Artificial intelligence in social media: a

catalyst for impulse buying behavior?

Afiqah Amin

Abstract Afiqah Amin is based at the

Purpose – This study aims to explore the largely under-researched area of user perceptions of artificial School of Business and
intelligence (AI)-driven recommendations and how the visibility of AI-driven labels influences impulsive Economics, Universiti
buying behavior on social media platforms. Addressing a gap in existing literature, it further examines the Brunei Darussalam, Bandar
moderating effects of generational differences and the mediating role of consumer knowledge, providing Seri Begawan, Brunei
new insights into how these factors shape consumer responses to AI-driven content in social media.
Darussalam.
Design/methodology/approach – This study used a quantitative approach, using consumer surveys to
explore the direct and indirect impacts of AI-driven recommendations and the visibility of AI-driven labels
on impulsive buying behavior through social media. A conditional process analysis was conducted to
examine how generational differences act as a moderator and how consumer knowledge of AI serves as
a mediator in these relationships. This analysis integrated these factors into a single framework to provide
a detailed understanding of how consumers respond to AI-driven personalization in social media.
Findings – This study confirms that AI-driven personalization effectively nudges impulsive buying on
social media, with personalized recommendations impacting behavior more subtly than AI labels.
Consumer knowledge does not mediate this effect, while generational differences emerge as a
significant moderator; Millennials are found to be more responsive to both recommendations and labels
in comparison to Gen Z, possibly due to less familiarity with AI technologies present in social media.
Research limitations/implications – This study identifies limitations such as its reliance on consumer
perceptions from a questionnaire rather than direct interactions with AI-driven features. While insightful,
future research should incorporate actual user data, like clickstreams, and include a wider range of
generational cohorts to deepen the understanding of AI’s impact on impulsive buying behavior.
Originality/value – This study bridges gaps in the literature by examining the combined effects of AI-driven
recommendations and visibility of AI-driven labels on impulsive buying behavior, offering new insights into
the role of generational differences and consumer knowledge in AI-based social media marketing.
Keywords Artificial intelligence, Social media, AI-personalization system, Impulsive buying behavior,
Generational differences
Paper type Research paper

Introduction
Artificial intelligence (AI) is fundamentally reshaping the field of social media, where it drives
significant innovation and transformation in how businesses engage with consumers. With
its ability to analyze vast amounts of consumer behavior data, AI allows businesses to
understand their customers better, predict purchasing patterns and personalize marketing
strategies (Benabdelouahed and Dakouan, 2020). As AI continues to enhance the way
companies interact with their audiences, social media platforms have emerged as key
environments where these changes are most visible. Social media has transformed
consumers’ shopping habits by providing new ways to seek, choose and purchase
products (Constantinides, 2014). As consumers’ reliance on social media increases, AI-
driven tools are used to profile potential customers, analyze their habits and determine Received 24 October 2024
Revised 24 December 2024
motivations, enabling companies to offer products or services that precisely align with 30 January 2025
consumers’ needs and expectations. According to a McKinsey report in 2019, 50% of Accepted 19 February 2025

DOI 10.1108/YC-10-2024-2297 © Emerald Publishing Limited, ISSN 1747-3616 j YOUNG CONSUMERS j

 interviewed customers learned about products through social media platforms, and 25%
eventually purchased those products (Bu et al., 2019). This trend may strengthen because
consumers are increasingly dependent on social media (Chen and Shen, 2015; Wang and
Yu, 2017). Therefore, those who do not adopt social media as one of their marketing tools
are missing a chance to reach consumers (Dwivedi et al., 2021).
AI’s role in social media marketing goes beyond traditional strategies. AI systems enhance
marketing efficacy by generating insights from consumers, offering businesses an
unparalleled ability to predict and shape consumer behavior (Al-Ghamdi, 2021; Gkikas and
Theodoridis, 2019). In particular, AI’s use of knowledge-sharing activities is noteworthy, as it
allows AI systems to “learn” from consumer perception, further refining marketing
approaches and driving product and service innovation (Olan et al., 2021). Businesses
nowadays then harness AI’s capacity to analyze consumer preferences, trends and
sentiments, leading to more personalized and targeted marketing campaigns (Albinali and
Hamdan, 2021). Moreover, the effectiveness of these strategies often varies across different
age groups, with generational differences significantly moderating how consumers
perceive and react to AI-driven content. Younger consumers, for example, may display
more resilience to digital marketing tactics, necessitating different approaches to engage
them effectively (Hazzam, 2022).
AI-driven systems not only personalize marketing strategies but also function as subtle
‘nudges’ that influence consumer decision-making. Drawing on nudge theory (Thaler and
Sunstein, 2008), these AI-powered personalizations guide consumers toward certain
behaviors—such as impulse buying—by presenting targeted suggestions, promotions or
reminders for social media users. These nudges, while not overt, shape the decision-
making environment in ways that lead to spontaneous purchases without limiting consumer
choice. However, the impact of these nudges can be further amplified by the consumer’s
level of knowledge about AI, which might alter their responsiveness to AI-driven
recommendations and visibility of AI-driven labels.
AI-driven personalization has been shown to shape consumer behavior effectively, yet
much of the existing research focuses on its technical aspects rather than its marketing
implications. Studies such as Starke and Lee (2022) on conversational user interfaces and
recommender systems (RecSys) and Zhen et al. (2022) on adaptive transfer learning for IoT
recommendations emphasize optimization rather than application in consumer decision-
making. Similarly, while research has explored the intersection of consumer behavior, AI
technologies and knowledge-sharing in influencing purchasing patterns, significant gaps
remain in understanding how AI-driven recommendations and personalization contribute to
impulse buying behavior on social media platforms. Olan et al. (2021) highlighted the
exclusion of key factors such as marketing strategies, consumer buying power and AI tools’
broader utilization, limiting a comprehensive view of AI’s role in influencing consumer
decision-making. Additionally, while Duffett and Charles (2024) examined AI-driven
recommendations among Generation Y and Z, there is a lack of research exploring how
different generational cohorts, such as Baby Boomers and Generation X, perceive and
respond to AI-powered marketing strategies. Given that generational differences influence
digital engagement and trust in AI, understanding these variations is crucial. Furthermore,
as Bag et al. (2022) noted, AI and digital technologies play an increasingly vital role in
customer engagement, yet research is still needed to assess their consequences on
consumer behavior. Addressing these gaps, this study examines user perceptions of AI-
driven recommendations and AI-driven label visibility on impulsive buying behavior while
also exploring how generational differences and consumer knowledge moderate these
effects. The findings aim to provide deeper insights into the marketing implications of AI-
driven strategies and their influence on diverse consumer demographics.
The limitations above point to the need for further investigation into how AI-driven
recommendations across various platforms drive impulse purchases, particularly in the

j YOUNG CONSUMERS j
context of different consumer demographics and consumer knowledge of AI in social
media. Therefore, this study aims to bridge this gap by examining users’ perceptions of AI-
driven recommendations and how the visibility of AI-driven labels influence impulsive
buying behavior on social media platforms. Additionally, it explores how generational
differences and consumer knowledge moderate these effects, providing deeper insights
into the dynamics of AI-driven marketing strategies and their impact on different consumer
demographics.

