• No products in the cart.


What if artificial intelligence (AI) itself were used to investigate the current

literature on AI in marketing? That is what we do in this study!

Having received more than US$5 billion in venture capital investments in just the

past two years, artificial intelligence (AI) is poised to exert transformative effects on

markets and marketing around the world (PricewaterhouseCoopers, 2017;

Rangaswamy et al., 2020; Smart Insights, 2018). Marketing increasingly relies on its

algorithms, which mimic human cognitive functions and exhibit aspects of human

intelligence (Huang & Rust, 2018; Rangaswamy et al., 2020; Russell & Norvig, 2016;

Sterne, 2017), such that 72% of marketers cite AI as a business advantage. Consumers

benefit from these applications, in the form of decreased costs, more diverse service

channels, innovative breakthroughs, and opportunities for expanded human creativity

and ingenuity when tedious, repetitive tasks are performed by AI (Haenlein & Kaplan,

2019; PricewaterhouseCoopers, 2017; Smart Insights, 2018). This revolution of AI usage

in marketing, and its potential for producing superior value outcomes, has sparked

substantial research attention (Davenport et al.; 2020; Haenlein & Kaplan, 2019; Huang

& Rust, 2018), prompting, for example, applications of intelligent technology (Marinova

et al., 2017); descriptions of services enabled, facilitated, and delivered by various

technologies (Rust & Huang, 2012); investigations of AI-powered robotics (Lu et al.,

2020; Wirtz et al., 2018); explorations of AI-led marketing and sales strategies

(Davenport et al., 2020); considerations of how AI-enabled delivery can lead to cost

effective service excellence (Wirtz, 2020; Wirtz & Zeithaml, 2018); proposals of AI

enabled platform business models (Wirtz et al., 2019); investigations of the impact of AI

chatbot disclosures on customer purchases (Luo et al., 2019); considerations of effects

on workforces (Davenport & Kirby, 2015) and redefinitions of AI-enabled workplaces

(Chui et al., 2015); and discussions of digital technologies as driving forces of work and

life (McAfee & Brynjolfsson, 2016).

Despite this extensive list, marketing still lacks a cohesive understanding of how

AI technologies have been applied thus far and how they should be in the future

(Haenlein & Kaplan, 2019; Paschen et al., 2019). That is, it needs literature analyses to

scrutinize and synthesize the use of AI in marketing and pave a concrete path for

future-focused academic research. Objective, reflective analyses are crucial to

evaluate any extant knowledge base, identify knowledge gaps, and evaluate research

effectiveness and productivity (Huang & Rust, 2018; Russell & Norvig, 2016), for both

researchers and publication outlets or journals (Lowry et al., 2004, 2013). Indeed,

where and how scholars publish represent essential aspects of the identity of the

marketing discipline, reflecting its value systems, aspirations, paradigms, reward

systems, cultural conduct, and political hierarchy (Garfield, 2006; Lowry et al., 2004,

2013). Therefore, with this study, we investigate dominant research topics related to AI

in marketing, outlining key themes, influential publications, and networks among

authors and journals, to provide a clear view of the extant knowledge base.3

We apply two complementary analytical approaches to examine the evolution

and structure of the research field. First, to identify dominant, salient topics, we

undertake topic modeling and apply natural language processing, machine learning,

and statistical algorithms. Second, using scientometric techniques, we generate further

insights about authors and research networks, as well as establishing keyword co

occurrence rates, identifying landmark publications, and depicting the evolution of the

field over time (Lowry et al., 2013; Nikolenko et al., 2017; Zhao et al., 2019).

In turn, this study makes four contributions. First, the combination of advanced

topic modeling algorithms, established through text analytics techniques, and of

scientometric analysis, enables a highly objective, robust, structured, and

comprehensive review of this rapidly expanding research domain than traditional

literature reviews and analyses can provide (Vanhala et al., 2020). Second, our

systematic, data-driven approach reveals 10 dominant AI research topics in marketing,

which we can classify as consumer research or organization and strategy–related

research. Third, the scientometric analysis produces information maps of co-citation

clusters, landmark articles, conceptual and theoretical foundations, and reciprocal

interconnections of concepts on the basis of their paired presence in extant literature.

Such a comprehensive approach establishes a broad understanding and in-depth

insights, without restricting the analysis to any predesigned aspects. Fourth, we offer

an extensive discussion of research gaps that in turn produces a robust agenda for

increasing the depth and breadth of AI research in marketing.

After we establish the conceptual underpinning of this study in Section 2, we

explicate the methodological details in Section 3. Section 4 contains descriptive details

of the reviewed publications. In Section 5, we review the salient topics of research on

AI in marketing, and then in Section 6, we report the scientometric analysis. Finally, we

discuss the results and conclude with concrete recommendations for examining

multiple overlooked facets of AI in marketing in Section 7.

Template Design © VibeThemes. All rights reserved.