Analyzing the Impact of Artificial Intelligence and Sustainability on Gen Z Consumer Purchase Intentions: A Case Study of L’Oréal Cosmetics Indonesia

The dynamic landscape of consumer behavior, shaped by technological advances and sustainability considerations, has led to a reassessment of retail strategies, especially in the beauty industry. This study centers on the intersection of Artificial Intelligence (AI) and sustainability, particularly concerning Generation Z (Gen Z) consumers in the Indonesian beauty market. It explores how these factors influence Gen Z’s purchase decisions, offering insights for beauty brands to adapt strategically. Garnier, a L’Oreal subsidiary, faces heightened competition in the dynamic beauty market, especially with the emergence of local beauty products, adding complexity to its business landscape. This intensifies the need for strategic responses to maintain a competitive edge in the cosmetics industry. The research assesses the impact of AI technology, specifically using Garnier Skin Coach AI, on Gen Z’s purchase intentions for Garnier skincare products in Indonesia. It also examines the influence of sustainability on Gen Z’s preferences and purchase decisions in the Indonesian beauty market, adopting the Stimulus-Organism-reaction (SOR) model. Conducting a quantitative study with 400 Gen Z respondents, the research utilized online surveys through Qualtrics XM and analyzed data using Structural Equation Modeling (SEM) in SPSS AMOS 26.0. The findings highlight the substantial impact of AI technology, especially in enhancing hedonic values. Accurate information retrieval and interactive engagement create nuanced elements that heighten the appeal. Sustainability initiatives focusing on eco-friendly and cruelty-free practices significantly affect preferences, indicating a growing preference for sustainability-enriched experiences and affecting purchase intention. To enhance Garnier Skin Coach AI, a comprehensive strategy is recommended. This involves refining User-Centric Design, educating users, and boosting purchase intention through perceived utilitarian value. The proposed tactics align with customer preferences, encourage personalized interactions, integrate predictive skin insights, and the addition of e-wallet features.

Introduction

The global beauty industry stands at the intersection of innovation, aesthetics, and evolving consumer preferences. Beauty and personal care market value is estimated to reach more than $500 billion by 2022. By 2026, the market value is expected to reach more than $600 billion (Fig. 1) (Statista, 2022).

Fig. 1. Beauty and personal care market value worldwide from 2015 to 2028 by category. Source:  https://www.statista.com.

The beauty industry operates in both a global and local context. While global brands like L’Oréal dominate the market, local beauty product brands are emerging as significant players, especially in diverse regions like Indonesia. The local beauty market in Indonesia reflects a unique blend of cultural diversity and consumer preferences.

The digital revolution and the implementation of an omnichannel retail strategy have also contributed to increasing the market value of the industry (L’Oréal, 2019). Including 6 main product types: skin care, hair care, makeup, perfume, and hygiene products, the global beauty market recorded an annual growth rate of 6% in 2022 with market turnover estimatedat 250 billion euros. Skincare is the industry’s leading product category, accounting for 41% of the market in 2022, followed by haircare and makeup categories, accounting for 22% and 16% of the cosmetics market, then the rest are fragrance and hygiene products that contribute 11% and 10% in the beauty market (Table I) (L'Oréal, 2022).

Table I. Breakdown of the Market by Business Segment (%) in 2022

Category	Value
Skincare	41%
Haircare	22%
Makeup	16%
Fragrances	11%
Hygiene products	10%

Gen Z (born between the mid-1990s and the early 2010s) is a demographic generation that has grown up during a period of tremendous technological developments and considerable global change. This generation is distinguished by their digital fluency, social media participation, and distinct values and preferences from their predecessors (Seemiller & Grace, 2016). Gen Z customers have been shaped by their exposure to a hyper-connected world in which information and experiences are easily accessible via digital platforms. This exposure has shaped their expectations and behaviors, particularly in the area of consumerism (Kumar & Lim, 2018). As a result, understanding the subtleties of Gen Z customer behavior has become critical for organizations seeking their attention and commitment. One distinguishing feature of Gen Z consumer behavior is their increased knowledge of sustainability and environmental concerns. This generation grew up during a time when climate change, plastic pollution, and ethical consumption were all hot topics. As a result, brands that demonstrate a commitment to sustainability and responsible business practices are increasingly appealing to them (Smith & Johnson, 2020). When making purchase decisions, Gen Z customers prefer eco-friendly products, ethical sourcing, and corporate social responsibility (CSR) programs (Tuten & Bosnjak, 2021).

Garnier faces the challenge of knowing and responding to the features of Gen Z consumers in this dynamic market as a global cosmetic brand operating in Indonesia. Furthermore, the brand must navigate the intricate interplay between technical developments such as Artificial Intelligence (AI) and its target audience’s sustainability objectives. Garnier must assess the influence of these elements on Gen Z customer purchasing intentions in the Indonesian beauty sector to remain competitive and relevant.

Garnier’s Skin Coach application is at the vanguard of the modern skincare environment, harnessing the power of Artificial Intelligence (AI) technology developed over two decades of painstaking research. The Skin Coach AI app, which is supported by a large database collecting facial data from over 15,000 people, can give exact skincare advice suited to the user’s particular skin issues, all with the simple act of taking a single picture. Garnier Skin Coach AI (Fig. 2) analyses the six aspects of healthy skin, which are glowing, bounce, good pore quality, smoothness, evenness, and no black spots. This AI technology will analyze these six factors and provide customers with a detailed and personalized roadmap to help customers improve on all of them.

Fig. 2. Garnier skin coach AI. Source:  https://www.garnier.co.id/skin-coach.

The Garnier Complexion Coach app, which provides an unrivaled skincare experience, incorporates powerful AI and machine learning approaches to completely examine the user’s complexion before designing tailored skincare regimens. Furthermore, the program allows users to track their skincare progress and set up periodic reminders to stick to their individualized skincare routines; thus, the intersection of technology and skincare knowledge is exemplified.

In addition to the goal of customization, Garnier has embarked on a significant sustainability journey, recognizing the increasing relevance of ecologically responsible products among today’s consumers, particularly Gen Z. These strategic initiatives demonstrate Garnier’s commitment to aligning with the increasing interests and ideals of environmentally conscious consumers (Fig. 3) (Korhonenet al., 2020)

Fig. 3. Garnier’s sustainability activities. Source:  https://www.garnier.co.id.

