Abstract
Chatbots (sometimes just called “bots”) are the subject of much corporate and public interest today. Many enterprises are looking to get started with chatbot development initiatives to improve communication efficiency as well as reduce operating costs. Current research indicates constantly growing interest in this area and forecasts that 70% of office employees will interact with chatbots daily in 2022. This chapter reports on the challenges inherent in chatbot integration projects and identifies key operational factors for successful chatbot projects, as well as highlighting issues of strategic significance. Different technology adoption and project management models are explored, analysed, and applied in the context of chatbot implementation, and based on an in-depth case study, a model is put forward to aid the manageability of chatbot implementation in other similar environments.
TopIntroduction
There is a growing belief that chatbot technology can improve organizations in a variety of ways (Goasduff, 2019). Chatbots leverage artificial intelligence (AI) and can automate different business functions and repetitive tasks. Higher employee productivity, personal attention to customers, and effective communication between employees are mentioned as the most anticipated benefits that conversational bots can deliver. Furthermore, successful chatbot implementation cases demonstrate a dramatic reduction in per-query cost, considerable improvement in response times, ability to offer 24x7 customer service, and significant increases in customer satisfaction (Srinivasan et al., 2018).
Despite the undoubted benefits of chatbots, there are relatively few success stories. Companies tend not rush to adopt new technology, and prefer to observe and learn from others. In most cases, successfully launched chatbots automate only a particular business process or activity. Also, bots mostly supplement pre-existing reliable processes, hitherto supported by other information systems. Leading technology corporations, such as IBM, Google, Amazon, and Microsoft, provide cloud services that allow technical teams to deploy AI-powered conversational bots relatively quickly. However, available online services and open source software cannot solve company-specific problems that require implementation of custom bot behavior, data preparation and integration (Gwendal et al., 2020). Another difficulty is the need to change the existing working processes and employees’ behavior. A lack of experienced specialists, high deployment costs, and overall transformation complexity are the key factors that are currently inhibiting many companies from embarking upon their chatbot initiatives (Srinivasan et al., 2018). In this chapter, however, research objectives (ROs) concern identifying and discussing how companies can move forward with chatbot projects. The objectives are, first, to identify key factors for successful chatbot projects (RO1); and second, to design a new framework or model for chatbot implementation in the corporate environment (RO2).
Following this introduction, there are six further sections. The next section provides a context for the research upon which this chapter is based, examining published material on a number of related themes. The selected research method is then discussed and detail of the EVRAZ case study is provided. Then, building upon concepts and success factors evident in the extant literature and in existing models, a provisional model for chatbot project implementation is put forward. This is then applied and developed in the following section based on the EVRAZ case study findings. Finally, the conclusion summarizes the key themes of the chapter, and discusses relevant theoretical perspectives in the context of the main research findings.
Key Terms in this Chapter
Natural Language Processing (NLP): An area of science which explores computer and human interactions, and develops solutions which enable computers to process information in a human’s language.
Agile: A family of project management methodologies that aim to solve bureaucracy-related problems in software development and achieve high productivity. Unlike the traditional sequential and formalised approach, agile development is iterative, relies on communication between individuals and frequent product releases.
Scrum: A software development and project management framework based on agile principles. It employs an iterative approach, breaks the full workload into small parts called “sprints”, which can be easily assessed by the project team and completed in a short period (usually 2 weeks). Each sprint ends with the release of a new feature presented to customers. Scum consists of a set of formal procedures and rules that determine how software developers should manage their planning meetings, formulate product requirements, estimate work, communicate within the team and with the customers. By following these rules teams can achieve hyper-productivity by dramatically reducing time spent on communication and documentation, involving customers in the process, and reviewing plans after each iteration.
Unified Theory of Acceptance and Use of Technology (UTAUT): A technology adoption framework employing elements of other popular IS adoption models. The model views technology acceptance through user perception. It is widely applied by researchers to study technology adoption in practice by exploring behaviour of individual users.
Net Promoter Score (NPS): A marketing metric to measure customer loyalty. The measurement is based on asking the customer only one question “How likely is it that you would recommend X to a friend or colleague?”, where X represents the product, service, or company, which NPS is measuring.
Five Factor Model (FFM): A psychology model, which describes the personality of people and measures individual aspects of their behaviour. It comprises five factors: openness to experience, conscientiousness, extraversion, agreeableness, neuroticism.
Machine Learning (ML): A set of activities which leverage modern computer algorithms to process data and solve vast numbers of applicable or scientific tasks.
ChatBot: A software program which can automate certain routines by interacting with humans in their natural language.
Task Technology Fit (TTF) Model: A model that evaluates the impact of technology in supporting individuals in performing specific tasks.
Technology Organization Environment (TOE) Framework: A technology adoption and IS strategy development model. It is based on three groups of factors which influence the technological innovation decision making process: Technology, Organisation, Environment.