Theoretical background
Nudge theory
Nudge theory, initially articulated by Thaler and Sunstein, posits that subtle environmental
adjustments can significantly influence decision-making without impinging on freedom of
choice (Thaler and Sunstein, 2008). Extending these foundational ideas, they outlined
various techniques such as anchoring, which involves providing initial manipulative
information as a reference; priming, which uses indirect cues just before decision-making;
and framing, which presents information in a way that highlights either positive or negative
aspects to influence decisions (Thaler and Sunstein, 2008). Echoing these sentiments,
Ivankovic and Engelen (2024) demonstrated that marketing strategies using nudge theory
use environmental contexts—both physical and digital—to effectively shape consumer
behaviors. This is evident in retail settings where strategic product placements and online
interface designs guide user interactions and decision-making processes (Ivankovic and
Engelen, 2024). Furthermore, Koch et al. (2024) provided empirical support through a
digital experiment that highlighted how digital nudges can vary in effectiveness across
different purchasing stages to favorably steer consumer choices.
These newer insights complement existing research by Weinmann et al. (2016), who
discussed the application of design elements in digital environments to guide user choices
subtly, and by other researchers who have identified the effectiveness of nudges in online
retail contexts. For instance, Binu Christeena and Preetha (2019) stated that a customer’s
intention to buy online is influenced by an array of factors which are positioned in a specific
direction but do not compel customers; rather, these factors allow them to make choices in
their own way. Campbell (2022) added that advertising promotions on online platforms are
definitively considered a form of nudging; for example, reminder emails that alert individuals
to products they viewed or added to their online shopping carts but did not purchase,
signaling an incomplete transaction. In retail design, strategies that incorporate nudges
enhance consumer experiences and drive sales, as discussed by Lee and Chu (2023), and
in social media usage, nudges prompt repetitive behaviors, contributing to compulsive
consumption(Lervik-Olsen et al., 2024). The integration of AI in social media platforms,
fostering habitual engagement, further necessitates the exploration of nudge theory in these
settings to understand its role in influencing spontaneous consumer decisions through
cognitive biases and emotional stimuli (Gyulai and Revesz, 2023).
The growing presence of AI integration in social media platforms offers grounds for
investigating how nudge theory can stimulate impulsive purchasing behavior. This
theoretical framework is crucial for understanding the nuanced mechanisms through which
AI-driven social media facilitates spontaneous consumer decisions, highlighting the need
for further scholarly inquiry into this area.

Impulse buying behavior

Impulse buying is defined as an unplanned purchase that is characterized by a
spontaneous decision to buy, typically triggered by external stimuli and emotional
responses (Rook, 1987). Businesses are increasingly turning to Social Networking sites to
attract, connect with and engage customers profitably (Singh et al., 2023). Other than that,

j YOUNG CONSUMERS j
 the use of social media influencers amplifies consumer trust and enhances the perceived
value of brands, which, nurtured through carefully crafted promotional content on social
media platforms, significantly heightens the likelihood of impulse buying in online
environments (Zhang et al., 2022).
Consequently, in recent years, impulsive buying behavior has become increasingly
prominent within the context of social media. Social media platforms serve as a critical
environment that fosters impulsive purchasing decisions among consumers (Yang et al.,
2024). It is the convenience of online shopping, facilitated by social media platforms, that
contributes to impulsive buying behavior. For instance, hedonic browsing of social media
allows the state of immersion and enjoyment while browsing social media influences online
impulsive buying behavior, particularly on platforms like Instagram (Husada et al., 2023).
Esthetic appeal, scarcity promotions and discounted prices stimulate impulse buying
behavior due to positive emotional responses from social media platforms such as
Instagram among Gen-Z (Muhammad et al., 2024). Furthermore, consumers who spend
more time on social media or are closely connected to their mobile devices are more likely
to engage with mobile marketing offers, impacting their purchasing behavior (Oyedele and
Goenner, 2024). Based on the above literature, the seamless integration of shopping
features within social media platforms allows consumers to make purchases with minimal
effort, further promoting impulsive buying tendencies. This is in line with a study by Zafar
et al. (2021) that highlighted how social media marketing uses various cues to stimulate
impulse buying, suggesting that the environment fostered by social networks is highly
conducive to unplanned purchases.
Other cues that have been implemented by social media platforms are leveraging AI to
refine user interactions and strategically personalizing content to align with individual
preferences. This AI-driven mechanism meticulously analyzes extensive user data, ensuring
that the content delivered is highly relevant, thereby significantly enhancing user
engagement and satisfaction. AI algorithms efficiently curate and elevate high-quality
content while simultaneously suppressing less pertinent information, markedly improving
the relevance of users’ feeds (Mohamed et al., 2024). This personalized approach extends
beyond mere convenience; it is a calculated strategy that deepens user engagement and
fosters community building by connecting individuals with friends, groups and content that
resonate with their specific interests (Abdul and Baig, 2024). As social media continues to
integrate sophisticated AI technologies, understanding these dynamics becomes important
for marketers aiming to exploit impulsive buying behaviors. The forthcoming discussion
analyzes the mechanisms of personalization of AI-driven social media content, exploring the
degrees of recommendation these technologies provide, and examines the visibility of AI-
driven labels such as ‘Recommended for you’ in user interfaces.

Hypotheses development
Personalization levels of artificial intelligence-driven recommendations and visibility
of artificial intelligence-driven labels
AI is revolutionizing marketing by improving customer interactions and enhancing process
efficiencies. It enables businesses to precisely target their advertising through AI-driven
algorithms and bots. This precision allows for highly personalized content, which is
demonstrated by predictive analytics and customer segmentation. AI also boosts online
shopping with advanced search engines that quickly deliver relevant results, predicting
consumer preferences (Rashid and Kausik, 2024). Studies suggest that advertisements
tailored to users’ specific interests significantly enhance engagement, such as click-through
rates, and therefore influence consumer behavior, particularly in competitive markets (Ullah
et al., 2023). For instance, a young male frequently browsing gaming-related content is
more likely to encounter advertisements for games, which increases the probability of
engagement and purchase. Furthermore, AI-generated content often outperforms

