satisfaction.

The advent of the digital revolution has fundamentally transformed the nature of economic interactions and consumption patterns. Within this context, e-commerce has emerged as a central pillar of the modern economy, generating profound changes in buying and selling practices. Contemporary competition in this domain is no longer predicated solely on price or product variety; rather, the primary focus has shifted toward the creation of unique and personalized experiences for individual customers. In this trajectory, Artificial Intelligence (AI), as a transformative technology, plays an unparalleled role in redefining interactions between businesses and consumers [1]. Leveraging advanced algorithms such as machine learning, AI is capable of analyzing vast volumes of users’ behavioral data, enabling the identification and prediction of individual preferences. These capabilities are operationalized through intelligent recommender systems, advisory chatbots, and demand forecasting tools, all of which serve the overarching objective of personalizing the shopping experience [2]. At the core of this transformation lies the concept of consumer experience, which is widely recognized as a key determinant of sustainable success in online retail. Customer experience encompasses the totality of interactions with a brand, spanning from the search stage to purchase and after-sales services. Empirical evidence indicates that the quality of this experience exerts a direct influence on customer satisfaction and, consequently, on repurchase intention and loyalty [3].

In this regard, AI has evolved into a critical instrument for enhancing customer experience. By streamlining purchasing processes, delivering contextually relevant recommendations, and providing immediate responses to inquiries, AI not only improves transactional efficiency but also fosters a sense of being understood and valued among customers. Nevertheless, the impact of AI on purchase intention is not necessarily direct and appears to operate through more complex mechanisms. One such mechanism is social media engagement. When customers derive positive experiences from interacting with intelligent systems, they exhibit a greater propensity to share these experiences on social media platforms [4]. Such active engagement—manifested through content creation, commenting, or recommendations—can strengthen customer commitment and influence the decision-making processes of others.

In the Iranian context, examining these relationships is of particular significance. Although Iranian online retailers have taken steps toward adopting AI-driven solutions [5], the industry continues to face multiple challenges, including trust-related concerns (e.g., product quality, timely delivery, and privacy), infrastructural constraints, and a shortage of specialized expertise [6]. These conditions raise a critical question: do existing theoretical models regarding the impact of AI on purchase intention hold within the distinctive context of Iran? Furthermore, through what mechanisms does this influence operate, and what roles do variables such as consumer experience, social media engagement, and customer satisfaction play in this process?

The present study aims to examine the impact of AI on purchase intention by considering the mediating roles of consumer experience, customer satisfaction, and social media engagement among customers of Iranian online retail stores equipped with AI-driven systems. From a theoretical perspective, this research contributes to the enrichment of the indigenous body of knowledge in the field of e-commerce by empirically testing a structured model within the Iranian context. From a practical standpoint, it offers valuable insights for managers and decision-makers in the online retail industry, enabling them to better understand the mechanisms through which AI influences customer behavior. Such understanding can support more targeted investments in AI technologies and inform the development of effective strategies aimed at enhancing customer experience and fostering greater customer loyalty.

Amin [16] examined the role of AI in social media, finding that AI-driven personalization effectively stimulates impulse purchases on these platforms. Personalized recommendations influence consumer behavior more subtly than generic AI labels. Consumer knowledge was not found to mediate this effect, whereas generational differences emerged as a significant moderating factor.

Bilal et al. [15] tested an advanced model in which AI influences consumer experience and social media engagement, both of which, in turn, lead to customer satisfaction and ultimately purchase intention. The study emphasized the strategic use of social media to enhance customer experience and drive engagement.

Ziakis and Vlachopoulou [17] highlighted the capabilities of AI in personalized marketing and the optimization of digital marketing strategies. Their research demonstrated that AI-driven analysis of complex data enhances customer engagement and satisfaction, thereby generating a competitive advantage.

Zhang et al. [18] examined the user experience and social interaction aspects of AI-enabled tools, including chatbots, augmented reality applications, and virtual influencers. Their findings indicated that human-centered and affective design of these tools significantly enhances positive user perceptions and social engagement. Sullivan and Wamba [19] conducted research within a specific context the emerging market of Nepal focusing on organizational capabilities. Their studies demonstrated that, in addition to enhancing customer experience, AI can strengthen an organization’s adaptability to market changes and its innovative performance, while also highlighting challenges such as digital literacy and infrastructural limitations.

