A Research Review on Artificial Intelligence in Marketing

Abstract

Today, AI is playing an increasingly important role in optimizing marketing decisions, improving consumer experience, and establishing and maintaining customer relationship. Although researchers in this field have made certain achievements, due to the short application history of AI in marketing, there are some obstacles for research in this field such as lack of systematic integration of research results and core literature, confusion of related concepts, unclear research focus and future direction. This review focuses on the above deficiencies, helps readers to systematically sort out the knowledge framework of AI research in marketing and to comprehensively understand the current research progress, and hopefully provides reference or inspiration for future interdisciplinary research.Firstly, in order to show the overview of the existing research intuitively, this review analyzes 690 papers from the WOS database, through the CiteSpace knowledge maps. We find that the “marketing + AI” research is mainly based on three aspects: the introduction of basic AI technology, marketing data analysis using AI, and the emerging “AI+service” ideas. Besides, the research in the marketing field can be summarized into four hot topics: service interaction, product design, controversial issues, and data analysis.Secondly, we switch to a micro perspective to sort out the concept of AI. As previous research argues, the definition of AI can be summarized into two categories, namely, a human-centered approach and a rationalist approach, and scholars can focus on different definition perspectives according to what they emphasize. In order to define AI more clearly, this review compares AI concept with AI related algorithms or models like machine learning, and with other technology like IoT, robotics, etc. Researchers in this area should pay attention to these differences.Thirdly, this review further summarizes and comments on previous research in theories, themes and methodologies. The theories involve anthropomorphism, social presence, compensatory consumption, etc. Based on the four hot topics obtained above, this review separately discusses the role of AI in the service, the possibility of AI in product design, the controversy arising from the application of AI, and the significance of AI’s data analysis function for marketing. Furthermore, the coding result of research methods shows that empirical papers dominate now and few researches use actual AI technologies to prove its conclusions.Finally, we discuss future research directions in terms of themes and methodologies. For example, how will consumers’ expectation of the relationship between themselves and AI change over time? What characteristics of consumers and environmental factors can make the humanoid appearance and social skills of service robots produce better results? How to benefit from AI technology without hurting consumer welfare? In each stage of marketing, how to use AI technology to conduct more accurate data analysis? How to further standardize the overall measurement tools and variables in this area? etc.