j YOUNG CONSUMERS j
human-curated content in engagement metrics such as likes, shares and click-through
rates (Manoharan, 2024). This shift to predictive marketing models gives businesses a
competitive edge by ensuring marketing interactions are both timely and relevant
(Kumar et al., 2024). Overall, AI helps organizations deeply understand and respond
to their audience’s needs, refining engagement strategies to drive consumer behavior.
This literature focuses on the personalization capabilities of AI, particularly in
recommendation systems and label visibility, which are critical in enhancing targeted
advertising and influencing impulsive buying.
Research demonstrates that personalization in marketing, particularly through AI-driven
recommendations, significantly influences consumers. By enhancing the relevance of
suggested products (Yoon and Lee, 2021), AI algorithms use sentiment analysis and
consumer preferences, alongside advanced image and text analysis, to identify consumer
trends and feedback. These capabilities facilitate the development of more targeted and
effective marketing strategies (Al-Ghamdi, 2021). Gkikas and Theodoridis (2022)
noted that the customer funnel continues to expand, with consumers increasingly able
to discover desired products through personalized suggestions and predictions
across multiple mediums and channels. For instance, in mobile commerce platforms,
personalized recommendations significantly increase the likelihood of spontaneous
and unplanned purchases among mobile users in an e-commerce platform (Gao and
Li, 2022). Other than that, Ampadu et al. (2023) stressed the significant role of
personalized recommendation systems in enhancing consumer experiences and
fostering loyalty by suggesting products based on previous shopping behaviors,
affecting both prepurchase and post-purchase activities. The integration of AI
technology across various online platforms enhances the capacity to personalize user
experiences and optimize marketing efforts.
Moreover, AI-driven recommendations enhance consumer purchase intentions and
satisfaction within online retail environments. Each of these applications uniquely
contributes to improving the online shopping experience, thereby boosting consumer
purchase intention, satisfaction and overall engagement (Dai and Liu, 2024). Cui et al.
(2022) argued that personalized recommendations enhance consumers’ impulse to
purchase by enriching both the utilitarian value and the flow experience on mobile short
video platforms. When consumers encounter recommendations closely aligned with their
preferences, they are more likely to feel an urgent need to make spontaneous purchases. In
conjunction with AI-driven recommendations, the visibility of AI-driven labels – such as
“Suggested Post,” “Recommended for You,” “Customers Who Bought This Also Bought”
and “Recommended Based on Your Watch History” – plays a crucial role in shaping
consumer perceptions and behaviors. These labels, tailored to align with consumers’
interests, foster a sense of connection, making consumers more likely to act on impulse
when presented with products that resonate with them.
Based on the above literature, AI technologies’ capacity to offer highly personalized content
significantly influences consumer purchasing behaviors, particularly in scenarios involving
impulsive buying through social media mediums. Therefore, recommendations as well as
the visibility of AI-driven labels stand out as a pivotal tool in social media, acting as a
catalyst for impulsive purchases. Echoing this, Dodoo and Wu (2019) emphasized the
effectiveness of personalized advertising through social media, which enhances the
relevance of marketing messages. Drawing on the insights, the following hypotheses are
proposed:
H1. AI-driven recommendations in social media positively influence impulsive buying
behavior among consumers.
H2. The visibility of AI-driven labels in social media positively influences impulsive buying
behavior among consumers.

j YOUNG CONSUMERS j
 Consumer knowledge
The previous section has determined how impulse buying behavior can be influenced by
the personalization of AI content in social media, particularly through AI-driven
recommendations and the visibility of AI-driven labels. This influence is closely tied to
consumer knowledge, which Kolyesnikova et al. (2010) described as ’product-related
knowledge,’ encompassing the levels of experience, expertise and familiarity that
consumers develop with a product. Building on this, Ateke and Didia (2018) defined
consumer knowledge as a multidimensional construct, highlighting the importance of
factors such as experience, expertise and familiarity that consumers rely on when making
informed purchase decisions. Consumers’ understanding of persuasive techniques,
including their ability to recognize marketers’ goals, strategies and methods, affects how
they perceive and respond to targeted advertisements (Dodoo and Wu, 2019; Mohan,
2020). According to Mohan (2020), when consumer awareness about a product’s unique
benefits increases, it can lead to improved product evaluations and a greater likelihood of
purchase. Therefore, consumer knowledge mediates the relationship, clarifying the success
of marketing campaigns. The significance of this mediation is evident in its ability to explain
varying impacts of marketing strategies across different consumer segments, suggesting
that marketing efforts must be tailored to the initial knowledge levels of target consumers to
maximize effectiveness and ensure consumer satisfaction and loyalty.
Beyond enhancing knowledge of products and brands, social media platforms also expose
consumers to AI-driven personalization systems. The effectiveness of AI personalization is
contingent upon the consumer’s knowledge about the products and the AI systems
embedded in the social media platforms themselves. As consumers become more aware of
how AI systems function, such as how it can track online behavior and personalized
marketing efforts, this knowledge could either enhance or reduce the effectiveness of AI-
generated marketing strategies, particularly regarding impulse buying tendencies. Despite
its benefits in enhancing decision-making, consumer knowledge does not always curb
impulsive purchases. Al Mutanafisa and Retnaningsih (2021) observed that although
consumer product knowledge theoretically should enhance the ability to choose wisely,
intense promotional tactics can overpower this knowledge, driving emotional and impulsive
buying behaviors. For example, understanding how AI-driven labels like ’recommended for
you’ or ’trending’ are generated affects whether consumers trust these labels or see them
as manipulative (Dodoo and Wu, 2019). This complex interplay indicates that consumer
knowledge interacts with external marketing efforts in multifaceted ways. Furthermore, Yoon
et al. (2013) identified that higher product knowledge can reduce the perceived value of
recommendation agents, as these consumers may already possess established
preferences or expertise, rendering AI personalization suggestions less valuable or even
redundant.
Consequently, consumers with more knowledge about AI systems may respond differently
to AI-generated marketing stimuli, making AI knowledge a key mediator in consumers’
brand perceptions. This hypothesis posits that as the level of personalization in AI
recommendations and the visibility of AI-driven labels increase, consumers with higher
knowledge about the products are more likely to engage in impulse buying:
H3a. Consumer knowledge mediates the relationship between the personalization level
of AI recommendations in social media and impulse buying behavior.
H3b. Consumer knowledge of AI mediates the relationship between the visibility of AI-
driven labels in social media and impulse buying behavior.

