Abstract:

Artificial Intelligence is one of the most promising emerging technologies. This Paper examines the global view of the advancing role of AI in Marketing. The Paper analyses AI from the country, company, and consumer perspectives. It recognizes the existence of economic inequalities across nations and the multiplicity of resources for AI adoption. The country-level analysis emphasizes the economic disparities across countries challenged by lack of economic resources necessary for AI adoption. The company-level analysis focuses on glocalization because while the required hardware that underlies these technologies may be ubiquitous, their application mandates adaptation to local cultures. The consumer-level research focuses on customer privacy and ethics, given the abundance of data stored with AI technologies.

Through these three lenses, the research paper examines two significant dimensions of AI in Marketing: 1) Human-Machine Interaction (HMI), and 2) Automated analysis of Text, Audio, Images, and Video.

Artificial Intelligence incorporates programs, algorithms, systems, and machines that mimic intelligent human behavior. AI technologies predominantly include Machine Learning, Natural Language Processing, and Neural Networks, and enable machines to autonomously sense, comprehend, act, and learn vis human-machine learning interaction (HMI) (Davenport, Guha, Grewal, &Bressgott,2020)2.

The researchers believe it is critical to observe AI technologies via a global lens because 1) The growth of AI technologies and corresponding data collection has created a cycle of consumer data collection that has capacitated AI technologies to become more effective & efficient. 2) At the company level, deploying these technologies across the globe requires localization efforts by firms. 3) AI technologies have the potential to either shrink or widen the gap between developed and developing countries.

Human- Machine Interaction

Refers to the innumerable ways in which humans & machines interact via touch, gestures, voice, and sensors. HMI is enabled by cognitive technologies such as computer vision, machine learning, speech recognition, natural language processing, and robotics. These technologies are increasing their capabilities of performing human tasks and also augmenting human abilities by strengthening their cognitive abilities. In this symbiotic relationship, humans can focus on feeling tasks while machines can aid humans in better decision-making. For instance, Chatbots allow customized, individual, bi-directional communication between humans and machines. Relevant HMI offers a wide range of applications that include the Internet of Things (IoT) and Smart devices (Hoffman & Novak, 2018) virtual and augmented reality (Wedel, Bigne, & Zhang, 2020)4, personalization (Tong, Luo, & Bo, 2020)5, virtual reality experiences (Kang, Shin, & Ponto, 2020)6, and facial recognition (e.g., Amazon’s cashier-less Go stores).

The country-level focus of HMI: Economic Inequalities

AI development, adoption, and usage vary based on a country’s economic resources. Since it is an advanced and expensive technology, most leading companies in this space are based in developed countries. Resultantly, AI may act as a global divider or unifier. AI firms constantly seek a balance between global technologies and local adaptation (Jobin, Lenca, & Vayena, 2019)7. Prior studies have observed HMI- based and algorithmic- driven marketing strategies to obliterate incorporating local needs, customs and nuances particular to a nation in its design.. For instance, during the pandemic, the unavailability of reliable and high-speed internet, sparse network coverage, and single low-end cellphone limited the learning by students in developing countries (Wall Street Journal, November 24, 2020)8.

Country- specific HMI can address economically disadvantaged customers and reduce inequality. For example, HMI-based telehealth services reduce administrative burdens, empower patients with at-home care, determine at-risk patients, and more. E-bay’s HMI translation service has helped lower language barriers and thus been associated with a significant increase in global trade and a reduction in economic inequality across the globe (Meltzer, 2018)9.

The company-level role of HMI: Glocalization

Glocalization is an approach where the forces of globalization are co-shaped with local cultures via strategic adaptation (Thompson & Arsel, 2004)10. Most HMI platforms support English and a few other major languages, which helps globalization via language standardization. However, the research proposes that customization of AI technologies, such as machine learning, and natural language processing, will require adaptation to local culture. Glocalization may also enhance global operations by incorporating heterogeneous preferences and local consumer experiences particular to the culture. For example, Netflix has designed its machine learning-powered algorithms to develop programming adapted to various local consumer tastes (Smith & Telang, 2018)11.