Ebrahimi and Ebrahimi [20] investigated the impact of AI on enhancing customer experience through social media marketing. Their findings indicated that the targeted application of AI in social media marketing can lead to significant improvements in customer experience. Technologies such as augmented reality and chatbots act as intermediary tools, facilitating and deepening digital interaction processes.

Pourhasan Anpar and Hosseinzadeh Shahri [21] examined brand revitalization, customer experience, and their effects on service quality and customer loyalty in Tehran. The results revealed that brand revitalization and customer experience management exert significant direct effects on service quality and customer loyalty. Additionally, service quality was found to have a significant impact on customer loyalty in Saman Bank. The study also confirmed the mediating role of service quality in the relationship between brand revitalization, customer experience management, and customer loyalty.

Tajdedi Paskiaby et al. [22] examined the impact of social media marketing on brand loyalty, considering the mediating roles of brand identification and customer satisfaction. The findings indicated that social media marketing, brand identification, and customer satisfaction each have a significant effect on customer brand loyalty. Additionally, social media marketing and brand identification significantly influence customer satisfaction, and social media marketing also has a significant impact on brand identification. Further analysis of the reported path coefficients demonstrated that social media marketing and brand identification exert significant indirect effects on customer loyalty. Similarly, social media marketing was found to have a significant indirect effect on customer satisfaction.

Karaei [23] examined the impact of customer value on satisfaction, repurchase intention, and customer commitment in retail businesses among all customers of small enterprises. The results indicated that customer value has a significant positive effect on customer satisfaction, repurchase intention, and customer commitment.

Bidram and Salehi [24] investigated the impact of AI on purchase intention, considering the mediating roles of perceived value, satisfaction, and consumer engagement on social media. Their findings indicated that AI positively and significantly influences purchase intention by enhancing perceived value and increasing consumer engagement. Moreover, perceived value was identified as a critical mediating variable between AI, satisfaction, and ultimately purchase intention, highlighting its pivotal role in strengthening the effects of AI on consumer behavior.

Momeni-Rad et al. [25], in designing a social marketing model for sustainable consumer behavior, identified five categories of key factors: economic and market management factors (including competitive strategy and influencer marketing), managerial factors (such as customer relationship management), technological factors (including innovation and technological infrastructure), financial and human capital, and advertising and branding factors.

Davoudi and Maleki [26] specifically examined the relationship between customer experience, customer citizenship behavior, and purchase intention in Iran’s retail industry. Their findings indicated that a positive customer experience has a direct and significant effect on both the customer’s willingness to cooperate and recommend the brand (citizenship behavior) and on repurchase intention. The study also highlighted the necessity of considering cultural and geographic factors when generalizing results.

Mirfazli et al. [27] investigated the impact of AI-driven customer experience on e-satisfaction, customer trust in online purchases, and online purchase intention. The results demonstrated that AI-enhanced customer experience positively and significantly affects all three variables, indicating that AI not only improves the quality of customer experience but also exerts a substantial influence on key determinants of online consumer behavior.

Eslami-Farsani [28], through a review of AI applications in customer needs analysis, emphasized the central role of this technology as a driving force for e-commerce development, particularly in developing countries such as Iran. The study also identified contextual factors influencing e-commerce growth, including lifestyle patterns and the availability of payment infrastructures.

Mesrahi [29] specifically examined the impact of AI on subjective norms (perceived social pressure) and purchase intention in online retail. The findings indicated that AI technologies not only exert a direct effect on purchase intention but also indirectly influence customer decisions by reinforcing positive subjective norms.

Rousta et al. [30] confirmed the causal chain linking AI, satisfying customer experiences, and repurchase intention in online retail stores. The study further demonstrated that AI technologies, by creating satisfying experiences, significantly increase the likelihood of customer repurchase.