Generational differences
Social media is often used for purchasing purposes and there is a positive relationship
between social media and customer relationships, as well as social media and purchasing

j YOUNG CONSUMERS j
behavior (Shah et al., 2019). Online shopping platforms attract many younger customers
who are more likely to engage in impulsive buying behaviors (Gupta et al., 2023). Past
studies have emphasized the importance that demographic factors, notably age,
significantly influence consumer preferences with brands on digital platforms. For example,
Gen Z uses social media more intensively than Gen Y for purposes like education,
entertainment, shopping and socialization, while both generations show comparable levels
of information seeking (Mude and Undale, 2023). In addition, hedonic shopping motivation
positively impacts online impulse buying among Gen Z consumers (Chetioui and El Bouzidi,
2023). Younger consumers, particularly from Gen Z, exhibit distinct responses to social
media marketing, which diverges markedly from older generations’ engagement patterns
(Hazzam, 2022). This divergence is further evidenced in studies indicating that Millennials,
having grown up fully immersed in digital contexts, display a unique propensity for
engaging with brands via technology, particularly on platforms like Instagram (Bolton et al.,
2013; Casalo  et al., 2020). Conversely, a recent study challenges previous assumptions by
showing no significant age-related differences in the drivers of compulsive social media
use, suggesting that factors like technological nudging and the fear of missing out are
uniformly influential across different age groups (Lervik-Olsen et al., 2024).
In summary, the relationship between personalization of AI-driven social media content and
impulse buying is moderated by generational differences. When taking into account the context of
this study, the influence of AI-driven social media on impulse buying behavior can be measured,
particularly when considering the moderating effects of generational differences. Therefore:
H4a. Generational differences moderate the relationship between the AI-driven
recommendations in social media and impulse buying behavior.
H4b. Generational differences moderate the relationship between the visibility of AI-
driven labels in social media and impulse buying behavior.
Figure 1 presents the conceptual model and the corresponding hypotheses. While the
existing literature provides substantial insights into the role of digital platforms in reshaping
consumer behavior, it also highlights several critical gaps that warrant further investigation.
First, previous studies tend to examine individual AI features, such as recommendations and
label visibility, in isolation, rather than considering their combined influence on consumer
behaviors, particularly impulse buying. For instance, Al-Ghamdi (2021) examines consumer
trends and feedback, while Gkikas and Theodoridis (2022) explore the customer funnel.
However, these studies fail to assess the collective impact of AI-driven personalization on

Figure 1 Conceptual framework

j YOUNG CONSUMERS j
 consumer decision-making. While Binsaeed et al. (2023) discuss the significance of digital
technology in AI adoption and suggest that knowledge sharing plays a pivotal role in shaping
consumer engagement with AI, there remains a notable lack of empirical research on how
consumer knowledge mediates the relationship between AI-driven personalization in social
media and consumer choices. Understanding this mediation is essential for developing more
effective marketing strategies that cater to different levels of consumer awareness. Moreover,
although some studies highlight generational differences in digital behavior and technology
adoption, there is limited research on how these differences moderate the influence of AI-
driven social media nudges. This gap is particularly significant given the varying adoption
rates and media consumption habits among different generational cohorts. Nair and Trehan
(2024) reinforces this point by examining how AI-powered social media analytics contribute to
customer segmentation, further underscoring the influence of generational variations in digital
engagement. Finally, the absence of a comprehensive model that links AI-driven social media
features with impulse buying behavior, while accounting for intervening variables such as
consumer knowledge and demographic moderators, presents a major research gap. The
conceptual framework proposed in this study directly addresses these limitations by
visualizing the hypothesized relationships between AI-driven recommendations, the visibility of
AI-driven labels, consumer knowledge, generational differences and impulse buying behavior.

Methodology
This research adopted a quantitative methodology, primarily using consumer surveys to s
explore the relationships between specific nudge components—namely, AI-driven
recommendations and visibility of AI-driven labels—and their impact on impulsive
purchasing behavior. The study also examined the moderating effects of generational
differences on these relationships. According to Mandolfo and Lamberti (2021), quantitative
approaches are prevalent in impulse buying research, offering essential insights into the
underlying motivations and perceptions of impulsive buying behaviors.
The survey was conducted in Brunei Darussalam, enabling the inclusion of a culturally
homogeneous participant pool which provides a unique context for examining consumer
behavior in a Southeast Asian setting. Conducting the survey in Brunei Darussalam is
particularly significant due to the high internet penetration and social media usage in the
region, which mirrors trends in consumer behavior observed globally but within a localized
framework. Such a focused study setting also enhances the generalizability of the findings
within similar socioeconomic contexts in Southeast Asia, reflecting real-world conditions
more accurately than more diverse and geographically dispersed samples. As highlighted
by Taylor et al. (2018), such diversity ensures a wide applicability of the survey results,
reflecting real-world conditions more accurately than controlled experimental setups could.
Moreover, the naturalistic settings in which these surveys were conducted allows for an
authentic understanding of how consumers perceive and interact with AI elements in their
daily online activities. This study adhered to ethical research practices. At the beginning of
the survey, participants were presented with a statement informing them that proceeding
with the survey would be considered as their consent to participate. This process ensured
that participants were fully aware of their involvement and its voluntary nature. Anonymity
and confidentiality of their responses were maintained throughout the study.
Using a convenience sampling method, the study capitalized on the accessibility of specific
target populations via online platforms, a common approach within the social sciences due
to its practical advantages in quickly gathering data on perceptions and attitudes (Golzar
et al., 2022). The survey, distributed through social media platforms, reached a diverse
sample of 447 respondents, enhancing the robustness and representativeness of the data
collected. Table 1 provides a thorough breakdown of the demographic features of the
respondents.

j YOUNG CONSUMERS j
 Table 1 Demographics information
Category Frequency %

Gender
Male 100 22.4
Female 339 75.8
Prefer not to say 8 1.8
Year of birth
1965–1980 10 2.2
1981–1996 224 50.1
1997–2010 209 46.8
2010–2024 4 0.9
Time spent on social media
Less than 2 h 53 11.9
3 to 4 h 166 37.1
More than 4 h 228 51.0
Source(s): Author’s own work

In the study, age was further categorized into two principal age groups: those under
24 years old and those 24 years and older. 24 years is currently the boundary between the
demographic cohort known as the younger generation, Gen Z (born between 1997 and
later) and the older generation, previous cohorts (e.g., Gen Y and X - Millennials). This
approach was chosen due to the specific demographic characteristics of the sample.
Notably, individuals born between 1965 and 1980 (Gen X), as well as those born
between 2010 and 2024 (Gen Alpha), represent only a minor fraction of the sample –
2.2% and 0.9% respectively. Consequently, these respondents were integrated into
the broader category of older consumers and younger consumers to maintain
statistical robustness and coherence in the analysis, thereby avoiding the potential
bias and variability that might arise from the low representation of these age groups.
This methodological decision ensures a more consolidated and representative
analysis of age-related differences in the context of the study.
The initiation of the survey involved an introduction to AI-driven recommendations, defined
as suggestions made by social media platforms based on users’ perception using AI
algorithms. To illustrate the visibility of AI-driven labels, examples such as “Suggested Post”
and “Recommended for You” were provided. This preparatory section was critical to clarify
key constructs and enhance participants’ understanding, ensuring the reliability of the
responses collected. Furthermore, pilot testing with a diverse group of participants was
crucial in refining the survey’s design, assessing the potential influence of initial
explanations on participants’ responses and making necessary adjustments to minimize
bias (Ruel et al., 2016).
To ensure the rigorous assessment of each construct, the methodology used measurement
scales that were carefully selected and adapted from established research. The scale for
AI-Driven Social Media Recommendations and Visibility of AI-driven labels was based on
the work of El Alam and Bitar (2024), which assesses users’ perceived relevance and trust
in these recommendations and measures participants’ ability to recognize their perception
of the influence of such labels. Consumer Knowledge was measured using items also
adapted from Flynn and Goldsmith (1999), focusing on users’ perceived understanding of
AI’s role in online behavior and confidence in their knowledge compared to peers. Finally,
Impulsive Buying Behavior was gauged using a scale from Beatty and Ferrell (1998),
capturing the frequency of unplanned purchases and other aspects of spontaneous buying
influenced by social media. Each construct was evaluated using a five-point Likert scale,
facilitating the effective quantification of consumer attitudes and behaviors towards AI-