The Consumer-level role of HMI: Ethics & Privacy

AI technologies are often capable of identifying anonymized data and may result in algorithmic bias. The effectiveness of AI technologies in marketing is often based on their ability to collect swamps of individual-level data (Bleier, Goldfarb, & Tucker, 2020; SAS, 2021)12. Today, AI-enabled technologies collect more data than ever and track customer behavior both, offline and online through various mobile and connected devices. For example, Amazon’s granular data collection efforts provide a rich, 360-degree view of offline and online customer shopping behavior (via a partnership with physical retailers). As AI becomes more embedded in everyday devices, it may nudge humans to delegate decision-making that is barely perceptible.

AI may also take over tasks that humans perform. AI and automation may surpass humans in specific tasks such as innovation and design, not only predicting content but also designing and developing this content. Once AI takes over the creative content, humans will be required to ‘‘sensemaking, that is, understanding which problems should or could be addressed”. This will shift the focus of the types of problems the firms will seek to solve. The intensified use of machine labor will vary across countries, and the cost-benefit ratio of machines vs. humans could be considerably more favorable for high-income countries (e.g., Germany) compared to low-income countries (e.g., Senegal) (Frey & Osborne, 2017)13. This might further exacerbate distrust and alienation by humans, leading to speciesism towards artificial intelligence agents and racism against them. Thus, as AI advances, many individuals and organizations will likely advocate for the ethical treatment of artificial Intelligence(MacLennan, 2014)14. Regulation will be evoked to counteract privacy aspects due to HMI. However, the global reach of AI makes consensus on international regulations particularly challenging.

In essence, AI technologies involve two main players: humans and machines. Machines automate and predict, while humans apply their unique insight and use machine-generated predictions to solve marketing-related problems more efficiently and/or more profitably (Ma & Sun, 2020)15.

Automated Analysis of Text/Audio/Images/Video

Automated technologies such as NLP, computer vision, and web scraping have the capability to extract rich insights from unstructured data such as text/ audio/ images and video. In the future, experts predict “autonomous intelligence” may fully automate the collection, processing, and utilization of all these forms of data and will significantly augment human competencies (Davenport, 2018)16.

Automated analysis may offer many benefits but could also lead to data-related concerns. Automating data collection processes (e.g., robots completing survey responses) could lead to data distortion and biased inferences. Likewise, these types of algorithms have also been accused of (automatically) creating misleading content, such as social media trolling or fake reviews (Shao et al., 2018)17.

The country-level role of automated analysis: Economic Inequality

AI-enabled technologies can offer a wider range of services to underserved populations bridging economic inequalities and acting as a unifier between rich & poor.

AI’s automated processing capabilities can help optimize marketing communication across various local (first) languages. AI’s speech recognition and speech-to-text capabilities can help emerging markets circumvent the challenges posed by low literacy and serve previously underserved populations (e.g., Google translate).

However, AI’s autonomous abilities may also lead to a class-based divide between (1) the masses who work for algorithms (i.e., the low-wage gig worker), (2) a smaller privileged professional class who have the skills and capabilities to design and train algorithmic systems, and (3) an elite set of ultra-wealthy technocrats who own the algorithmic platforms that run the world (Walsh, 2020)18. Developing and maintaining AI algorithms is a complicated and expensive process that requires domain expertise. This acts as an entry barrier for a lot of firms. Since only a select set of companies possesses the capabilities, it may perpetuate this divide (e.g., Apple, Amazon, Tesla). In sum, AI’s autonomous abilities may intensify the class-based divide. Inequalities may also rise if some countries need more AI technologies that align with the composition of their industry structure.

The company-level role of automated analysis: Glocalization

AI-based text/audio/ image/ video analytics will likely be more successful if locally applied. Prior research suggests Integrated Marketing Communications will be more successful if it incorporates varied cultures, languages, and socio-demographics (de Villiers, Tipgomut, & Franklin, 2020)19. Thus, it is not surprising that AI-based automated applications such as call agents, self-service terminals, chatbot applications, and voice-based interactions a largely local and exist across all major languages (Dawar, 2018)20.

In general, automated techniques such as text mining, sentiment analysis, emotion detection, and speech recognition are becoming increasingly language-specific and signify a shift toward glocalization. The automated analysis should deftly identify regional differences and adapt to local needs.

The consumer-level role of automated analysis: Ethics & Privacy

The fact that automated analysis requires copious amounts of data to fuel its machine-learning algorithms increases the risk (and costs) and occurrences of data breaches even though it may be not a leading cause for them.

To mitigate such risks, while harvesting the power of automated analysis, firms are becoming increasingly reluctant to store and process individual-level data and, instead, have shifted their focus to meta-data generated through statistical analysis (Wieringa et al., 2021)21. In this approach, firms collect customer data but only store its statistical properties rather than the raw data.

Data privacy could be safeguarded by having the analysis take place on the user side. This approach, known as federated machine learning, was first proposed by Google in 2016 and has received considerable interest among a wide array of AI researchers (e.g., Konecˇny´, McMahan, Ramage, & Richtárik, 2016; Yang, Liu, Chen, & Tong, 2019)22.

Automated analysis has the potential for algorithmic bias. The bias can result in systematic and repeated errors that can lead to unfair and biased outcomes, such as the privilege of one group over other. There are multiple drivers of algorithmic bias, and a prominent one is its tendency to use past data to recommend future actions. This can be corrected by employing hybrid sequential decision-making structures suggesting that humans accept or reject.

The rapid pace of AI adoption presents several opportunities for future research by Marketing scholars. The Paper proposes questions for future research in each of the three levels of the global prism by Marketing research scholars.

Future Research

HMI at the country level has the potential to combat inequalities by optimizing processes, integrating systems, and monitoring resource utilization by offering front-end intuitive, smart, engaging, and culturally relevant interfaces.

Implementation of cognitive technologies can be challenging, particularly in emerging economies. Firms can pool their financial, human, technological, and data resources. The collaborative approach may reduce competitive threats and enable collective solutions.

Understanding cultural differences in AI-related data collection among countries is an essential topic for future research. The paper indicates that researchers are keen to pursue research in this area to preserve customer privacy and maintain ethics to minimize privacy breaches.

Conclusion

Advancements in AI technologies are enhancing the capability of a growing number of firms to collect, store, analyze and utilize a vast variety of customer information (Rust, 2020)23. The Paper concludes that at the country level, AI technologies have the potential to both increases as well as decrease economic inequality. At the firm level, AI technologies have begun to transform various aspects of marketing by glocalizing their applications. At the consumer level, AI technologies raise concerns about ethics and privacy, leading to an increased need for regulation, education, and training.

Looking forward, a growing number of firms will advance toward ‘‘autonomous intelligence” in a glocalized fashion. Once Intelligence is automated, machines, robots, and other forms of AI-based machine learning will be able to integrate with existing information management systems to augment human analytical competencies (Davenport, 2018)24. When this state of autonomy is reached, many firms will be transformed into AI-fuelled organizations that employ human–machine collaboration systems designed to harness and act on data-driven insights locally.

The global prism sheds new light on the country, firm, and consumer-level implications of AI technologies. The forward-looking research agenda helps motivate and direct future research in this critical domain. Thus, while work remains to be done, examining the role of AI technologies in marketing from a global standpoint is worthy of the effort required to provide deeper insights.

KEYWORDS

Artificial Intelligence, Global Marketing, Inequality, Glocalization, Ethics and privacy, Human-machine interaction, Automated analysis of text, Audio, Images, Video.

References:

1. Kopalle,P., Gangwar, M., Kaplan A., Ramchandran D., Reinartz W., Rindfleisch A. Examining artificial intelligence (AI) technologies in marketing via a global lens: Current trends and future research opportunities. International Journal of Research in Marketing 39 (2022) 522-540.

2. Davenport, T., Guha, A., Grewal, D., & Bressgott, T. (2020). How artificial intelligence will change the future of marketing. Journal of the Academy of Marketing Science, 48, 24–42.

3. Hoffman, D. L., & Novak, T. P. (2018). Consumer and object experience in the internet of things: An assemblage theory approach. Journal of Consumer Research, 44(6), 1178–1204.

4. Wedel, M., Bigne, E., & Zhang, J. (2020). Virtual and augmented reality: Advancing research in consumer marketing. International Journal of Research in Marketing, 37(3), 443–465.

5. Tong, S., Luo, X., & Bo, X. (2020). Personalized mobile marketing strategies. Journal of the Academy of Marketing Science, 48, 64–78.

6. Kang, H. J., Shin, J. H., & Ponto, K. (2020). How 3D virtual reality stores can shape consumer purchase decisions: The roles of informativeness and playfulness. Journal of Interactive Marketing, 49, 70–85.

7. Jobin, A., Lenca, M., & Vayena, E. (2019). The global landscape of AI ethics guidleines. Nature Machine Intelligence, 1(9), 389–399.

8. Wall Street Journal, November 24, 2020. Think remote learning is hard? Try using a phone in an Indian village.

9. Meltzer, J. P. (2018). The impact of artificial intelligence on international trade. Brookings, 18 December. https://www.brookings.edu/research/the-impact-of artificial-intelligence-on-international-trade/ Accessed date: 27 September 2020

10. Thompson, C. J., & Arsel, Z. (2004). The starbucks brandscape and consumers’ (anticorporate) experiences of glocalization. Journal of Consumer Research, 31 (3), 631–642.

11. Smith, M.D., & Telang, R. (2018). Data can enhance creative projects—just look at Netflix. Harvard Business Review, January 23 (https://hbr.org/2018/01/data can-enhance-creative-projects-just-look-at-netflix).

12. Bleier, A., Goldfarb, A., & Tucker, C. (2020). Consumer privacy and the future of data-based innovation and marketing. International Journal of Research in Marketing, 37(3), 466–480.

13. Frey, C. B., & Osborne, M. A. (2017). The future of employment: How susceptible are jobs to computerisation? Technological Forecasting & Social Change, 114, 254–280.

14. MacLennan, B. J. (2014). Ethical treatment of robots and the hard problem of robot emotions. International Journal of Synthetic Emotions, 5(1), 9–16.

15. Ma, L., & Sun, B. (2020). Machine learning and AI in marketing – connecting computer power to human insights. International Journal of Research in Marketing. forthcoming.

16. Davenport, T. H. (2018). The AI advantage: How to put the artificial intelligence revolution to work. Cambridge, MA: MIT Press.

17. Shao, C., Ciampaglia, G. L., Varol, O., Yang, K.-C., Flammini, A., & Menczer, F. (2018). The spread of low-credibility content by social bots. Nature, 9, 4787. Shankar, V. (2018). How artificial intelligence (AI) is reshaping retail. Journal of Retailing, 94(4), vi–xi.

18. Walsh, M. (2020). Algorithms are making economic inequality worse. Harvard Business Press.

19. de Villiers, R., Tipgomut, P., & Franklin, D. (2020). International market segmentation across consumption and communication categories: Identity, demographics, and consumer decisions and online habits. Book Chapter from Promotion and Marketing Communications, July 2020.

20. Dawar, N. (2018). Marketing in the age of Alexa. Harvard Business Review, 96(3), 80–86.

21. Wieringa, J., Kannan, P. K., Ma, X., Reutterer, T., Risselada, H., & Skiera, B. (2021). Data analytics in a privacy-concerned world. Journal of Business Research, 122, 915–925.

22. Konecˇny´ , J., McMahan, H. B., Ramage, D., & Richtárik, P. (2016). Federated optimization: Distributed machine learning for on-device intelligence. Computer Science.

23. Rust, R. T. (2020). The future of marketing. International Journal of Research in Marketing, 37(1), 15–26.

24. Davenport, T. H. (2018). The AI advantage: How to put the artificial intelligence revolution to work. Cambridge, MA: MIT Press