In Indonesia’s skincare market, Garnier’s commitment to sustainability is significant. Gen Z consumers in Indonesia, like elsewhere, appreciate brands that prioritize sustainable practices (Rogerset al., 2019). Garnier’s efforts in this direction not only align with these consumer preferences but also contribute to a cleaner and more responsible beauty industry.

Problem Statement

Garnier, a subsidiary brand of L’Oréal, operates in the Indonesian beauty market, which has growing concerns around digitalization and sustainability and is facing a critical business issue. This issue revolves around comprehensively understanding and strategically responding to two pivotal factors that significantly influence consumer behavior: Artificial Intelligence (AI) and sustainability considerations. Garnier finds itself in the complex terrain of navigating these intertwined dynamics, particularly about the preferences of Gen Z consumers, who wield substantial influence, especially in the Indonesian market. According to skincare revenue data compiled by Compas for the second quarter of 2022, specifically sourced from sales within the e-commerce channel, Garnier occupies the fourth position with a revenue of 34.5 billion Indonesian Rupiah. It is noteworthy that this positioning places Garnier behind the top three brands, which predominantly trace their origins to local brand entities (Fig. 4) (Compas, 2022).

Fig. 4. Top 5 Perawatan Wajah Terlaris di E-commerce Kuartal II–2022. Source:  https://compas.co.id.

These symptoms and gaps encompass a range of challenges and opportunities that warrant comprehensive investigation.

This study aims to address the following research questions:

1. Does artificial intelligence (AI) technology influence the purchase intentions of Gen Z consumers when buying Garnier’s skincare products?
2. Does sustainability, encompassing eco-friendly raw material and cruelty-free product formulations, impact Gen Z consumers’ preferences and purchase intentions in Garnier’s skincare products?

This research aims to provide valuable insights into evolving consumer behavior and its impact on the beauty industry.

Literature Review

Artificial Intelligence (AI) in Beauty Market

Artificial Intelligence (AI) is widely used in skincare and beauty solutions because it adds an extra layer of analysis that allows the customer to make the best possible purchase. Artificial Intelligence (AI) is the ability of a digital computer or computer-controlled robot to perform tasks commonly associated with intelligent beings, according to the definition (Copeland, 2020). Machines, particularly computer systems, simulate human intelligence processes. Expert systems, natural language processing, and speech recognition are examples of AI applications.

The beauty industry has consistently performed successfully over time and differs significantly from other consumer industries. The beauty sector may engage consumers of various colors, ages, and backgrounds, ranging from beauty aficionados to new beauty product advocates. As more technology, such as skin scanning technologies, enters the mainstream, marketers will find it much easier to reach new customers.

Currently, several companies are establishing platforms that combine technology into the beauty and skincare markets. They are attempting to provide a better customer experience through the combination of data and personalization (BeautyTech.jp 2018). To remain industry leaders, top firms like L’Oreal, Unilever, Estee Lauder, P&G, COTY, and Shiseido have all implemented new methodologies that include AR, AI, and even VR.

Consumers nowadays spend a lot of time exploring the web on their smartphones and desktop computers. Because so many people are online, Beauty Tech is the ideal approach for the skincare sector to stay relevant in a saturated consumer environment. It allows people to try and discover new skincare and beauty items that they might have overlooked in their regular lives. Since these new digital tech services are bringing solutions to our fingertips, technology is redefining the notion of beauty and how the average consumer interacts with it.

To improve consumer experiences, the beauty sector has increasingly implemented AI-powered personalization tactics. Artificial intelligence (AI) technologies provide virtual try-on experiences, allowing consumers to see how beauty products would look on their skin, hair, or nails (Wanget al., 2019). Such virtual simulations have been shown to positively influence purchase intentions by lowering uncertainty and enhancing product selection confidence (Chiang & Trivedi, 2020).

Furthermore, beauty businesses are using AI-powered chatbots and virtual assistants to provide real-time guidance and suggestions based on particular skin types, tones, and concerns. This individualized contact increases consumer engagement and trust, which influences purchasing decisions (Xuet al., 2020).

Data and artificial intelligence (AI) are enabling new sorts of personalization and customization for consumers through beauty technology. Companies are incorporating these services into their websites and point-of-sale apps, improving the intimacy of their interaction with the consumer. The goal is for the consumer to believe that a certain company understands their wants and delivers them with the ideal product. Data suggest that online search demand for applications and innovative Beauty Tech has been stable, indicating that there is still room for exploration (Vanzella, 2019).

A big shift has occurred in retail to highlight the social construct of interaction between retailers and their consumers as a great bar-setter of customer pleasure. Simply said, the customer’s word has become more valued, and retail has evolved to suit consumer demand. Personalization, along with clean beauty and the expansion of e-commerce, has been a prominent priority for retail (Fine, 2019).

Personalization of Skincare Product

Personalization is beneficial to merchants because it allows them to sell products that their customers want and will buy. Certain retail enterprises may fail or fail to fulfill quota if no tracking or personalization is done. When we talk about retail, we mean both e-commerce and traditional brick-and-mortar stores. Value gain for the supplier can take the shape of increased expertise and enhanced customer satisfaction. Leeet al. (2019). Knowledge augmentation can be accomplished through surveys, email marketing, and even the incorporation of digital gadgets into physical storefronts. “Retailers can then better observe customer behavior, collect customer data, understand their needs, and provide personalized services” (Wetzlinger & Werner, 2017). This personalization may bring value for customers, but it may also raise privacy concerns.

Sustainability of Skincare Product

Sustainability has acquired significant attention and relevance in the skincare sector in recent years, driven by both customer demand and a growing worldwide awareness of environmental challenges (Finkbeineret al., 2010). Environmental and ethical practices of brands, particularly those in the cosmetics and skincare industries, are increasingly being investigated (Zhuet al., 2020). The creation of products with a focus on lowering their environmental impact is an important part of sustainability in skincare. This includes employing environmentally friendly components, lowering resource-intensive processes, and reducing waste creation (Zhuet al., 2020). Sustainable skincare products frequently include cruelty-free and ethically sourced components, which appeal to customers who value animal welfare and ethical sourcing (Rogerset al., 2019).

Another major area where sustainability strategies come into play is packaging. Sustainable skincare firms frequently use recyclable or biodegradable packaging materials to reduce the environmental impact of product packaging (Ghisolfiet al., 2013). To fit with sustainable packaging practices, single-use plastics are being reduced, and refillable packaging choices are being implemented (Finkbeineret al., 2010).

