed such that the aspired value shifts. A VoD then appears in two situations: if the new working principle requires new stakeholders (not part of the founding problem frame) to become involved, or if not, all involved stakeholders accept the new frame. In11addition, I deployed design practices to create new organisational infrastructure which fosters innovation implementation success. These practices inform a sense of shared ownership and novel organisation designs, but they also introduce challenges that require further investigation.Contributions and GuidelinesOne principal contribution to literature is the reconceptualisation of service innovation implementation. Instead of three sequential phases, ‘elaboration’, ‘championing’ and ‘production’ (Perry-Smith & Mannucci, 2017) are three reiterating micro-processes. These micro-processes constitute two innovation-to-implementation process streams. In one process stream, innovation teams solve ‘innovation challenges’ (Dougherty & Hardy, 1996) through concept elaboration and production. In the other stream, championing in the organisation sphere aims to solve ‘innovation-to-organisation challenges’ (Dougherty & Hardy, 1996). In line with this conceptualisation, I propose to define the VoD in this context as ‘when concept development terminates because champions fail to gather the required resources for further development because of innovation-to-organisation challenges’.Second, I propose a classification of three types of organisational units involved in innovation. In service organisations, achieving innovation requires mitigating gaps between (1) explorative units, (2) support resources, and (3) operational units. I challenge whether the dichotomous conceptualisation of a VoD does justice to the complexities of achieving alignment for reform within service organisations.The findings add to a growing body of knowledge that considers the role of institutions in realising (service) innovation. I add that, besides on an ecosystem level, organisational level ‘Logics matter when coordinating resources’ (Edvardsson et al., 2014) in service innovation. I identify three issues where misalignment between organisational logics hampers innovation implementation: innovation priorities, innovation processes and problem frames. I propose that besides contextual, spatial, and organisational boundaries (Antons & Piller, 2015), organisational logic boundaries can trigger a Not Invented Here attitude.Insights from this study suggest a complicated relationship between design innovation and the successful implementation of these innovations, which I call the ‘Design Implementation Paradox’. Design principles and practices related to experimentation, experiential learning, and embracing12diversity contribute to implementation success. Practices related to embracing diversity, user-centricity and materialisation contribute to resolving innovation-to-organisation challenges and mitigating logic conflicts, and thus to implementation success. However, design can also contribute to a VoD when reframing leads to a shift in the stakeholder field or when champions cannot convince involved stakeholders of a new frame. This study represents an initial exploration into this relation, but more research is needed.The final contribution to theory is 10 organisational barriers identified that contribute to the VoD in a service organisation. For example, by exhibiting how an internal innovation ‘marketplace’ encourages competing behaviour as opposed to collaborative behaviour, which hinders innovation implementation.The insights inform six guidelines for managers, specifically for those who shape organisational conditions, to design organisational infrastructure that promotes innovation implementation. These guidelines describe organisational infrastructure that contributes to mitigating the VoD:1. To resolve innovation-to-organisation problems, service organisations can use innovation hubs because this infrastructure facilitates the required social dynamics.2. To avoid a Not Invented Here attitude, the infrastructure of these innovation hubs can promote institutionalisation and legitimisation of innovation concepts.3. To motivate aligned innovation processes and ‘implementable’ concepts, the infrastructure of these hubs must act as a ‘shadow of the future’.14. To align decisions making across organisational units, a service vision - which describes what value the organisation wants to create in the future - should be formulated and shared.5. To ensure alignment between resource allocation and the innovation vision, and to spot potential VoD issues, centralised innovation portfolio management can be applied.6. To align the innovation portfolio with the current technological and organisational system, the service system-fit framework can be applied.1 An example of such infrastructure is when incentives of innovation hubs relate to implemented innovations, not merely proposed concepts.13This research emphasises the need to study implementation in design research, if designers aim to realise societal impact. Design education needs to adjust to fit the more strategic role that design is assuming. If design is indeed going ‘beyond design’ (Dorst, 2019b) to contribute to solving the world challenges, then we need to go beyond teaching future designers how to generate innovative interfaces, products, and systems. We need to teach them how to contribute to implementation and, ultimately, impact. This implies assuming a broader understanding of design, offering students tools and skills to become more sensitive to organisational context and helping them understand what influences implementation and what strategies they may pursue to achieve implementation. This requires a realisation that the road to implementation is paved with team players and that besides being great pitchers, designers need to learn how to knock the ball out of the park.Above all, this research emphasises the limits of the ‘rogue innovator’ narrative and provides principles for organisational leaders of service organisations that face transformation to mitigate their dependence on innovation champions and instead design organisational infrastructure that facilitates innovation implementation.