Integrating AI into marketing and advertising research involves leveraging machine learning algorithms, data analytics, and computational tools to gain insights into consumption, consumer studies, and consumer research including consumer behavior, how to optimize marketing campaigns, and enhance the personalization of advertising content. Brand et al. (2023) for instance, explore the potential of Large Language Models (LLMs) in market research applications, focusing on GPT3.5 ability to mimic consumer preferences and behavior. Reisenbichler et al. (2023) explore the integration of LLMs in creating advertising content for search engines, highlighting the development of an application layer over OpenAI’s GPT models to generate ad text tailored for search engine advertising and predict advertising costs. Similarly, Spatharioti et al. (2023) explore the impact of LLMs on consumer product research, particularly in the context of search engine use. Hayes et al. (2021) compared graduate students’ sentiment analysis of social media posts on Nike’s “Dream Crazy” ad featuring Colin Kaepernick to that of two major tools, Crimson Hexagon and LIWC, finding that these tools’ accuracy in categorizing sentiments as positive, neutral, or negative was only about one-third, equivalent to random chance.
Four academic literature reviews on the topic of the increasing relevance of AI in marketing are noteworthy. We present each in brief. A more extensive review can be found here.
Vlačić et al. (2021) review 164 articles, identifying AI’s impact on marketing and advertising research across four overarching themes:
- Marketing Channels: AI enhances efficiency in marketing channels (Bock et al., 2020; Wirtz et al., 2018), with applications in predicting customer preferences and improving product distribution, as seen with North Face and Amazon (Sjödin et al., 2018).
- Marketing Strategy: AI reshapes strategic marketing, balancing massification and customization (Du et al., 2003), and integrating luxury brands with mass markets (Kumar et al., 2020c; Paul, 2019).
- Performance: AI’s predictive capabilities are superior for performance evaluation (Bock et al., 2020; Russell & Norvig, 2016; Syam & Sharma, 2018) and impact competitive advantage and customer value creation (Paschen et al., 2020; Riikkinen et al., 2018; Diaz, 2017).
- Segmentation, Targeting, and Positioning (STP): Advances in STP utilize AI for customer base management, leveraging neural networks for market segmentation (Fish et al., 1995; 2004; Ha et al., 2005) and customer profiling (Lei & Moon, 2015; Wu et al. 2015; Belk, 2016; Pitt et al., 2018).
Mustak et al. (2021) provide an in-depth analysis of AI in marketing, identifying ten dominant research topics divided into consumer and organization/strategy-related themes. Consumer research includes AI in sentiment analysis (Chong et al., 2016; Zhang et al., 2018), customer satisfaction (Ansari & Riasi, 2016; Baumann et al., 2012), eWOM (Pantano et al., 2019), brand management (Haryanto et al., 2015), customer loyalty (Ballestar et al., 2019), and relationship management (Bejou et al., 1996). Organizational themes cover AI in B2B automation (Cascio et al., 2010), market performance (Erevelles et al., 2016), service innovation (Yu, 2020; Van et al., 2019), and strategic marketing (Lin & Kunnathur, 2019; Netzer et al., 2012).
Mariani et al. (2021) conducted a systematic review of AI research across marketing, consumer research, and psychology, identifying eight key thematic areas in AI research: (1) memory and computational logic; (2) decision making and cognitive processes; (3) neural networks; (4) machine learning and linguistic analysis; (5) social media and text mining; (6) social media content analytics; (7) technology acceptance and adoption; and (8) big data and robots. This work showcases the diverse and expanding scope of AI’s application in these fields.
Notably, all three literature reviews demonstrate a rapid and exponential growth in the evolution of academic publications on the topic of AI in marketing over time, with Vlačić et al. (2021) pointing to a significant spike in 2017 onwards.
Finally, Haleem et al. (2022) analyzed 217 academic publications, highlighting AI’s diverse applications in marketing segments like personalized marketing, predictive analytics, customer segmentation, automated decision-making, and customer engagement.
In industry research, AI in market research covers a range of topics. From how AI will impact jobs (Qualtrics 2018 surveys industry perspectives) through AI’s potential to both enhance and disrupt marketing and sales (Deveau et al., 2023; see also Ruden, 2023 on how AI can optimize paid search campaigns) and to practical guides designed to help those already working with AI or looking to leverage it in their businesses (IAB AI Standards Working Group, 2021).
Ostler & Kalidas (2023) emphasize the potential of Large Language Models for efficiency and data analysis, predicting the use of LLMs in three primary use cases. Firstly, to enhance efficiency by automating tasks, such as the manual coding of open-ended responses. Secondly, to improve capabilities in areas like analyzing vast amounts of data. Lastly, LLMs create new opportunities, such as generating multiple versions of a concept and using another AI system to evaluate each, thereby identifying the most effective option. This mirrors Analytics Partners (2023)’s view on Generative AI’s capabilities and the need for substantial human involvement, oversight, and review of AI outputs, without which AI cannot effectively develop reliable and precise commercial analytics models.