The development and distribution of skin care products is also environmentally friendly. To lower their carbon footprint, businesses are progressively adopting energy-efficient manufacturing processes, procuring renewable energy, and improving transportation logistics (Makvandiet al., 2020). Furthermore, ethical supply chain methods guarantee that materials are sustainably acquired, with an emphasis on fair labor conditions and environmentally responsible processes (Nasiret al., 2021).

Consumer concern about the environmental and ethical consequences of skincare products has increased demand for sustainability labels. Sustainability labels, such as cruelty-free, organic, and vegan labels, give consumers certainty about the ethical and environmental aspects of skincare products (Zhuet al., 2020). These labels not only guide purchase decisions but also build brand loyalty among environmentally-conscious consumers (Verssimoet al., 2020).

The presence of sustainability labeling on items in the Indonesian skincare sector represents a growing trend among Gen Z consumers. Sustainability labeling acts as a physical indication of a brand’s commitment to ethical and environmentally responsible practices, and this generation actively seeks out products that align with their ideals (Compas, 2022).

Gen Z in Indonesia

Gen Z, defined as people born between the mid-1990s and the early 2010s, is a significant and powerful demographic generation in Indonesia. The findings of the 2020 Population Census are dominated by Gen Z, which accounts for around 27.94% of Indonesia’s population or approximately 74.93 million people. With a population of over 75 million, Gen Z accounts for a sizable share of the country’s population. This generation grew up during a time of rapid technical breakthroughs, easy access to knowledge, and increased connectivity via digital platforms (Kusumawijaya, 2017).

In Indonesia, as in many other parts of the world, the features of Gen Z are defined by their digital fluency, social media involvement, and a distinct set of beliefs and interests. They are well-known for their active participation in online forums, where they exchange their thoughts, seek information, and express themselves (Ardisara, 2019). Understanding the dynamics of this generation is critical for businesses operating in Indonesia since their consumer behavior is heavily influenced by their unique characteristics and experiences (Putriet al., 2020).

Consumer Perceived Value

The appraisal of the discrepancies between what consumers spend and what they receive while buying is referred to as consumer-perceived value. It includes several dimensions, such as perceived utility value and perceived hedonic value.

1. Perceived utility value is related to practical benefits such as time and cost savings, as well as convenience.
2. Perceived hedonic value is concerned with experiencing aspects such as pleasure, relaxation, and involvement.

Perceived value in the context of technology systems is determined by aspects such as perceived simplicity of use and perceived utility. AI technology improves perceived value in online purchasing by giving accurate recommendations and personalized experiences.

Consumers, particularly the environmentally conscious Gen Z, have made sustainability a top priority. By assisting in the assessment of a product’s environmental impact, AI plays an important role in addressing sustainability problems. Artificial intelligence-powered systems can examine a product’s lifecycle and supply chain, highlighting the potential for environmentally friendly enhancements (Prakash & Saini, 2017). Furthermore, AI promotes transparency in raw material procurement and ethical production procedures, which resonate with consumers who value sustainability (Dangelico & Pontrandolfo, 2015). Brands that effectively convey their sustainable practices using AI-driven marketing campaigns might influence the preferences and intentions of Gen Z customers to support such brands (Berger & Ribeiro-Navarrete, 2020).

In the context of our research, we built a structural equation model to analyze the complex link between AI technology in online shopping platforms, environmental concerns, and customer purchase intents in Gen Z. Our investigation digs into the mediating roles of perceived hedonic and perceived utilitarian value, as well as the impact of sustainability considerations. Furthermore, we want to understand and evaluate the distinct effects of perceived hedonic value, perceived utilitarian value, and sustainability concerns as mediators in this relationship. Our findings provide practical insights for platform enterprises as a well-defined research trajectory for AI technology and sustainability, as well as facilitating the enhancement of consumer shopping perceived value, ultimately contributing to the sustainability of online shopping services.

AI Technology Experience

Through intelligent recognition and search, suggestions, and virtual customer support, AI technology improves the online purchasing experience. As a result, consumers will encounter three types of AI Experiences:

1. Accuracy Experience: AI makes use of huge amounts of data to assist consumers in swiftly finding products using text, voice, or image searches. This level of precision may be seen in the precise product recommendations offered by AI-powered search engines on platforms such as Taobao and Jingdong.

2. Insight Experience: Machine learning is used by AI to customize website content depending on user preferences, providing personalized solutions and accurate forecasts of user demands. Personalized product recommendations on Internet shopping platforms reflect this insight experience.

3. Interactive Experience: AI-powered virtual customer service assistants take the place of human customer service representatives, providing natural language interactions and assisting customers in making educated decisions. AI virtual assistants such as Echo and Tmall Genie exemplify this interactive experience.

The above accurately depicts the engaging experience that AI marketing technology provides to consumers (Yin & Qiu, 2021).

The SOR Model

The Stimulus-Organism-reaction (SOR) model (Fig. 5), first suggested by Mehrabian and Russell (1974), explains how environmental stimuli (S), such as images and sounds, influence our interior emotions and cognitive processes (O), ultimately leading to a reaction (R).

Fig. 5. Mehrabian and Russell the SOR model. Source: An approach to environmental psychology; MIT press: Mehrabian and Russell (1974).

The SOR model was applied to the shopping context (Donovan & Rossiter, 2010), while Erogluet al. (2001) expanded it to online shopping platforms to explore how preferences and cognitive states influence online purchasing behaviors. The SOR model has been used to investigate the following aspects of online shopping:

1. The effect of website environment on consumer behavior (Erogluet al., 2003).
2. How does the image of a website (for example, security, convenience, and entertainment) influence perceived quality and purchase intentions (Yang, 2009)?
3. The impact of consumers’ emotional states (such as pleasure and impulse) on their spending (Floh & Madlberger, 2013).
4. How the structure of a website affects purchase intentions through user engagement and acceptance (Lorenzo-Romeroet al., 2016).
5. The relationship between the ambiance of a website and customer satisfaction (Sanjeev & Chandan, 2017).
6. The reputation and image of e-commerce platforms (Fikriet al., 2019).
7. How does the image of an online store affect consumer behavior (Yun & Good, 2007)?
8. The impact of technology, such as artificial intelligence (AI), on the internal mechanics of purchase intention (Cui & Lai, 2013).