j YOUNG CONSUMERS j
 driven social media interactions and their subsequent impulsive purchases. Prior to
finalizing the questionnaire, pilot testing was conducted with a small but diverse group of
participants. This step was crucial for assessing whether the initial explanations influenced
their understanding or predisposed them to respond in a certain way (Ruel et al., 2016).
Feedback from pilot testing led to adjustments in the wording to further neutralize any
potential bias. This comprehensive approach not only ensures the reliability and validity of
the measurements but also significantly contributes to the existing literature on impulse
buying and AI-driven marketing.
The analysis was conducted using Jamovi software, selected for its robust capabilities in
executing conditional process analysis, which was pivotal for evaluating the moderation
effects hypothesized in our study. Conditional process analysis, which integrates mediation
(explaining how an effect occurs) and moderation (identifying when an effect occurs), offers
a detailed understanding of a research by enabling researchers to ascertain how
mechanisms vary according to context or individual differences (Hayes et al., 2017; Hayes
and Rockwood, 2020), thereby enhancing the comprehension of the determinants of
consumer behavior. In the context of this study, it enables a comprehensive examination of
the direct and indirect effects of the independent variables—AI-driven recommendation
and visibility of AI-driven labels—on impulsive purchasing behavior, with consumer
knowledge acting as the mediator and generational difference as moderator. Furthermore,
the significance of moderation in this study was determined through interaction terms within
a multiple regression framework. Specifically, the interaction between generational
differences and the variables of AI-driven recommendations and visibility of AI-driven labels
was analyzed. Statistical significance was assessed using p-values obtained from these
regression models, with results indicating the presence of moderation effects where the
path coefficients varied significantly between different age groups. Furthermore, these
moderation effects were robustly identified through the utilization of interaction plots, which
visually depict the moderation effect by illustrating the interaction of the independent
variable with the moderator across different levels of the moderator. This graphical
representation aligns with the reported statistical findings, providing a visual confirmation of
the moderated relationships. Such methodological framework as conditional mediation
analysis is robust, revealing complex relationships within the data that simpler analytical
techniques might overlook (Hayes, 2013), which is essential in fields like marketing where
consumer behavior is influenced by multiple intertwined factors.

Results
The present study aims to explore the relationship between AI-driven recommendations and
visibility of AI-driven labels in social media and impulsive buying behavior, focusing on
moderated generational differences and mediated consumer knowledge effects. The
following section starts with the necessary power analysis, followed by the reliability and
validity test, and then proceeds with the main analysis.

Power analysis, reliability and validity test

A power analysis was conducted to determine the required sample size, underpinning the
hypothesis of this study that AI-driven recommendations and visibility of AI-driven labels,
and the moderating effect of generational differences and the mediating effect of consumer
knowledge, impact impulsive purchasing behavior. The power analysis suggested a
sample size of around 64 participants would be necessary to achieve 80% power,
assuming the effect size of 0.5 is used. This relatively modest sample size is primarily
attributed to the large effect size observed, making this study robust and reliable.
Regarding common method bias, Harman’s Single Factor Test was used, which is a widely
recognized method for assessing the extent of such bias in survey data. The analysis was

j YOUNG CONSUMERS j
conducted by performing an exploratory factor analysis on all the items in the data set,
constrained to extract only one factor. The results of this test revealed that a single factor
accounted for 31.1% of the total variance. Given that this value is significantly below the
commonly used threshold of 50%, it suggests that common method bias does not pose a
substantial threat to the validity of this study’s findings. This indicates a robust variance
structure where the measured constructs are distinct and not predominantly influenced by
any single underlying methodological bias.
In terms of reliability, all individual scales show adequate internal consistency, with
Cronbach’s alpha values above 0.7 as well as the overall Cronbach’s alpha for all scales
combined also being acceptable at 0.797. In terms of internal consistency, the AI-driven
social media recommendations variable had good internal consistency with a composite
reliability (CR) of 0.8909 but moderate convergent validity with an average variance
extracted (AVE) of 0.4900. Impulsive buying behavior (CR ¼ 0.9412, AVE ¼ 0.6400) and
consumer knowledge (CR ¼ 0.9000, AVE ¼ 0.5625) exhibited both high reliability and good
convergent validity. However, the visibility of AI-driven labels showed lower convergent
validity, with a CR of 0.7297 and an AVE of 0.3600, suggesting the need for further
refinement of its measurement items. Overall, while most constructs are robust, the visibility
of AI-driven labels requires additional attention to better capture the intended concept.
The inadequate convergent validity observed in the visibility AI-driven label variable
necessitated the assessment of discriminant validity. This evaluation was conducted using
the Fornell–Larcker criterion alongside the Heterotrait-monorait ratio (HTMT) ratio.
According to the Fornell–Larcker criterion, the square root of the AVE for each construct
exceeded its correlations with other constructs, indicating distinctiveness. Specifically, the
constructs demonstrated the following results: AI-Driven Social Media Recommendations
(HAVE ¼ 0.700), Visibility of AI-driven Label (HAVE ¼ 0.600), Impulsive Buying Behavior
(HAVE ¼ 0.800) and Consumer Knowledge (HAVE ¼ 0.750). Additionally, the HTMT ratios
were uniformly below the threshold of 0.85, with the maximum ratio observed at 0.5 between
AI-Driven Social Media Recommendations and Consumer Knowledge. These findings
substantiate the discriminant validity of the constructs, confirming that they are uniquely
representative and effectively delineate different facets of consumer behavior and
perceptions within the context of AI-driven social media. The results bolster the reliability
and validity of the measurement model.