This study is classified as applied in terms of purpose and descriptive-survey in terms of data collection. Data were gathered using both library research and fieldwork methods. Since the research aims to determine the relationships among study variables, it employs a correlational design. According to Table 1, the study sample comprises 420 online store customers, of whom 59% are male and 41% are female. The dominant age group is 26 to 35 years, representing 44% of respondents. The majority hold a bachelor’s degree (44%) and are employed (58%). Regarding economic status, 44% of participants report a monthly income between 10 and 20 million IRR. Concerning online shopping behavior, most respondents (37%) have 3 to 5 years of online shopping experience, with Digikala being the most popular online store (51%). About 35% of respondents make 1 to 5 online purchases annually. In terms of AI system usage, product recommendation systems are the most widely used, with 46% of respondents reporting frequent use. This demographic and behavioral composition indicates that the sample primarily consists of young, educated, employed individuals who are familiar with emerging technologies, which may influence the study’s findings regarding the impact of AI on online shopping behavior.

The statistical population of this study comprises users of four selected online retail platforms (Digikala, Modiseh, Digistyle, and Technolife) who have experience interacting with the AI systems of these stores and maintain contact with them through social media platforms. A non-probability purposive and convenience sampling method, supplemented by snowball sampling, was employed. The sample size was determined using Cochran’s formula for an infinite population, yielding 384 participants. To account for potential attrition, 420 questionnaires were distributed. The primary data collection instrument was a questionnaire consisting of 24 items measured on a five-point Likert scale. The AI construct was assessed using 8 items adapted from Schepman and Rodway [31]. Consumer experience was measured with 3 items drawn from Agarwal and Singh [32]. Social media engagement was evaluated through 5 items based on Habibi et al. [33] and Laroche et al. [34]. Purchase intention was assessed with 4 items derived from Davis [35], and consumer satisfaction was measured with 4 items adapted from Oghuma et al. [36]. To establish the validity of the measurement instrument, three types of validity were examined: content validity, convergent validity, and discriminant validity. Content validity was ensured by confirming the alignment of measurement indicators with existing literature and was further validated through expert review by faculty members. Convergent validity pertains to the principle that indicators of the same construct should exhibit moderate to high intercorrelations. Following Fornell and Larcker [37], acceptable convergent validity is indicated when the Average Variance Extracted (AVE) exceeds 0.50. Discriminant validity was assessed by comparing the square root of the AVE for each construct with its correlations with other latent constructs (see Table 2). For reflective constructs (latent variables), the square root of the AVE must exceed the construct’s correlation with any other construct in the model. Reliability of the questionnaire was evaluated using two criteria: Cronbach’s alpha coefficient and Composite Reliability (CR). Cronbach’s alpha values for all variables exceeded the minimum acceptable threshold of 0.70. CR, unlike Cronbach’s Alpha which implicitly assumes equal weighting of indicators relies on the actual factor loadings (indicating the strength of the relationship between the latent variable and its observed indicators, ranging from 0 to 1) and thus provides a more accurate measure of reliability. CR (which reflects the degree to which a set of two or more indicators collectively measure a common factor) should exceed 0.70 to demonstrate adequate internal consistency of the construct. The complete results of the reliability and validity analyses are presented in Tables 2 and 3.

Based on the aforementioned discussions and the results obtained from the outputs of SMART-PLS software, Tables 2 and 3 indicate that the measurement instruments demonstrate satisfactory validity—including content, convergent, and discriminant validity—as well as reliability, as evidenced by factor loadings, CR, and Cronbach’s alpha coefficients.

The results presented in Table 4 indicate that the saturated model exhibits superior fit compared to the estimated model. Although the SRMR index for both models falls within the acceptable threshold (below 0.08), its value in the saturated model (0.075) is more favorable than in the estimated model (0.09). Other fit indices, including d_ULS, d_G, and the chi-square statistic, also suggest better performance for the saturated model. Notably, the approximately 100-unit difference in chi-square values between the two models implies the potential omission of some significant relationships in the estimated model. The NFI index, which is below the recommended threshold (0.9) in both models, highlights the need to revise and improve the model structure. Overall, while the model demonstrates a moderate level of fit, the results indicate that the saturated model possesses stronger predictive capability. For the estimated model, achieving an improved fit may require revising the inter-variable relationships and, potentially, increasing the sample size.

The data presented in Table 5, which illustrate the analysis of the research hypotheses, were obtained in accordance with Figs. 2 and 3. The results indicate that the first hypothesis is supported: with a path coefficient of 0.60 and a t-value of 17.31, AI has a significant effect on consumer experience. For the second hypothesis, the analysis shows that AI significantly influences social media engagement, as evidenced by a path coefficient of 0.61 and a t-value of 6.76. The third hypothesis is also supported, with a path coefficient of 0.60 and a t-value of 7.21, indicating that consumer experience has a significant effect on consumer satisfaction. The fourth hypothesis results demonstrate that social media engagement significantly impacts consumer satisfaction, with a path coefficient of 0.61 and a t-value of 6.21. Finally, the fifth hypothesis is confirmed, as consumer satisfaction significantly affects purchase intention, with a path coefficient of 0.59 and a t-value of

Table 6 examines the indirect relationships and the role of mediating variables within the conceptual model of the study. The primary objective of this analysis is to identify the underlying mediating mechanisms through which the independent variables exert their influence on the dependent variables.

The results presented in Table 6 indicate that all indirect effects are significant at the 99% confidence level (p < 0.001, with t-statistics exceeding 1.96). The strongest indirect effect pertains to the impact of AI on customer satisfaction through social media engagement, with a path coefficient of 0.207. Additionally, AI exerts a notable indirect effect on customer satisfaction via customer experience (path coefficient=0.174). In longer pathways, social media engagement and customer experience influence purchase intention through customer satisfaction, with path coefficients of 0.095 and 0.090, respectively. These findings confirm that the mediating variables play a critical role in explaining the relationships among the primary constructs and that meaningful mediation mechanisms exist within this model. The calculated mean path coefficients and standard deviations for each path further support the stability of these relationships across different samples.

AI, as one of the advanced technologies of the digital era, has played a pivotal role in transforming various industries, including e-commerce. By providing intelligent and automated solutions, AI has enhanced online buying and selling processes and substantially improved the consumer experience.

Results concerning Hypothesis 1. the findings indicate that AI has a significant impact on consumer experience in online stores. These results are consistent with those of Rousta et al. [30] and Sullivan and Wamba [19]. Accordingly, AI enhances the relevance and efficiency of the shopping experience by personalizing recommendations based on past behavior. By providing intelligent and real-time support (e.g., chatbots), it reduces purchase barriers. Furthermore, through preference prediction and process optimization, AI significantly increases customer satisfaction and loyalty.

Results concerning Hypothesis 2. the findings indicate that AI has a significant impact on social media engagement in online stores. These results are consistent with those of Ebrahimi and Ebrahimi [20]. Bilal et al. [15], and Amin [16]. Accordingly, AI enhances user interaction with posts by generating and personalizing engaging and relevant content. By analyzing user behavior and optimizing publication timing, it maximizes both reach and engagement. Furthermore, through the identification and amplification of influencers and the facilitation of viral campaigns, AI significantly expands the breadth and depth of social media participation. Results concerning Hypothesis 3. consumer experience has a significant impact on customer satisfaction in online stores. The findings of this study align with those of Pourhasan Anpar and Hosseinzadeh Shahri [21] as well as Ziakis and Vlachopoulou [17]. Accordingly, consumer experience, by fulfilling both functional and emotional expectations, forms the foundational basis of satisfaction. Positive and memorable experiences reinforce feelings of achievement and pleasure, fostering a sense of attachment. Ultimately, a seamless and value-creating experience significantly enhances both the consumer’s mental image of the brand and the intention to repurchase.

Results concerning Hypothesis 4. participation in social media has a significant impact on customer satisfaction in online stores. The findings of this study are consistent with those of Tajdedi Paskiaby et al. [22]. Accordingly, social media engagement satisfies consumers’ emotional and identity-related needs by fostering a sense of belonging and social connection. Through direct interaction with the brand and the receipt of immediate feedback, it enhances feelings of being valued and heard. Furthermore, participation in content creation and social endorsement significantly strengthens brand trust and satisfaction.