Conceptual Framework

Previous studies have conducted this research using the SOR framework (Figs. 6 and 7). Yin and Qiu (2021) investigate AI marketing on online buying platforms and investigate its impact on consumers’ perceived utilitarian and hedonic values, with implications for precision marketing. The study confirms that AI marketing technology improves consumers’ perceived value while shopping online. This highlights the need for organizations to continue to invest in AI technology and optimize its applications, such as intelligent search and recommendations. This has the potential to enhance the shopping experience and boost consumer purchasing urges. Recognizing such limitations, future research should investigate additional variables and moderating effects to acquire a thorough knowledge of the impact of AI marketing on consumer behavior. AI marketing affects consumers’ perceived value in online buying, highlighting the significance of long-term AI management for precision marketing and improved shopping experiences.

Fig. 6. The SOR framework in previous study, AI technology and online purchase intention source: AI technology and online purchase intention: Structural equation model based on perceived value (Yin & Qiu, 2021).

Fig. 7. The SOR framework in previous study, the influence of sustainable positioning on eWOM and brand loyalty. Source: The influence of sustainable positioning on eWOM and brand loyalty: analysis of credible sources and transparency practices based on the S-O-R Model (Mim & Stacy, 2022).

Sustainability Initiative has the potential to address local issues related to global sustainability issues (such as health, mobility, and biodiversity loss). Local actors frequently plan, execute, and oversee these. New paradigms for doing, thinking, and organizing are offered by sustainability initiatives (e.g., social, technological, economic, socio-technical, or socioecological). For instance, these could be initiatives, goods, methods, strategies, or technological advancements (Bennettet al., 2016; Gorissenet al., 2018). Their emphasis can be on urban agroecology, climate-smart cities, or green design, depending on the context and agency (individual or collective) (Pereiraet al., 2018). Sustainability Initiatives are essential for transformations because they have the power to gradually unite to move prevailing regimes toward more sustainable paths, promoting transformative change (Pereiraet al., 2018; Lamet al., 2019). Sustainability initiatives are referred to by different names in different research domains: social innovations (Westley & Antadze, 2010; Mooreet al., 2015), grassroots innovations (Seyfang & Smith, 2007), seeds of a good Anthropocene (Bennettet al., 2016), transition experiments (Canigliaet al., 2017; Sengerset al., 2019), and transition initiatives (Frantzeskakiet al., 2016; Gorissenet al., 2018). Sustainable positioning was supported by reputable sources such as social influencers and government organizations. It had a favorable effect on brand attachment, trust, and identification among Gen Z people. Also, higher-income groups and females were more likely to convert to sustainable brands (Mim & Stacy, 2022).

Based on previous research, this study’s conceptual framework (Fig. 8) uses the Stimulus-Organism-Response (SOR) model, which is established in consumer behavior research. The SOR framework has been widely recognized and applied in consumer behavior research to understand the relationships between stimuli, internal cognitive processes, and behavioral outcomes. It offers a structured approach for analyzing the impact of Artificial Intelligence (AI) and sustainability considerations on Gen Z consumer purchase intentions in the context of the cosmetics sector, utilizing L’Oréal as a case study.

Fig. 8. Proposed framework.

Research Hypothesis

The stimulation of AI marketing technology encourages customers to make sophisticated purchasing decisions (Fanet al., 2018), which can save consumers time and money when buying (Zeithaml, 1988), making shopping selections more accurate and conducive to improving consumers’ shopping efficiency (Rosenberg, 2018). The use of artificial intelligence technologies such as visual recognition, speech recognition, and machine vision can provide better insight into consumer behavior in five areas: problem recognition, information collection, alternative evaluation, purchase decision-making, and post-purchase evaluation (Janet al., 2018), providing consumers with a more efficient consumption reference and enriching and smoothing the value perception form of the entire consumption process. Accurate and expanded information stimulation during the engagement with an e-commerce platform can broaden the boundaries of consumers’ target selection. In terms of pleasure, respect, and attention, the visual impact of online purchasing and surfing can match customers’ individualized, customized information needs. Artificial intelligence can search for words, images, and voice using machine vision and deep learning technology, allowing consumers to accurately identify product features, enriching the consumer search experience, saving consumers’ shopping time, reducing boredom in the shopping process, and increasing consumer interest in the consumption process (Aakash & Panchal, 2019). The following hypotheses are put forward:

• Hypothesis 1a (H1a): The improvement of the accuracy experience of Garnier Skin Coach AI technology while analyzing and recommending skincare products positively influences the formation of consumers’ perceived utilitarian value.
• Hypothesis 1b (H1b): The improvement of the accuracy experience of Garnier Skin Coach AI technology while analyzing and recommending skincare products positively influences the formation of consumers’ perceived hedonic value.

Machine learning is a crucial subfield of artificial intelligence. It can accurately analyze customers’ preferences and purchasing needs, push individualized information to existing and new customers, and deliver more effective buy suggestions to consumers, according to an evolutionary behavior algorithm based on empirical data (Ma & Sun, 2020).

However, the intelligent push of information must affect the purchase behavior of consumers by some means. The information that is pushed must make customers see the utility and effectiveness of the information or bring a particular level of engagement and enjoyment of physical and mental pleasure into the buying activity, or the purpose and significance of shopping will be lost (Liuet al., 2019). As a result, the following hypotheses are put forward:

• Hypothesis 1c (H1c): The improvement of insight experience of Garnier Skin Coach AI technology while analyzing and recommending skincare products positively influences the formation of consumers’ perceived utilitarian value.
• Hypothesis 1d (H1d): The improvement of insight experience of Garnier Skin Coach AI technology while analyzing and recommending skincare products positively influences the formation of consumers’ perceived hedonism.

Consumers can have a positive or negative emotional connection to and be influenced by the power provided by artificial intelligence, even if they are aware that this power does not represent genuine emotional contact between humans (Guerra, 2018). According to an analysis of the Jingdong AI robot’s application status, some scholars believe that AI technology can help with the automation of consumer feedback management and that emotional analysis powered by AI can help marketers better respond to consumers, give robot customer service in e-commerce platforms better intelligence quotients, and bring a better consumer value experience to consumers’ online shopping. As a result, the following hypotheses are put forward:

• Hypothesis 1e (H1e): The improvement of user interaction experience of Garnier Skin Coach AI technology while analyzing and recommending skincare products positively influences the formation of consumers’perceived utilitarian value.
• Hypothesis 1f (H1f): The improvement of accuracy of user interaction with Garnier Skin Coach AI technology while analyzing and recommending skincare products positively influences the formation of consumers’ perceived utilitarian value.

