Artificial Intelligence (AI) in retail : the AI-enabled value chain

Oosthuizen, Kim (2022-04)

Thesis (PhD)--Stellenbosch University, 2022.

Thesis

ENGLISH SUMMARY: The competitive landscape is shifting for retailers, and many are scrambling to stay ahead of the competition by investing in new technologies like Artificial Intelligence (AI), automation, robotics and blockchain. Traditional retailers face disruption from competitors that can deliver value to their customers faster through these new technologies. AI, in particular, is earmarked to transform retailing, and its influence on retail is projected to be substantial. However, empirical research on AI in retail remains limited. This study investigates how AI is transforming the retail value chain through a qualitative two-stage research design, using four articles to answer the research question: How is AI transforming the retail value chain? The Leavitt Diamond model and the jobs-to-be-done theory are used to answer the research question. First, this study used all the Leavitt Diamond Model variables (i.e. structure, technology, tasks and people) to examine how AI transforms the retail value chain. The process offered a more comprehensive view of the organisational factors that need to be considered when adopting AI in the retail value chain. Previous research typically focuses on only one of these components. Articles one and three broadens our understanding of applying jobs theory and outcomes-based innovation in the context of AI in the retail value chain. In article one, the jobs-to-be-done approach was used as a lens to conceptually cluster the jobs AI technologies can perform in the retail value chain. The article conceptually proposed four AI technology dimensions that can fulfil most of the roles in the “traditional” retail value chain. Article one introduced a conceptual framework to understand AI's role in the retail value chain proposing an alternative AI-enabled value chain. Article two conducted a detailed review of AI's different tasks across the retail value chain. In article three, an outcomes-based approach was used to present a framework of four outcomes for applying AI in the retail value chain and tested the association between the AI outcome and the value chain stage. Therefore, this study proposes the appropriate theoretical lens to understand better the use of AI in the retail value chain. However, it also improves this framework in the final chapter, presenting an adapted conceptual lens. Finally, article four aimed to understand retailers' challenges better when implementing AI, using Leavitt’s Diamond Model. The overall findings suggest that AI transforms the retail value chain in three ways. First, the iterative nature of AI requires the shape of the retail value chain to change from linear to circular, with data and insight at the core of the successful value chain. Second, AI changes how retailers attain goals in the retail value chain through achieving specific outcomes. The outcomes are dependent on where AI is applied in the retail value chain. Third, there is a complex interplay between structure, technology, people and tasks when implementing AI into the retail value chain, transforming how retailers operate. This study broadens the understanding of how new technologies impact value chains in general and retail value chains in particular. For retailers to successfully implement AI into their business, they need a clear understanding of how it impacts people, organisational structure, other technology, and organisational tasks. This study created a framework of eight imperatives retailers need to consider when implementing AI, offering a holistic view of the consideration needed across people, structure, tasks and technology to ensure successful integration of AI into the business.

AFRIKAANSE OPSOMMING: Geen opsomming beskikbaar.

Please refer to this item in SUNScholar by using the following persistent URL: http://hdl.handle.net/10019.1/124902
This item appears in the following collections: