1. Introduction

The rapid development of technology, particularly in the field of artificial intelligence (AI), has changed many aspects and meanwhile shifts in human history. Those who take advantage of what AI enables and apply that knowledge to their businesses and careers will have a greater gap than those who don’t [1]. Therefore, this is a chance to be a pioneer in the next-gen marketer. According to Paul, marketing needs to become a marketer plus machine to optimize strategies and attract modern consumers [1].

The Key Opinion Leader (KOL) strategy is one of the strategies that benefit greatly from AI integration. KOLs, also known as thought leaders, are individuals with significant influence and credibility within their respective communities [2]. The name KOL comes from the theory of the Two-step flow of communication [3] and the key characteristics of KOLs include originality, uniqueness, and a strong presence in their respective fields [4]. Therefore, they play a crucial role in healthcare, media, and marketing [2,5,6] and many other industries. Nowadays, the internet has provided contemporary individuals with more comprehensive channels for information acquisition, and the empowerment of AI has further enhanced the efficiency of search. According to data from the statistical website Statista, as of April 2024, the number of global internet users reached 5.44 billion, and the number of social media users reached 5.07 billion. Such a vast social media user base inevitably generates economic benefits, and this significant profit potential has been detected by many brands as a business opportunity. However, as consumer purchasing decisions are changing, many people turn to social media to seek feedback and reviews from other users before, as a foundation for their decision-making [7]. KOLs are thus considered a powerful marketing tool to increase brand awareness, engagement, and ultimately sales [8]. With the convenience of online social media platforms, this new-age marketing method utilizes the booming online social networks through word-of-mouth marketing [9].

This paper aims to explore how AI can enhance and optimize KOL marketing strategies. By leveraging the capabilities of AI, businesses can unlock new possibilities in the selection, content generation, deployment, and performance analysis of KOL-driven campaigns. The objective is to provide a comprehensive understanding of how AI can be integrated into KOL marketing, ultimately leading to more effective and impactful campaigns.

The rest of this paper is organized as follows. Section 2 provides an overview of KOL marketing and figures out the differences between KOL and influencers. Section 3 gives an elaboration on the application of AI in KOL marketing, which is divided into four parts. Section 4 discusses the problems and potential challenges of AI-driven KOL marketing. Section 5 concludes the paper.

2. Overview of KOL Marketing

2.1. History and Development of KOL Marketing

KOL marketing has experienced rapid growth over the past few years. Its history can be traced back to the following stages: early exploratory stage, rapid growth stage, standardized development stage, and AI-driven stage.

In the early stages of KOL marketing, the main social media platforms relied on were Facebook, Instagram, and Xiaohongshu. Brands are beginning to experiment with leveraging the communication power of KOLs by inviting them for content promotion and brand endorsement. For example, in 2015, Becca Cosmetics collaborated with beauty maven and YouTuber Jaclyn Hill on a limited-edition highlighter called Champagne Pop. Jaclyn Hill is considered a KOL for her expertise and influence in the beauty space, and the collaboration broke Sephora's sales record.

With the rise of short videos and other emerging social media, KOL marketing has rapidly become one of the mainstream marketing methods. Brands begin to gain insights into the match between KOLs and audiences and optimize KOL selection and content strategies. Then, Multi-Channel Network agencies emerged, which provide comprehensive operational support for KOLs. As an illustration, consider TikTok, which has launched the "Star Platform" that focuses on KOLs. TikTok has developed an advertising ecosystem providing brands with comprehensive marketing support, encompassing integrated marketing, lifecycle marketing, and performance marketing strategies for both business-to-business (B2B) and business-to-consumer (B2C) markets [10].

KOL marketing faces several challenges, including fake followers, transparency issues, and ethical concerns. In response, various industries are progressively adopting standardized mechanisms such as key performance indicators (KPIs) and credit evaluations. As brands increasingly prioritize the quality of KOL content and the consumer experience, they are seeking deeper content collaborations. Despite these challenges, the risks associated with KOL marketing are generally lower than those of traditional influencer marketing, as the lack of transparency can quickly reveal fraudulent KOLs.

The use of AI technology in KOL marketing became more widespread after 2023 with the generative AI appearing. In previous years, selecting the right KOL often incurred significant costs for brands. However, AI offers the potential to both increase revenue and reduce expenses. Additionally, the study shows that AI bots can be as efficient as trained salespeople and even four times more efficient than inexperienced salespeople: Let the virtual AI assistants have an initial interaction with the customer before handing over the most promising leads to a human salesperson [11].

2.2. The Difference between KOL and Influencer Marketing

KOLs are a subset of influencers, indicating that numerous KOLs can be considered as influencers, yet not every influencer is a KOL. The key distinction between KOL marketing and influencer marketing lies in the depth of expertise.

KOLs are recognized as authoritative figures who possess profound knowledge and credibility within a particular domain or industry. They are viewed as subject matter experts whose opinions and recommendations carry significant weight. In contrast, influencers may not necessarily have specialised expertise, but rather gain followership based on their personal experiences, opinions, and ability to connect with their audience.

This difference in the level of expertise is the defining characteristic that sets KOL marketing apart from the broader concept of influencer marketing.

3. The Application of AI Technology in KOL Marketing

3.1. Automated KOL Selection

In the era of big data, information is updated rapidly, and the types of data are becoming increasingly complex. To help companies extract valuable insights and identify suitable Key Opinion Leaders (KOLs) for their products, organizations can utilize advanced machine learning (ML) algorithms. These algorithms can uncover new and latent features of the data [12], allowing for quick matching based on the brand's objectives and audience characteristics.

Several organizations, such as Digital Business Lab, specialize in providing KOL insights. They employ big data analytics to ensure the authenticity of audiences, helping businesses avoid targeting fake followers. AI can also rank and score potential KOLs using a weighted combination of audience metrics, content, and performance, aiding companies in selecting KOLs that align with their marketing objectives and budget.

Predictive analytics and machine learning are increasingly vital for measuring the potential impact of KOL collaborations and calculating return on investment (ROI). By analysing historical data and past campaign performance, these tools can forecast key metrics such as expected reach, audience engagement, and conversion rates. This provides businesses with informed decision support when partnering with KOLs, ensuring that investments are strategically aligned with marketing goals.

However, traditional methods of screening KOLs—based solely on metrics like retweets, comments, likes, and follower counts—are becoming ineffective in today's market, which is saturated with inauthentic online profiles. To address this, EDGE, a member of Publicis Groupe, has launched Fluency, a KOL content marketing solution tailored for the Chinese market. Fluency leverages data insights for planning, screening, evaluating, and tracking KOL content, emphasizing data quality to deliver more reliable results.

This platform balances content quality with performance metrics, particularly beneficial for KOL screening in thematic marketing campaigns across various verticals. Fluency has successfully helped Publicis Groupe identify high-quality KOLs for product recommendations, significantly boosting consumer engagement and reducing overall interaction costs by more than five times the industry average. This case illustrates the efficacy and practicality of the Fluency platform for KOL screening, demonstrating that the integration of predictive analytics and machine learning with processed data can yield more accurate and efficient KOL marketing strategies, ultimately leading to higher ROI.

3.2. AI-Powered Content Creation and Optimization

AI-powered tools can analyse KOL's past content performance, audience preferences, and industry trends to generate new content ideas tailored to the target market. This content-generation process can be further enhanced by leveraging the capabilities of the OpenAI language model, which is a powerful tool for generating AI content [13].

The OpenAI language model is trained on a large amount of text data, which enables it to learn the patterns, style, and tone of the training data and generate new text accordingly. This tool utilises a recurrent neural network (RNN) architecture, making more accurate predictions compared to traditional rule-based chatbots [13]. The tool employs advanced machine learning algorithms that can analyse and comprehend natural language, enabling it to generate grammatically correct, error-free, and tailored to a specific target audience.

Furthermore, OpenAI can help optimise the generated content for search engines, ensuring that the content reaches a wider audience. By fine-tuning templates, the tool can generate more content traffic. The sophisticated machine learning models underpinning this tool go beyond simple rule-based systems, allowing for more nuanced and contextual language generation that is better suited to the needs of the target audience [13].

Jasper, an innovative company that emerged with the AIGC wave, has quickly become a rising star in the marketing space with its conversational AI technology based on the OpenAI API, valued at $1.25 billion since its inception in 2021.Jasper's AI tools not only simplify the copywriting process for marketers, but also enable ads through more than 50 built-in templates, blogs, e-commerce, emails, videos and websites, and other multi-scenario copy generation to meet the stylised needs of platforms such as Facebook, Google, Amazon, Instagram, LinkedIn and more. It is especially worth mentioning that Jasper AI's personalised service can adjust the tone of the copy according to the KOL's personal style and strengthen the connection with the audience, which not only improves the marketing effect, but also provides more creative space for the KOL. Through the deep integration of technology and marketing, Jasper shows the unlimited potential of AI in KOL marketing and leads the innovative change of the marketing industry.

According to Gurgul et al., machine learning models with natural language processing (NLP) data integration can enhance forecasting performance, and various NLP techniques can capture nuanced market sentiment [14]. Taking advantage of this capability, AI sentiment analysis can monitor reactions and feedback to KOL content, enabling brands to quickly identify and address any potential issues or negative sentiment. This real-time insight can also help brands correct and optimise their content strategy and engagement with KOLs.

3.3. Data-Driven User Analysis and Personalised Content Recommendation

Data-driven user analytics involves the use of large amounts of user-generated data to understand user behaviours, preferences and interactions to develop KOL strategies for specific groups of users. Such analytics are essential for the development of intelligent systems that can provide personalised recommendations, enhance the user experience and support the decision-making process [15]. Meanwhile, the concept of user profiling in the field of artificial intelligence refers to analysing the expected number of potential buyers and data based on the user's age, gender, industry, personality preferences and other characteristics [16]. In addition, data-driven 'personas' are those that combine an emotional understanding of people with logical thinking, using large amounts of user data and avant-garde computational methods to help organisations make more accurate choices when processing information and communicating with machines [17]. This precision marketing model develops more result-oriented and action-oriented plans for more precise, high-return investment in marketing activities. It combines traditional marketing with the AI big data model based on full knowledge and understanding of customer information, targeting customer preferences, and targeting specific products [18].

3.4. Comprehensive Campaign Performance Tracking and Analysis

Effective AI-driven KOL marketing relies on comprehensive campaign performance tracking and analysis. To achieve this, scalability, and key performance indicators (KPIs) are crucial for marketers to create efficient campaigns and accurately interpret results [19].

Adwan et al. proposed a digital marketing data analytics model that addresses this need [20]. This model analyses website performance, social media metrics, email marketing performance, customer data for targeting and personalization, and customer journey analysis to assess campaign effectiveness and inform strategy. The framework is particularly relevant for e-commerce platforms like Taobao and Amazon, enabling them to recommend KOL marketing products aligned with customer preferences based on their interaction data.

Furthermore, organizations can leverage data-driven media planning strategies that utilize attribution models. These models collect touchpoint data and track customers from initial contact to potential new orders, allowing marketing teams to identify the most effective creative assets and channels [21]. This data-driven approach not only enhances the relevance of marketing efforts to customer preferences but also improves the accuracy of market forecasts.

Customer journey analytics involves analysing customer behaviour across channels and time to understand its impact on business outcomes [22]. This is particularly relevant in the context of AI-powered KOL marketing, as it allows for a deeper understanding of how customers interact with brand messages and influencers. It can also help identify weak spots and enhance the user experience. By examining the consumer journey from initial exposure to post-purchase satisfaction, businesses can segment their consumer base and tailor their marketing efforts to each customer's demographics, interests, and needs [23]. This is where AI-driven KOL marketing can play a crucial role. By leveraging AI to analyse customer journey data, businesses can identify the most influential KOLs for specific customer segments and optimize their outreach strategies. Through this analysis, businesses gain insights into customer behaviour at every stage, enabling them to fine-tune their strategies for optimal customer satisfaction and marketing ROI. This data-driven approach to KOL selection and engagement, empowered by AI, represents a new chapter in brand marketing, where customer understanding and personalized experiences drive success.

4. Problems and Potential Challenges

The utilization of big data marketing and customized content allows for the quantitative evaluation of advertising and social media campaign effectiveness in the digital domain. However, these data-driven practices have also given rise to concerning trends, such as the proliferation of fake news and the misuse of advertising activities [24].

Although AI-enabled KOL marketing offers companies convenience by delegating certain tasks to AI, it also presents challenges and potential drawbacks. With AI-driven data analysis, user recommendations become increasingly precise, coupled with the influence of KOLs and extensive advertising, leading to the establishment of deceptive marketing schemes. This results in substandard product quality for users and issues with post-sales support.

Furthermore, the development of AI models raises privacy concerns for users. As highlighted by Liu and Terzi, the more sensitive information a user discloses, the higher their privacy risk [25]. Individuals may express dissatisfaction with KOL marketing promotions and the information silos that result from these practices.

In summary, while data-driven marketing techniques can enhance campaign effectiveness, the integration of AI and KOL influence also introduces risks related to misinformation, deceptive practices, and user privacy. Companies must carefully navigate these challenges to ensure ethical and transparent marketing strategies that prioritize the needs and trust of their customers.

5. Conclusion

Marketing is undergoing a methodological, technical, and cultural paradigm shift that augments and amplifies traditional outbound marketing with a more modern, inbound marketing approach. This transition involves adopting sophisticated, data-driven, and customer-centric marketing techniques, as opposed to solely relying on interruptive, one-way communication channels.

The integration of AI in KOL marketing has revolutionized the way brands approach influencer marketing. By leveraging AI-powered tools, businesses can optimize KOL selection, content creation, and campaign performance tracking, leading to more effective and impactful marketing strategies. Additionally, the use of AI in KOL marketing has the potential to increase revenue and reduce costs by improving marketing decisions, automating simple tasks, and enhancing customer engagement.

However, it also raises concerns about privacy, misinformation, and deceptive practices. With AI-driven data analysis, user recommendations become increasingly precise, coupled with the influence of KOLs and extensive advertising, leading to the establishment of deceptive marketing schemes. This results in substandard product quality for users and issues with post-sales support. Furthermore, the development of AI models raises privacy concerns for users, as the more sensitive information a user discloses, the higher their privacy risk.

To navigate these challenges, companies must prioritize transparency, ethics, and customer trust. By doing so, they can harness the power of AI-driven KOL marketing to build stronger relationships with their target audience, drive business growth, and stay ahead of the competition.