According to Manole (2023) the market research industry is witnessing a transformative shift with the integration of AI. First, AI-Powered Market Research for Price Optimization is revolutionizing how businesses determine pricing strategies by meticulously analyzing market data to balance value and profitability. Simultaneously, Emotion AI is being employed in Customer Experience (CX) Market Research, enabling companies to identify and enhance key customer journey touchpoints through emotional and feedback analysis. The Rise of AI-Powered Survey Technology is another significant trend, where AI is streamlining the processing of free-text survey responses, transforming them into comprehensive and actionable insights. Additionally, Inclusivity through AI Models is gaining momentum, focusing on understanding and catering to a diverse range of customer demographics, thereby personalizing business strategies. Lastly, AI’s application in Social Media Market Research is proving invaluable for analyzing user behavior and engagement, as well as predicting content performance, thereby refining targeted marketing strategies.
Moriarty (2023) identifies three prominent ways AI can enhance marketing: First, by analyzing digital trends and consumer sentiments to inform market research and strategy development. Second, through predictive analytics for efficient and insightful creative testing (see also Ho et al., 2022 discussing how machine learning (ML) can predict new product 4innovation success). Finally, AI optimizes campaign performance by providing real-time insights, enabling rapid adjustments and improvements.
Huff & Bonde (2022) discuss how the emerging field of AI-enabled consumer intelligence (AICI) is reshaping how businesses understand and interact with consumers. Since these platforms are designed to enable enterprises to gather insights from various data sources, both external (like social media, web data, and consumer data) and internal (such as CRM and website data), they can help brands grapple with the volume of consumer-generated data and uncover predictive and prescriptive insights.
Deveau et al. (2023) highlight how gen AI is revolutionizing marketing and sales, especially in B2B and B2C sectors, emphasizing the potential of gen AI in areas like customer experience, productivity, and growth. They discuss the automation of a significant portion of sales functions, the importance of advanced sales technology and hyper-personalization, and the impact of AI on various aspects of the customer journey. In contrast, Kerwin (2023) reflects on the cautious or wary approach with which marketers address Gen AI, underlining the importance of addressing crucial aspects including formulating a clear strategy, ensuring adequate training for usage, allocating the necessary budget, and understanding the legal implications involved before incorporating the outputs of AI tools. Relatedly, Hardcastle (2023) discusses the unintended consequences of Gen AI content in marketing, including the inadvertent production of harmful content, the impact of inconsistent brand tone of voice, the environmental impact of generating absurd quantities of content, and how an overreliance on Gen AI tools can stifle creativity.
Alongside the possible benefits of integrating AI into marketing, there are also voices calling to consider the inherent challenges in AI’s multifaceted influence. For instance, Campbell (2023) critiques the limited scope of existing AI application frameworks in market research citing The Cross-Industry Standard Process for Data Mining (CRISP-DM) as an example. He calls for a voluntary agreement among market research companies and practitioners to guide the adoption of these technologies and ensure they are deployed ethically and responsibly. Cooke & Passingham (2022) discuss the need to shape ethical frameworks to retain the industry’s ability to self-regulate (see more in chapter 5 on ethics).
AI advertising, namely, “brand communication that uses a range of machine functions that learn to carryout tasks with intent to persuade with input by humans, machines, or both” (Rodgers, 2021, 2)2 carries with it much promise (and perhaps also peril). Huh et al. (2023) identify four domains of advertising practice and research that are undergoing profound transformations due to the influence of AI, creating new avenues for exploration, and necessitating further research efforts (see Figure 4):
(1) Consumers’ experience of advertising. AI is revolutionizing advertising by enabling novel human-AI interactions with virtual chatbots, influencers, and brand spokespersons. This transformation prompts inquiries into how consumers engage with AI persuasion agents, distinguish them from humans, and the societal implications of this AI-powered world. Additionally, generative AI technology is reshaping search advertising, altering search behaviors and result presentation. Lastly, the emergence of AI powered virtual influencers raises questions about their impact on social influencer marketing and whether existing advertising theories remain relevant in this evolving landscape.
(2) Societal and policy implications related to the truthfulness of AI-generated content. The rise of generative AI technology raises concerns about the spread of fake information in AI-generated content. Researchers can address this through advertising literacy training and considering regulations for ethical AI-powered advertising.
(3) The analytical considerations of data and algorithms. AI advertising poses ethical concerns regarding data collection, transparency, and bias. Research is needed on evolving data privacy issues, covert data capture’s implications, and consumer awareness. Opacity in advanced AI models hampers transparency, affecting ethical advertising. Algorithmic bias in data-driven advertising and its societal impact also require examination.
(4) The functioning of the advertising industry. AI’s impact on the economy, especially job displacement, is a long-debated issue. While some jobs have been replaced by automation and AI, the advertising industry has been relatively insulated. However, the rise of generative AI raises questions about the future of creativity in advertising and which tasks will still require human intervention, impacting revenue models, organizational structures, and client-agency relationships. As an example, consider Fond et al. (2021) use of deep learning models to improve emotion-based advertising in digital media.