Main analysis
The overall fit of the model is shown in Table 2. The R-squared value of 0.244 indicates that
the model explains 24.4% of the variance in impulsive buying behavior, which is a modest
but reasonable proportion for social science research. The F-statistic (F ¼ 28.4, p < 0.001)
is significant, suggesting that the overall model fits the data well and that the predictors (AI

Table 2 Overall model

Relationship Estimate SE Beta p-value Supported

AI label ! impulsive buying (direct) 0.09 0.05 0.09 0.065 Not supported
AI recommendations ! impulsive buying (direct) 0.5 0.06 0.40 <0.001 Supported
Mediation: AI label ! consumer knowledge ! impulsive buying 0.008 0.02 0.04 0.649 Not supported
Mediation: AI recommendations ! consumer knowledge ! impulsive buying 0.003 0.007 0.02 0.654 Not supported
AI label ! impulsive buying (generation 0, older) 0.20 0.07 0.20 0.003 Supported
AI label ! impulsive buying (generation 1, younger) 0.02 0.07 0.01 0.832 Not supported
AI recommendations ! impulsive buying (generation 0, older) 0.5 0.08 0.7 <0.001 Supported
AI recommendations ! impulsive buying (generation 1, younger) 0.4 0.08 0.6 <0.001 Supported
Note(s): R 2: 0.227; Adjusted R 2: 0.224; F-statistic: 65.19 (p < 0.001); Durbin-Watson: 1.976; Skew: 0.255; Kurtosis: 3.202
Source(s): Author’s own work

j YOUNG CONSUMERS j
 recommendations, visibility of AI-driven labels, generational differences and consumer
knowledge) collectively contribute to explaining impulsive buying behavior. Therefore, while
some individual hypotheses are not supported, the model overall demonstrates a good fit in
capturing the relationships between the variables of interest.
The results of the conditional analysis provide insights into the relationship between AI-
driven recommendations and visibility of AI-driven labels and impulsive buying behavior,
with generational difference acting as a moderator and consumer knowledge as a mediator.
First, Hypothesis 1, which posited that AI-driven recommendations would positively
influence impulsive buying behavior, is supported. The direct effect of AI-driven
recommendations in social media on impulsive buying behavior is significant (Estimate ¼
0.485, p < 0.001), indicating that it increases impulsive purchases. However, Hypothesis 2,
which suggested that the visibility of AI-driven labels would positively influence impulsive
buying, is not supported. The direct effect of AI-driven labels on impulsive buying behavior
is not significant (Estimate ¼ 0.09, p ¼ 0.065), as the p-value is slightly above the 0.05
threshold.
The mediation hypotheses, Hypothesis 3a and Hypothesis 3b, are not supported. Consumer
knowledge does not mediate the relationship between AI-driven recommendations and
impulsive buying behavior, nor does it mediate the relationship between the visibility of AI-
driven labels and impulsive buying behavior, as both indirect effects are insignificant. This
suggests that this factor directly influences impulsive buying without being mediated by
consumer knowledge.
As discussed previously in the methodology chapter, the moderating role of age was
analyzed by categorizing participants into two principal age groups: those under 24 years
old and those 24 years and older. 24 years is currently the boundary between the
demographic cohort known as the younger generation, Gen Z (born between 1997 and
later) and the older generation of previous cohorts (e.g., Gen Y and X - Millennials). For AI-
driven recommendations, the moderation by generation was statistically significant, as
evidenced by the differential path coefficients: the younger generation exhibited a weaker
effect (Estimate ¼ 0.436, p < 0.001) compared to the older generation (Estimate ¼ 0.533,
p < 0.001). This variation in path coefficients supports hypothesis 4a, suggesting that AI-
driven recommendations exert a stronger influence on impulsive buying behavior among
the older generation. The significant difference in coefficients, along with their respective
p-values, confirms that generation acts as a moderator in the influence of AI-driven
recommendations on impulsive buying behavior. As the relationship between AI-driven
recommendation is moderated by generational difference, its significance value across
various levels is further investigated through a slope test. This is demonstrated in
Figure 1 with the simple slope analysis of the interaction.
The interaction graph presented in Figure 2 demonstrates the varying impact of AI-driven
recommendations on impulsive buying across two generational cohorts. Both the older and
younger generations show increased impulsive buying with higher levels of AI-driven
recommendation exposure, as depicted by their positive slopes. However, the continuous
blue line for the older generation displays a steeper slope, indicating a stronger influence of
AI-driven recommendations on their impulsive buying behavior compared to the younger
generation, represented by the less steep orange line.
While visibility of AI-driven labels does not have a strong enough influence on impulsive
buying behavior for the entire population studied, similarly, for the visibility of AI-driven
social media labels, the effect on impulsive buying behavior also varied by generation,
though the pattern differed. While the visibility of AI-driven labels showed no significant
influence on impulsive buying among the younger generation (Estimate ¼ 0.01520, p ¼
0.832), it had a substantial positive effect on the older generation (Estimate ¼ 0.20395, p ¼
0.003). The statistical significance of this difference supports hypothesis 4b, indicating that

j YOUNG CONSUMERS j
 Figure 2 Interaction plot for AI recommendations and generational differences on
impulsive buying behavior

the older generation is more responsive to the visibility of AI-driven labels. This finding
highlights how generational differences can significantly modulate the impact of AI features
on consumer behavior, particularly in how these features are perceived and acted upon
across different age groups. Furthermore, the moderation effects on generational
differences regarding the relationship between visibility of AI-driven labels and impulsive
buying behavior are robustly identified through the utilization of interaction plots, as shown
in Figure 3. The interaction plot demonstrates the moderating effect of generational

Figure 3 Interaction plot for visibility of AI-driven labels and generational differences on
impulsive buying behavior

j YOUNG CONSUMERS j
 differences. The continuous blue line shows the older generation’s impulsive buying
behavior increases with greater visibility of AI-driven labels, demonstrating a significant
responsiveness to AI-driven recommendations. Conversely, the orange line for the younger
generation remains relatively flat, highlighting that AI-driven labels have minimal impact on
their impulsive buying behaviors.

Discussion
On the premise of existing literature, this research develops a conceptualized model that
illuminates the significant role of AI-driven personalization, specifically AI-driven social
media recommendation and visibility of AI-driven labels in social media in exploring
consumer impulsive buying behavior. It also explores the moderating effects of generational
differences on the relationship between AI-driven personalization and impulsive buying
behavior. Furthermore, this study looks at how consumer knowledge mediates the
relationship between AI-driven personalization with impulsive buying behavior. The findings,
analyzed through the lens of nudge theory, offer key insights into how subtle cues
embedded in AI-driven systems influence consumer behavior, particularly in the context of
impulsive purchases.