Linet al. (2016) discovered that transparency is an important driver of sustainability perceived value, which can generate brand loyalty. An ecologically conscious consumer wants complete transparency across the retail chain and is eager to learn where and how items are manufactured, as well as their design origins and quality (Amedet al., 2019). Consumers exposed to advertising with very clear information had more favorable attitudes about sustainable product promises, according to studies (Chenet al., 2015; Fiore & Kim, 2013). Furthermore, a high level of perceived sustainability transparency can assist a company in gaining customers’ respect for socially responsible initiatives and avoiding greenwashing allegations (Liet al., 2019; Jianget al., 2014). As a result, the following hypotheses are put forward:

• Hypothesis 1e (H1g): The improvement of Garnier’s transparency efforts in explaining the Sustainability Initiative of their products positively contributes to the development of consumers’ perceived utilitarian value.
• Hypothesis 1f (H1h): The improvement of Garnier’s transparency efforts in explaining the Sustainability Initiative of their products positively contributes to the development of consumers’ perceived hedonic value.

Empirical research has shown that the perceived value of an online store’s image in an online shopping environment can influence both conscious and impulsive purchasing behavior (Jiang & Zhao, 2013). A vast number of research have found that buy intention is dependent on the product’s symbolic and functional features, whereas utilitarian value is reflected in the practicality, convenience, and cost-saving experienced by consumers during the purchasing process (Batra & Ahtola, 1990; Zhuet al., 2009; Chenet al., 2015). When customers just focus on the visual benefits of the product itself, the utilitarian value will play a role, influencing consumers to buy goods based on their own needs during the online purchasing process (Fiore & Kim, 2013). The utilitarian value provided by technological convenience and improvements in shopping efficiency can increase consumer happiness, increase consumption desire, and stimulate re-consuming (Liet al., 2019). The subjective sensation of pleasure, curiosity, and relaxation gained by consumers while shopping is referred to as perceived hedonic value. The enjoyment of the purchase process can influence impulse consumption intention. The perceived function value and perceived hedonic value of the consumer process encourage online repurchase activity (Lorenzo-Romeroet al., 2016). Artificial intelligence improves marketing by making it more intelligent, efficient, conducive to customer decision-making, and capable of achieving a greater marketing result. As a result, the following assumptions are advanced:

• Hypothesis 2a (H2a): The perceived utilitarian value derived from Garnier Skin Coach AI technology experience and Sustainability Initiative positively influence Gen Z consumers’ purchase intentions for Garnier skincare products.
• Hypothesis 2b (H2b): The perceived hedonic value derived from Garnier Skin Coach AI technology experience and Sustainability Initiative Gen Z consumers’ purchase intentions for Garnier skincare products.

In summary, the hypotheses given in this study revolve around the important factors impacting the purchase intentions of Gen Z customers for Garnier skincare products and can be characterized using the SOR framework, as shown in Fig. 9. We investigated utilitarian and hedonistic values generated from the Garnier Skin Coach AI technology experience, as well as the impact of Garnier’s transparency efforts in discussing environmental and social impacts. We hope to get a better understanding of the complex interplay between technology, sustainability, and consumer behavior in the cosmetics sector by investigating these ideas. The findings of these hypotheses will provide useful insights for Garnier and other businesses looking to engage and resonate with Gen Z customers while also matching with changing consumer attitudes and expectations.

Fig. 9. Proposed framework and hypothesis.

Methodology and Data Collection

A quantitative method approach was used to empirically test the research hypotheses.

The study will use random sampling and the sample size will be calculated based on Slovin’s Formula to choose a representative and unbiased sample from the Indonesian Gen Z population. According to the 2020 Population Census data, Gen Z constitutes around 27.94% of Indonesia’s population, which is equivalent to approximately 74.93 million people. If Slovin’s formula is used to calculate the sample size, with a margin of error of 5%, the total sample size for this research is 400 respondents.

The major form of data gathering was through online surveys utilizing Qualtrics XM. The questionnaire was developed for Gen Z customers in Indonesia aged 18 to 26 who used skincare products and or were interested in trying Garnier Skin Coach. The study asked participants about their experiences with Garnier Skin Coach, their judgments of utilitarian and hedonic values, sustainability concerns, and purchase intentions for Garnier skincare products.

For this study, SPSS AMOS 26.0 was used to analyze the data. First, the scale’s sample data was subjected to a normality test, then a reliability test, and a validity test. Second, confirmatory factor analysis was performed on the Structural Equation Model (SEM) to validate that the data and model fit well. Finally, the theoretical model was subjected to overall and multi-group path coefficient analyses.

Result

The demographics of the study participants provide a comprehensive overview of the essential profile characteristics needed to understand the context and relevance of the research findings. Examining these demographics helps to identify the diverse perspectives present within the participant pool, which contributes to the study’s richness and depth. Of all the participants, 88.50% were female, and 11.50% were male. In terms of age, 31.25% were 18–20 years old, 36.25% were 21–30 years old, and 32.50% were aged 23–26 years old. Among the participants, 45.25% were university students, 27.00% were full-time workers, 11.00% were freelancers, 8.75% were part-time workers, and 5.75% were entrepreneurs. West Java had the highest number of participants (24.75%), followed by DKI Jakarta (19.50%), and so on. According to the survey, 16.50% of the respondents reported having a monthly household income between Rp 4,500,001 to Rp 6,500,000, while 15.00% reported having an annual household income between Rp 6,500,001 to Rp 8,500,000. Additionally, 15.00% of the respondents reported having an annual household income between Rp 1,500,001 and Rp 2,500,000, and so on. Regarding the frequency of searching for information about skincare products, about 39.50% of participants searched once a week, 29.00% searched once a month, 22.75% searched mostly every day, and 8.75% searched less than one month. For a detailed breakdown of the sample characteristics, please refer to Table II.