Results concerning Hypothesis 5. customer satisfaction influences purchase intention in online stores. The findings of this study align with those of Karaei [23]. Therefore, it can be concluded that customer satisfaction, by fostering positive attitudes and enduring trust toward the brand, provides a robust foundation for purchase decision-making. This satisfaction significantly increases the likelihood of brand repurchase and recommendation to others. Ultimately, a satisfied consumer pays less attention to competitors and may even be willing to pay a premium to maintain a favorable relationship with the brand.

Results concerning Hypothesis 6. the findings indicate that consumer experience significantly moderates the relationship between AI and customer satisfaction in online stores. These results are consistent with those of Mirfazli et al. [27]. Accordingly, consumer experience functions as a key value-transmission mechanism, translating AI’s technical capabilities into tangible and perceived benefits for the customer. AI enhances the quality and richness of this experience through personalization, efficiency, and improved interaction. Indeed, a positive and seamless AI-facilitated experience fosters feelings of control, pleasure, and convenience in the consumer. These positive affective responses directly and strongly enhance overall perceived value, thereby increasing ultimate satisfaction.

Results concerning Hypothesis 7. the findings indicate that social media engagement significantly moderates the relationship between AI and customer satisfaction in online stores. These results are consistent with those of Mirfazli et al. [27]. Accordingly, social media engagement functions as a social amplification channel, extending and deepening the value generated by AI. By producing personalized and interactive content, AI provides the foundation and motivation for active user participation. This engagement—through likes, comments, shares, and content creation—fosters social validation, a sense of belonging to the brand community, and strengthened customer identity. Ultimately, this AI-facilitated social interaction generates deeper, relational satisfaction rather than merely transaction-based satisfaction.

Results concerning Hypothesis 8. the findings indicate that customer satisfaction significantly moderates the relationship between consumer experience and purchase intention in online stores. These results are consistent with those of Saheb et al. [38] and Davoudi and Maleki [26]. Accordingly, satisfaction serves as a critical bridging mechanism between perceptions derived from consumer experience and subsequent behavioral decisions. A positive experience, if not translated into satisfaction, lacks the power to motivate action. Satisfaction crystallizes and reinforces the positive emotions and evaluations arising from the experience, transforming them into a stable attitude toward the brand. It is this enduring positive attitude that meaningfully drives strong motivation and intention for repurchase or recommendation.

Results concerning Hypothesis 9. the findings indicate that customer satisfaction moderates the relationship between social media engagement and purchase intention in online stores. These results are consistent with those of Tajdedi Paskiaby et al. [22]. Accordingly, customer satisfaction functions as the ultimate and valuable outcome of brand-related social interactions, such that mere engagement on social media, without generating satisfaction, has limited impact on purchase intention. Effective social media participation, through fostering a sense of belonging and two-way dialogue, cultivates positive feelings and deeper satisfaction with the brand. This socially derived satisfaction creates stronger emotional commitment and loyalty than purely transactional satisfaction. Ultimately, this type of satisfaction becomes a powerful driver of repurchase and brand advocacy, as it integrates the purchase decision into the consumer’s social identity.

Ultimately, AI functions as a primary driver in enhancing customer experience and fostering social engagement by personalizing the shopping journey and facilitating interactions. These two factors, as key mechanisms, translate the value generated by technology into profound and enduring satisfaction. This resultant satisfaction, in turn, significantly reinforces repurchase intentions and loyalty by shaping positive attitudes and emotional commitment. Therefore, AI not only enhances operational efficiency but also creates a value-generating cycle for achieving sustainable competitive advantage by enriching the emotional and social dimensions of brand interaction. Based on the extensive and comprehensive discussions presented, the following practical recommendations for industry and directions for future research are proposed:

The author appreciates the constructive comments and analytical suggestions provided throughout the review process, which contributed to improving the conceptual structure and interpretation of the empirical findings. Appreciation is also extended to the editorial team for their professional evaluation and guidance during the publication process.

This study was conducted independently and did not receive financial support from governmental, private, or non-profit institutions.

The data analyzed in this research were collected through questionnaires distributed among users of Iranian online shopping platforms. The processed data and measurement instruments used in the study are available from the corresponding author upon reasonable request.