Artificial intelligence personalization in social media as a nudge for impulsive buying

behavior
The findings of this study provide substantial evidence supporting the application of
nudge theory within the realm of AI-driven personalization in social media. Consistent
with prior research (Ullah et al., 2023; Manoharan, 2024), AI-driven recommendations
significantly influence consumer behavior by simplifying decision-making processes
and enhancing the relevance of the products presented. This study confirms that such
personalized recommendations serve as effective nudges that subtly encourage
consumers towards making impulsive purchases. These findings align with the
principles of nudge theory, which posits that minor modifications in the consumer
environment can significantly influence behavior without constraining choice (Thaler
and Sunstein, 2008).
A key element of this study is how consumers perceive AI-driven recommendations. The
findings suggest that these recommendations are perceived as seamlessly integrated into
the user experience, which enhances their subtlety and effectiveness as nudges.
Consumers reported finding AI-driven recommendations to be highly relevant and visually
appealing, likely due to the sophisticated algorithms that analyze behavioral patterns, such
as browsing history and demographic data (Kumar et al., 2024). Unlike the visibility of AI-
driven labels, which explicitly display messages such as ‘Recommended for You’ or
‘Suggested Post for You’, AI recommendations are seamlessly integrated into social media
feeds, appearing as ordinary posts without overt cues, thus blending more subtly into the
user experience. This technology’s ability to offer highly relevant and visually appealing
suggestions aligns with observations by Zhang et al. (2022), who noted the critical role of
personalized recommendations in triggering impulsive buying behaviors among mobile
commerce users. Moreover, the emotional responses elicited by AI-driven personalization,
which often bypass rational deliberation, underscore the influence of AI on impulsive buying
behaviors. This effect is particularly pronounced in environments rich in information, such as
social media platforms, where personalized recommendations help navigate through the
clutter (Dai and Liu, 2024). This aligns with the study’s findings that AI recommendations
nudge towards impulsive buying behavior, as AI recommendations present options that
seem tailor-made for individual preferences in social media.
Conversely, the findings on perceptions of the visibility of AI-driven labels reveal a different
dynamic. While labels such as ‘Recommended for You’ or ‘Suggested Post for You’ are

j YOUNG CONSUMERS j
explicitly designed to highlight curated content, they are perceived by consumers as overt
and deliberate marketing efforts. This perception diminishes their impact as effective
nudges, as they require conscious recognition and evaluation. While such labels provide
clear product information and promotional cues, they fail to evoke the same spontaneous,
emotional responses associated with personalized recommendations. This aligns with
the theory that effective nudges often work through instinctive, rather than deliberative,
processes (Thaler and Sunstein, 2008). Therefore, consumers in the study perceived these
labels as helpful for facilitating deliberate decision-making but not as triggers for immediate
purchasing behavior. This suggests that the lack of subtlety in these labels may reduce their
ability to function as behavioral nudges.
To enhance the effectiveness of AI-driven labels, it is essential to consider how
consumers perceive their credibility and relevance. Labels could be redesigned to
incorporate emotionally resonant cues, such as trust signals or visually engaging
elements, which may reduce cognitive load and encourage more automatic responses.
For example, labels that emphasize social proof (“X people purchased this today”) or
evoke urgency (“Limited stock available”) may align better with consumers’ instinctive
decision-making processes. By aligning label visibility with the subtlety of AI-driven
recommendations, marketers can bridge the gap between deliberate and impulsive
consumer behaviors.

The non-mediating role of consumer knowledge

The findings of this study challenge the conventional wisdom that consumer knowledge
serves as a mediating factor. Contrary to the theoretical underpinnings suggested by Osei-
Frimpong et al. (2019) and Garrido-Castro et al. (2024), which posited that increased
consumer knowledge fosters a deeper understanding and engagement with brands,
leading to more informed purchasing decisions, this study found no significant mediating
effect of consumer knowledge on impulsive buying behavior triggered by AI-driven
personalizations in social media.
This discrepancy may be illuminated by considering the role of consumer knowledge in a
social media context. While social media platforms indeed enhance consumer knowledge
through continuous engagement, increasing familiarity with brands and products (Cheung
et al., 2023; Frimpong et al., 2019), the same platforms also use sophisticated AI
personalization techniques that can effectively bypass the rational processing typically
enhanced by greater knowledge. Studies such as those by Kolyesnikova et al. (2010) and
Ateke and Didia (2018) described consumer knowledge as a multifaceted construct that
should theoretically empower consumers to make more deliberate choices. However, the
current findings suggest that the persuasive power of AI-driven personalization might
overshadow the mediating role of consumer knowledge.
The lack of mediation by consumer knowledge suggests that impulsive buying, particularly
when triggered by nudges such as AI-driven recommendations and visibility of AI-driven
labels, are highly visible in social media, and thus are not compelling enough as nudges to
alter behavior, particularly in the domain of impulsive purchases. It is possible that as
consumers gain knowledge about AI, they may also develop skepticism or cognitive
resistance toward AI-driven personalization in social media content, thereby reducing the
likelihood of translating this knowledge into impulsive actions. This suggests that emotional
nudges, such as personalized recommendations, are more influential in driving
spontaneous consumer behavior than rational, information-based nudges. Furthermore, the
potential for AI-driven labels to be perceived as manipulative rather than informative (Dodoo
and Wu, 2019) could further explain why consumer knowledge does not mitigate the
impulsive buying behaviors as expected.

j YOUNG CONSUMERS j
 Role of generational differences as a moderator
The findings validate and extend the existing literature on the moderating effects of
generational differences on the impact of AI-driven marketing in social media on
impulsive buying behavior. Prior research indicates significant variations in how
different age groups use and respond to social media, with demographic factors
crucially shaping consumer preferences with digital platforms (Shah et al., 2019; Mude
and Undale, 2023). Particularly, generational cohorts such as Generation Z and
Millennials, display divergent behaviors and engagements with social media marketing
(Hazzam, 2022; Bolton et al., 2013).
Results demonstrated that generational differences significantly influence the effectiveness
of AI-driven nudges. The study classifies the generational differences into two age groups:
those under 24 years old (younger consumers, encompassing of Gen Z and Gen Alpha)
and those aged 24 and above (older consumers, encompassing of Millennials). This
approach is justified by the sample’s demographic profile, predominantly composed of two
groups, with those born between 1981 and 1996 and then 1997 and 2010 representing the
Millennial and Gen Z cohorts accordingly, with minimal representation from other groups
in the study. These results are in line with the nudge theory, which posits that the
efficacy of nudges varies due to individual factors like age, experience and familiarity
with technology. Notably, older consumers demonstrated a pronounced responsiveness
to AI-driven recommendations and labels, possibly due to their relatively lesser
familiarity with or greater trust in AI-driven systems, which tends to simplify decision-
making and enhance credibility, thereby augmenting their propensity for impulsive
purchases (Lervik-Olsen et al., 2024).
Conversely, the younger consumers, adept at navigating digital environments, showed a
muted response to these nudges. This cohort’s familiarity with digital marketing tactics likely
engenders a degree of skepticism or indifference, suggesting that traditional personalized
marketing may be less effective. This aligns with Chetioui and El Bouzidi (2023), who noted
that a hedonic shopping motivation significantly impacts online impulsive buying among
Gen Z consumers, indicating a shift towards more experiential or emotion-driven
purchasing behaviors. Moreover, studies by Hazzam (2022) and others (Bolton et al., 2013;
Casalo et al., 2020) reinforced the unique digital engagement patterns of younger
generations, particularly Millennials and Gen Z, who engage distinctively with brands on
platforms like Instagram and are driven by their immersion in digital contexts from an early
age. However, the findings of Lervik-Olsen et al. (2024) introduce a contrary perspective by
suggesting there are no significant age-related differences in the drivers of compulsive
social media use, indicating a universal influence of factors such as technological nudging
and the fear of missing out across all age groups. This points to a possible underlying
uniformity in certain psychological responses to digital marketing across generations,
challenging the notion that older consumers are uniquely susceptible to such influences.
This dichotomy between generational responses necessitates an understanding of how
different age groups perceive and react to AI-driven nudges. The apparent contradiction in
the literature highlights the complexity of consumer behavior in the digital age, suggesting
that while some behaviors may be universally held, specific responses to marketing tactics
like AI-driven nudges may still vary significantly by age. This indicates a need for
marketers to adopt more tailored, innovative and emotionally engaging approaches,
particularly when targeting younger consumers, who are less influenced by traditional
marketing nudges. The critical integration of these findings with existing literature not
only validates the impact of generational characteristics on digital engagement
strategies but also calls for a broader examination of how these strategies might be
optimized for different demographic cohorts.