Table II. Characteristics of Respondents (N = 400)

Characteristics		Frequency	Percentage
Gender
Male		46	11.50%
Female		354	88.50%
Age
18 - 20		125	31.25%
21 - 23		145	36.25%
23 - 26		130	32.50%
Occupation
University student		181	45.25%
Entrepreneur		23	5.75%
Freelancer		44	11.00%
Full-time worker		108	27.00%
Part-time worker		35	8.75%
Others		9	2.25%
Location
DKI Jakarta		78	19.50%
West Java		99	24.75%
Banten		43	10.75%
Central Java		47	11.75%
DI Yogyakarta		27	6.75%
East Java		54	13.50%
Others		52	13.00%
Household Income	SES (Nielsen)
Less than Rp 1.500.000	Lower	41	10.25%
Rp 1.500.001 - Rp 2.500.000		60	15.00%
Rp 2.500.001 - Rp 3.500.000	Middle 2	48	12.00%
Rp 3.500.001 - Rp 4.500.000		44	11.00%
Rp 4.500.001 - Rp 6.500.000	Middle 1	66	16.50%
Rp 6.500.001 - Rp 8.500.000		60	15.00%
Rp 8.500.001 - Rp 11.000.000	Upper 2	31	7.75%
Rp 11.000.001 - Rp 20.000.000		23	5.75%
Rp 20,000.001 - Rp 30.000.000	Upper 1	9	2.25%
Rp 30,000.001 - Rp 40.000.000		9	2.25%
More than Rp 40.000.000		9	2.25%
Frequency of Searching Information about Skincare Products
Mostly everyday		91	22.75%
Once a week		158	39.50%
Once a month		116	29.00%
Less than once a month		35	8.75%

Data Analysis

To perform Structural Equation Modeling (SEM) accurately, it is necessary to examine the normality of the data. This involves checking the distribution of the data’s variables to see if they conform to a normal distribution curve. The normal distribution curve is a bell-shaped curve that represents the frequency distribution of a random variable. It is important to verify normality because SEM assumes that the data follows a normal distribution. If the data is not normally distributed, the results of the analysis may be biased, or the statistical tests used to evaluate the model may not be valid. Therefore, it is crucial to assess the normality of the data before conducting SEM. We used AMOS 26.0 to investigate and found that all latent variables scores of critical kurtoses (C.R.) fell within an acceptable range of −2.58 to +2.58 (Table III), as determined by the z-score criterion. This means that the data distributions maintained normality. This critical assessment follows the best statistical practices (Tabachnick & Fidell, 2007) and is essential for ensuring the accuracy and reliability of SEM outcomes. Additionally, the absence of significant outliers within the latent variables strengthens the credibility of the dataset, making it a strong foundation for robust and valid statistical inferences (Field, 2013).

Table III. Assessment of Normality

Variable	Minimum	Maximum	Skew	C.R.	Kurtosis	C.R.
AC1	2.000	5.000	−0.427	−3.488	−0.255	−1.042
AC2	2.000	5.000	−0.549	−4.483	−0.342	−1.397
AC3	1.000	5.000	−0.515	−4.207	−0.241	−0.985
IS1	1.000	5.000	−0.722	−5.895	0.547	2.233
IS2	1.000	5.000	−0.717	−5.850	0.446	1.821
IS3	2.000	5.000	−0.450	−3.678	−0.467	−1.909
IT1	2.000	5.000	−0.597	−4.876	−0.220	−0.900
IT2	2.000	5.000	−0.525	−4.288	−0.299	−1.219
IT3	2.000	5.000	−0.520	−4.243	−0.512	−2.091
SI1	2.000	5.000	−0.550	−4.489	−0.278	−1.133
SI2	2.000	5.000	−0.560	−4.572	−0.344	−1.404
SI3	2.000	5.000	−0.669	−5.459	−0.059	−0.242
SI4	2.000	5.000	−0.693	−5.657	−0.095	−0.388
UV1	1.000	5.000	−0.660	−5.387	0.025	0.103
UV2	1.000	5.000	−0.660	−5.388	0.210	0.858
UV3	1.000	5.000	−0.747	−6.102	0.455	1.857
UV4	1.000	5.000	−0.713	−5.824	0.239	0.977
UV5	1.000	5.000	−0.663	−5.415	−0.086	−0.352
UV6	2.000	5.000	−0.700	−5.718	−0.198	−0.809
UV7	1.000	5.000	−0.778	−6.355	0.487	1.989
UV8	2.000	5.000	−0.595	−4.858	−0.350	−1.429
UV9	2.000	5.000	−0.488	−3.985	−0.342	−1.396
HV1	2.000	5.000	−0.483	−3.945	−0.602	−2.460
HV2	2.000	5.000	−0.503	−4.108	−0.389	−1.589
HV3	1.000	5.000	−0.501	−4.087	−0.337	−1.378
HV4	1.000	5.000	−0.622	−5.082	−0.170	−0.694
HV5	2.000	5.000	−0.764	−6.235	−0.003	−0.013
HV6	2.000	5.000	−0.572	−4.670	−0.577	−2.356
HV7	2.000	5.000	−0.592	−4.836	−0.358	−1.460
HV8	2.000	5.000	−0.587	−4.793	−0.367	−1.500
CPI1	2.000	5.000	−0.761	−6.217	−0.027	−0.109
CPI2	1.000	5.000	−0.593	−4.840	−0.263	−1.075
CPI3	1.000	5.000	−0.607	−4.956	−0.034	−0.139
CPI4	1.000	5.000	−0.745	−6.083	0.148	0.606
CPI5	2.000	5.000	−0.804	−6.567	−0.135	−0.551
CPI6	2.000	5.000	−0.605	−4.939	−0.384	−1.569
CPI7	2.000	5.000	−0.589	−4.810	−0.349	−1.426
CPI8	2.000	5.000	−0.697	−5.694	−0.096	−0.391
Multivariate					926.140	167.973

Afterward, the data was further analyzed using AMOS 26.0. and revealed convincing results, as shown in Table IV, regarding the reliability and validity of the latent variables in the study. Factor loadings were carefully examined, and all items showed factor loadings above the commendable threshold of 0.7. This demonstrates a strong and reliable connection between the observed items and their underlying latent constructs, confirming the structural integrity of the measurement model. Furthermore, the Average Variance Extraction (AVE) values for each latent variable were above the threshold of 0.5, indicating a high degree of agreement among the items within these constructs. This suggests that they can effectively measure the latent variables.

