Article Preview
TopIntroduction
Major developments in Artificial Intelligence (AI) have resulted in many industries and organisations looking into AI applications in processes, operations, and customer interaction to achieve gains in efficiency and organisational outcomes (Borges et al., 2021). Conversational agents (CAs) are one such application of AI that can mimic human conversations through text or voice-based interfaces. Two types of CAs have been gaining traction in the world of technology: chatbots (text-based) and digital assistants (text or voice-based). The use of chatbots has become prevalent through major messaging platforms such as Facebook, Slack, WeChat, and WhatsApp. Facebook itself has over 10, 000 chatbots that interact with its users (Devaney, 2016). It is estimated that the market size of chatbots will reach $1.25 billion by 2025 (Grand View Research Centre, 2017). In the space of digital assistants, major technology giants such as Amazon, Apple, Google, and Microsoft have invested considerably in the development of digital assistants for consumer applications (e.g., Google Assistant, Microsoft’s Cortana, Amazon’s Alexa, or Apple’s Siri). The digital assistant offerings by these companies have been adopted by millions of consumers (Hao, 2018). Advancements in CA technology have gained the attention of organisations looking to use CAs in areas such as customer experience and internal operations (Meyer von Wolff et al., 2020). Interest for CAs among organisations is high, and Gartner predicts that chatbots, conversation user interfaces and virtual assistants will gain significant interest from customer service and support leaders (Omale, 2020), as well as transform the digital workplace in the next 2 to 10 years (Rimol, 2020).
It is evident that organisations strive to tap into the potential in leveraging CAs for benefits and value creation. CAs can provide customers with enhanced experience via personalised services and build strong relationships with customers (Huang & Rust, 2021). Moreover, CAs can transform organisational operations (Tarafdar et al., 2019). However, research into the applications of CAs in organisations is scant. More attention has been paid to the development, technical, and design aspects of CAs. The varieties of strategic applications of CAs in an organisational setting remain underexplored (Io & Lee, 2017). Early research on the application of CAs in organisations have looked at use cases in the consumer domain (Chung et al., 2018), workplace CAs for supporting employees (Feng & Buxmann, 2020; Meyer von Wolff et al., 2020), and machine learning (ML) techniques for CAs in various business domains (Bavaresco et al., 2020). Notably, different from traditional, task-oriented enterprise systems, CAs consisting of sophisticated cognitive and emotional features present new technical and organisational challenges when implementing CAs (Huang & Rust, 2021; Jang et al., 2021). Failed implementations can be costly, not only preventing benefit realisation but also detrimental to the reputation, as is the case with Microsoft chatbot Tay (Neff, 2016). It is vital for organisations to identify the needs for CAs, present justification evidence, and prepare an implementation plan to address risks ahead (Peffers & Santos, 2013). However, limited literature offers a holistic understanding of why and how organisations adopt CAs to support organisational capabilities - that is, to understand the rationale behind a CA investment and consider the foundations of successful implementation. The lack thereof limits the IS community’s ability to develop refined theoretical understandings of the factors that affect CA’s creation of value and their successful implementation, limiting the community's ability to provide evidence-based advice to practitioners about such matters.