Ford et al. (2023) conducted a comprehensive review aiming to map the field’s evolution. Analyzing AI advertising academic articles published between 1990 and 2022, they identify a publication trend wherein despite the infancy of AI advertising research, since 2018 there is enormous growth in academic articles related to AI advertising (see Figure 5). Analyzing 72 articles they identify four clusters as key focus areas of AI advertising research: AI-driven advertising innovation, Computational advertising (CA), Programmatic advertising (PA), Ad Effectiveness in AI Advertising.
Cluster 1: AI-driven advertising innovation features 18 articles. More than half are conceptual, discussing AI’s role in advertising and suggesting future research avenues (Kietzmann et al., 2018; Li, 2019; Coffin, 2022). These works, along with others, examine AI applications in various areas. For instance, Qin & Jiang (2019) focus on smart advertising, namely how AI is used in various stages in the advertising process such as consumer insight discovery, ad creation, media planning and buying and ad impact evaluation. Campbell et al. (2022) examine manipulated advertising, specifically how consumers respond to ‘synthetic ads’ that rely on deepfakes and generative adversarial networks (GANs). Bakpayev et al. (2020) focus on programmatic creative to show how AI ads are effective in rational appeal, yet they lack in emotional aspects compared to human-made ads. Deng et al. (2019) and Vakratas & Wang (2020) explore enhancing ad creativity through smart ads.
Cluster 2: Computational advertising features 19 articles. Articles in this cluster focus on how AI techniques like predictive modeling, programmatic advertising, algorithms, and machine learning are leveraged to enhance advertising efficiency, and explore how these technologies use data for tailored ads and targeting. For instance, Malthouse et al. (2019) merge programmatic advertising with recommender systems, creating content delivery algorithms. Similar algorithms for consumer profiling are discussed by Neumann et al. (2019), with applications in traditional and mobile advertising (Li & Du, 2012; Guitart et al., 2021). Chen et al. (2019) advance programmatic creative for automated ad creation, while Liu-Thompkins et al. (2020) and van Noort et al. (2020) focus on automating brand content. Araujo et al. (2020) examine brand-consumer experiences in this context. Zimand-Sheiner & Earon (2019) investigate the impacts of computational techniques on account planning through a qualitative study with managers.
Cluster 3: Programmatic advertising features 14 articles focusing on the increasingly popular programmatic advertising within AI-driven computational advertising. This automation of media buying enhances ad placement efficiency and performance. Samuel et al. (2021) explore its processes and consumer concerns related to personalization. Palos-Sanchez et al. (2019) address privacy concerns, finding no significant increase over time. Other research includes methodology development for search and display advertising optimization (Gong et al., 2017; Miralles-Pechuán et al., 2018), emphasizing ad relevance (Kononova et al., 2020) and consumer motivations (Lee & Cho, 2020). This cluster primarily examines website advertising in the context of programmatic advertising’s impact and consumer response.
Cluster 4: Ad effectiveness in AI advertising features four articles. Shumanov et al. (2022) conduct a mixed-method study to evaluate the effectiveness of AI-predicted personality-based ads, showing that for most personality types matching consumer personality with congruent advertising can lead to more effective consumer persuasion. Matz et al. (2019) assess image appeal in advertising computationally. Huh et al. (2020) explore viral ad diffusion, finding that high source trust among consumers enhances engagement with viral ads. Lastly, Roy et al. (2017) develop a Trust Score Algorithm for social media advertising, demonstrating that both advertiser and sender trust positively influence ad effectiveness.
In industry research, research focuses on how AI presents opportunities and challenges to advertisers and their agencies (e.g., Gralpois, 2023; Hsu & Lu, 2023; Vranica, 2023). Forrester’s “B2C Marketing CMO Pulse Survey” predicts AI will enhance creative agencies, potentially close digital shops, and trigger more reviews in the coming year. The survey anticipates the leading 10 agencies to invest around $50 million collectively in partnerships for developing bespoke AI solutions for clients (Stam, 2023). Erdem & Sidlova (2023) (see also Ostler et al., 2023) tested some fully AI-generated ads along those partially developed by GenAI (like script writing or image creation), using Kantar’s AI-powered ad-testing tool that leverages data from 250,000 real-world ad tests. They show how GenAI ads performed strongly, but quality was variable, thus arguing for the role of AI as partner that helps agencies and marketers explore new ways to be creative.
When delving into how AI is used for advertising research, however, research is scarce. Indeed, two call for papers have been recently issued on this topic: the Journal of Advertising Research, on “How Can Advertisers Leverage Ai And Generative AI?” and the International Journal of Advertising on “Evolution, Challenges, and Opportunities of AI-Generated Advertising.” As a first step to address this lacuna, in what follows, we present findings from 8 case studies, each showcasing a different aspect of advertising research and how AI can be used (with more or less success) to address these.
2 However, definitions of “artificial intelligence and advertising” vary widely, see Rodgers (2021) for elaboration.