Table IV. Analysis of Reliability and Validity

Latent variable	Items	Factor load	Composite Reliability	AVE
Accuracy (AC)	AC1	0.898	0.958231545	0.884360314
	AC2	0.894
	AC3	0.909
Insight (IS)	IS1	0.728	0.92236793	0.799737391
	IS2	0.85
	IS3	0.91
Interactivity (IT)	IT1	0.786	0.898205895	0.746303218
	IT2	0.772
	IT3	0.796
Sustainability initiative (SI)	SI1	0.814	0.914765366	0.729667697
	SI2	0.669
	SI3	0.712
	SI4	0.786
Utility value (UV)	UV1	0.651	0.931268837	0.601498136
	UV2	0.596
	UV3	0.648
	UV4	0.571
	UV5	0.662
	UV6	0.64
	UV7	0.625
	UV8	0.589
	UV9	0.662
Hedonic value(HV)	HV1	0.659	0.929584757	0.622793907
	HV2	0.648
	HV3	0.64
	HV4	0.65
	HV5	0.668
	HV6	0.638
	HV7	0.684
	HV8	0.672
Consumer purchase intention (CPI)	CPI1	0.589	0.927698684	0.617472536
	CPI2	0.672
	CPI3	0.643
	CPI4	0.636
	CPI5	0.706
	CPI6	0.722
	CPI7	0.773
	CPI8	0.715

Additionally, the internal consistency reliability was assessed through Cronbach’s α coefficient (Table V), with a value of 0.9735 for the entire sample data, exceeding the conventional benchmark of 0.7. This highlights the reliability of the measurement model and emphasizes the precision and consistency of the latent variable assessments.

Table V. Cronbach’s α Analysis of Reliability

ANOVA
Source of variation	SS	df	MS	F	P-value	F critical
Rows	4335.52	399	10.86596	37.75554	0	1.121103
Columns	110.8189	37	2.995107	10.40698	1.304560	1.411424
Error	4248.76	14,763	0.287798
Total	8695.099	15,199
Cronbach’s α	0.973514

In this study, the Structural Equation Model (SEM) was meticulously constructed using Amos 26.0 software to assess the model’s fit to the data. The outcomes of the fitting test underscored the mode’s relatively robust fitting performance. Evidenced by a CMIN/DF ratio of 5.196, along with goodness-of-fit indices such as AGFI (0.583), NFI (0.731), IFI (0.771), and CFI (0.770), as well as RMSEA value 0.103. This conformed to established standards for model fitting. The adherence to recommended fit indices lends substantial support to the model’s adequacy in representing the underlying data structure, thus substantiating its suitability for subsequent path analysis.

The model and path coefficients, visually represented in Fig. 10, provide a comprehensive view of the strength and direction of the relationships among the variables in the theoretical model. These path coefficients quantitatively characterize the impact and direction of influence that one variable exerts on another, contributing to a deeper understanding of the structural relationships.

Fig. 10. Model and path coefficient.

The hypothesis test results for the theoretical model, as presented in Table VI, offer insights into the significance test of the relationships examined within the study. Notably, the majority of these relationships exhibited significant p-values, with values lower than 0.001, affirming they have a significant effect among relationships. However, there are exceptions to this trend. Specifically, the relationships of H1C, H1D, and H2A did not yield p-values below the 0.001 threshold. This indicates that there is no significant effect on these relationships.

Table VI. Results for Hypotheses Verification

Hypotheses	Path	B (Estimate)	S.E.	C.R	P (<0.05)	Result
H1a	AC → UV	0.108	0.033	3.262	0.001	Positive significant
H1b	AC → HV	0.158	0.038	4.11	***	Positive significant
H1c	IS → UV	0.06	0.04	1.491	0.136	Positive insignificant
H1d	IS → HV	0.004	0.046	0.094	0.925	Positive insignificant
H1e	IT → UV	0.277	0.052	5.317	***	Positive significant
H1f	IT → HV	0.213	0.059	3.631	***	Positive significant
H1g	SI → UV	0.687	0.054	12.757	***	Positive significant
H1h	SI → HV	0.66	0.06	11.074	***	Positive significant
H2a	UV → CPI	−0.049	0.113	−0.436	0.663	Negative insignificant
H2b	HV → CPI	0.868	0.132	6.588	***	Positive significant

The results of the hypothesis testing within this study provide an intricate perspective on the intricate dynamics of AI marketing technology experiences using Garnier Skin Coach AI and Garnier Sustainability Initiative’s impact on consumers’ perceptions and purchase intentions in the realm of online shopping platforms. The significant validation of H1a and H1b underscores that heightened accuracy in AI marketing technology experiences fosters an augmented perception of both utilitarian and hedonic values during the shopping process. In contrast, the non-significant outcomes for H1c and H1d indicate that while increasing insight experiences with AI technology may enhance consumer understanding, it does not significantly influence the perception of utilitarian and hedonic values. Meanwhile, H1e and H1f highlight the pivotal role of increased online interaction with AI marketing technology, contributing positively to consumers’ perceived utilitarian and hedonic values. Furthermore, H1g and H1h demonstrate that the strengthening of sustainability initiatives within AI technology enhances the perceived utilitarian and hedonic values, emphasizing the importance of sustainability considerations in online shopping experiences.

Regarding purchase intentions, the non-significant result for H2a suggests that the perceived utilitarian value resulting from AI technology does not significantly drive consumers’ purchase intentions. In contrast, the significance of H2b reveals that hedonic values derived from AI marketing technology play a pivotal role in shaping purchase intentions. This nuanced distinction suggests that while the utility of AI technology may not directly influence purchase intentions, the pleasurable aspects associated with its usage significantly contribute to the formation of consumer purchase intentions. These findings collectively provide a comprehensive understanding of the intricate interplay between AI marketing technology, consumer values, and purchase intentions in the context of online shopping platforms, emphasizing the multifaceted nature of consumer decision-making processes.

The study employed the Sobel test method to investigate the indirect effects within the research framework. Consumer purchase intention was designated as the dependent variable, while accuracy experience, insight experience, interactive experience, and sustainability initiative served as the independent variables. The study also incorporated perceived utilitarian value and perceived hedonic value as mediating variables to verify their roles in mediating the entire causal pathway.

The research findings, which are summarized in Table VII, reveal the presence of important indirect effects, primarily driven by the concept of perceived hedonic value. Most of these relationships exhibited significant p-values, with values lower than 0.001, affirming for each influence path indicates that there are significant indirect effects of AI technology experiences, accuracy, interactive experiences, sustainability initiatives, and perceived hedonic value on consumers’ purchase intentions. These indirect effects are particularly relevant in the context of AI technology experiences and sustainability initiatives, as they play a pivotal role in shaping consumers’ purchase intentions.

Table VII. Intermediary Test of Perceived Utilitarian Value and Perceived Hedonic Value

Path	t-stat	P (<0.05)	Result
AC → UV → CPI	−0.4298714	0.6672892	Negative insignificant
AC → HV → CPI	3.5142989	0.0004409	Positive significant
IS → UV → CPI	−0.416571	0.6769922	Negative insignificant
IS → HV → CPI	0.0869489	0.9307121	Negative insignificant
IT → UV → CPI	−0.4321987	0.665597	Negative insignificant
IT → HV → CPI	3.1646074	0.0015529	Positive significant
SI → UV → CPI	−0.4333767	0.6647412	Negative insignificant
SI → HV → CPI	5.6441459	0.00000002	Positive significant

On the other hand, the indirect effect of perceived hedonic value is found to be statistically insignificant when it comes to insight experiences. This means that while hedonic value effectively mediates the relationships between accuracy, interactivity, sustainability initiatives, and purchase intentions, it does not perform the same mediating role for the relationship between insight experiences and purchase intentions.

When considering utilitarian value, the study reveals that both AI technology experiences and sustainability initiatives do not exert statistically significant indirect effects on consumers’ purchase intentions. This implies that utilitarian value does not play a significant mediating role in the connections between these variables and purchase intentions in the current context.

Discussion

The outcomes of the study have significant implications for businesses in the skincare industry, especially those targeting Gen Z consumers who are interested in Skincare Products. According to the study, the impact of Artificial Intelligence (AI) through Garnier Skin Coach AI experiences reaches beyond utilitarian functionalities, affecting the perception of hedonic value and significantly influencing purchasing intention towards skincare products. Additionally, the study analyzes the business value of sustainability initiatives. Aligning eco-friendly practices with hedonic values allows brands to tap into the ethical considerations of Gen Z consumers.

Influence of AI Technology on Purchase Intentions

This study confirms that Artificial Intelligence (AI) technology has a significant impact on the purchase intentions of Gen Z consumers for Garnier’s skincare products. The accuracy and interactive features of AI experiences play a critical role in this influence, particularly in hedonic values. The accuracy of AI-generated information and interactive AI technology shape consumer perceptions, contributing significantly to overall satisfaction and loyalty.

According to Pine and Gilmore (1998), hedonic experiences that provide pleasure and emotional engagement are fundamental in consumer decision-making processes. In the context of AI technology, accurate information retrieval and interactive engagement with users create nuanced elements that amplify the hedonic appeal. The joy derived from accurate and tailored skincare recommendations and cognitive stimulation experienced through interactive interfaces form the core of these hedonic values (Sundaret al., 2015).

Emotional engagement is also a key driver influencing purchase intentions. Research by Schmitt (1999) indicates that emotionally engaging experiences foster a stronger connection between consumers and brands, leading to increased purchase intentions. The nuanced pleasure and emotional resonance generated by accurate and interactive AI experiences significantly contribute to steering the purchase intentions of the discerning Gen Z demographic. Gen Z consumers, known for their tech-savvy and experiential preferences, tend to prefer AI interactions that evoke joy and emotional resonance. Incorporating these experiential elements into marketing can be a strategic move for skincare brands, particularly Garnier, to stand out in the competitive landscape.

Impact of Sustainability Initiatives on Preferences and Purchase Intentions

The research highlights the significant impact of sustainability initiatives, such as the use of eco-friendly raw materials and cruelty-free product formulations, on the preferences and purchase decisions of Garnier’s Skincare Products among Gen Z consumers. Sustainability initiatives go beyond practical considerations and significantly influence both utilitarian and hedonic values. This indicates a broader shift among Gen Z consumers, suggesting a growing preference for experiences enriched by sustainability aspects. To take advantage of this, businesses should explicitly communicate sustainability aspects in their product messaging. The integration of AI technology and sustainability initiatives emerges as a potent strategy, allowing brands to position themselves as both technologically advanced and socially responsible. AI-driven shopping experiences that incorporate sustainability initiatives have become key factors in influencing purchase decisions, as they provide joy, stimulation, and emotional engagement. By resonating with the experiential and ethical preferences of Gen Z, businesses can differentiate themselves and foster brand loyalty in the evolving skincare market.

Conclusions

The research concludes with two main points. First, Gen Z consumers’ purchase intentions towards Garnier’s skincare products are significantly influenced by AI technology. AI technology’s accuracy and interactive features play a vital role in driving consumer satisfaction and loyalty. Second, Gen Z consumers’ purchase intentions towards Garnier’s skincare products are significantly influenced by sustainability initiatives. These initiatives impact both utilitarian and hedonic values and are a crucial factor in influencing the purchase decisions of consumers in the context of online shopping platforms.

Recommendation

The results offer L’Oréal Indonesia, particularly the Garnier Division, insightful suggestions for improving the Garnier Skin Coach AI and sustainability programs. To overcome the constraints, a comprehensive strategy is suggested that includes improving User-Centric Design, incorporating features aimed at enlightening users, and increasing purchase intention through increased perceived utilitarian value. The suggested tactics are meant to improve the Garnier Skin Coach AI insight experience by making sure it matches customer preferences and encourages a more interesting and customized exchange. Furthermore, improving perceived utilitarian value and boosting buy intentions can be achieved by informing customers about AI’s predictive powers and integrating e-wallet features for easy transactions.

Limitations

This study examines the impact of Artificial Intelligence (AI) technology on the purchase intentions of Generation Z (Gen Z) customers, with a focus on Garnier skincare products in the Indonesian market. It also examines how sustainability issues, such as eco-friendly raw materials and cruelty-free product formulations, influence Gen Z’s skincare product choices and purchase patterns.

Nonetheless, it is critical to recognize the study’s inherent limitations. These limits include things such as sample size constraints in the Indonesian setting. Furthermore, the continuously changing landscape of consumer tastes and technology in the Indonesian cosmetics sector may have an impact on the research findings’ generalizability. It should be noted that the study focuses primarily on Gen Z consumers in Indonesia, and the findings may not be easily extended to other demographic cohorts or geographical regions.