j YOUNG CONSUMERS j
Conclusion
The findings offer several implications for marketers aiming to leverage AI-driven tools to
influence consumer behavior. Personalization emerges as a key driver of impulsive buying,
acting as a powerful nudge that can be optimized through sophisticated AI algorithms. By
aligning product recommendations with consumer preferences, AI systems not only reduce
cognitive load but also trigger the emotional impulses that drive spontaneous purchases.
This makes personalized recommendations a valuable tool for increasing impulsive buying,
particularly among older consumers who are more susceptible to these nudges. In contrast,
the limited impact of AI-driven labels suggests that marketers may need to rethink how they
use labeling as a nudge. Labels, while informative, do not generate the same impulsive
response as personalized recommendations. To make labels more effective as nudges,
marketers could experiment with incorporating more emotionally resonant cues or trust
signals that require less cognitive processing. For example, simplifying label content or
embedding them with visual cues that evoke trust might enhance their nudging potential,
particularly for younger generations.
This study acknowledges potential limitations related to the conceptualization provided in
the questionnaire and the focus on perceptions rather than direct interaction with AI-driven
features. While this approach provided valuable insights into consumer attitudes, future
studies could analyze actual user interaction data, such as clickstreams or browsing
behavior, to complement the findings and provide a more comprehensive understanding of
how these AI-driven elements influence impulsive buying behavior. Future research should
aim to include a more balanced representation of all generational cohorts, including Baby
Boomers, Generation X, Millennials and Generation Z, to better explore their unique
characteristics and roles as consumers. This would enable a deeper understanding of how
generational traits, such as digital nativity, technological adoption and purchasing
behaviors, influence the effectiveness of AI-driven recommendations and label visibility.
In conclusion, this study, aiming to bridge the gap in understanding the user’s
perceptions of AI-driven recommendations and the visibility of AI labels, underscores the
pivotal role of AI-driven personalization as a nudge in influencing impulsive buying
behavior on social media platforms. Personalized recommendations simplify decision-
making and evoke emotional reactions, proving particularly potent in encouraging
spontaneous purchases, especially among older consumers. While the study finds that
AI-driven labels in their current form do not act as strong nudges, there is potential for
enhancing their effectiveness through strategic modifications. Additionally, the research
highlights the significance of generational differences, with older consumers showing
greater responsiveness to both recommendations and labeling cues. Marketers are
advised to capitalize on these insights by customizing their AI-driven strategies to cater
to distinct demographic segments, ensuring that their nudges are both subtle and
impactful in steering consumer behavior.

References
Abdul, M. and Baig, A. (2024), “Artificial intelligence, emerging technologies and national security of
Pakistan”, CISS Insight Journal, Vol. 12 No. 1, p. 90.

Al Mutanafisa, T. and Retnaningsih, R. (2021), “The effect of sales promotion and knowledge on impulsive
buying of online platform consumers”, Journal of Consumer Sciences, Vol. 6 No. 1, pp. 77-91.

Albinali, E.A. and Hamdan, A. (2021), “The implementation of artificial intelligence in social media
marketing and its impact on consumer behavior: evidence from Bahrain”, Lecture Notes in Networks and
Systems, Vol. 194, pp. 767-774, doi: 10.1007/978-3-030-69221-6_58.

Al-Ghamdi, L.M. (2021), “Towards adopting AI techniques for monitoring social media activities”,
Sustainable Engineering and Innovation, Vol. 3 No. 1, pp. 15-22, doi: 10.37868/SEI.V3I1.121.

j YOUNG CONSUMERS j
 Ampadu, S., Jiang, Y., Gyamfi, S.A., Debrah, E. and Amankwa, E. (2023), “Perceived value of
recommended product and consumer e-loyalty: an expectation confirmation perspective”, Young
Consumers, Vol. 24 No. 6, pp. 742-766, doi: 10.1108/YC-08-2022-1597/FULL/PDF.
Ateke, B.W. and Didia, J.U.D. (2018), “Consumer knowledge and purchase intention of healthcare
product consumers in Rivers state”, International Journal of Business & Law Research, Vol. 6 No. 1,
pp. 1-7.

Bag, S., Srivastava, G., Bashir, M.M., Kumari, S., Giannakis, M. and Chowdhury, A.H. (2022), “Journey of
customers in this digital era: understanding the role of artificial intelligence technologies in user
engagement and conversion”, Benchmarking: An International Journal, Vol. 29 No. 7, pp. 2074-2098, doi:
10.1108/BIJ-07-2021-0415/FULL/PDF.
Beatty, S.E. and Ferrell, M.E. (1998), “Impulse buying: modeling its precursors”, Journal of Retailing,
Vol. 74 No. 2, pp. 169-191, doi: 10.1016/S0022-4359(99)80092-X.
Benabdelouahed, R. and Dakouan, C. (2020), “The use of artificial intelligence in social media:
opportunities and perspectives”, Expert Journal of Marketing, Vol. 8 No. 1, pp. 82-87.
Binsaeed, R.H., Yousaf, Z., Grigorescu, A., Samoila, A., Chitescu, R.I. and Nassani, A.A. (2023),
“Knowledge sharing key issue for digital technology and artificial intelligence adoption”, Systems, Vol. 11
No. 7, p. 316, doi: 10.3390/SYSTEMS11070316.

Binu Christeena, A. and Preetha, S. (2019), “Effects of nudge and purchase intention in online purchasing
of electronic products”, Indian Journal of Public Health Research & Development, Vol. 10 No. 1, p. 282,
doi: 10.5958/0976-5506.2019.00056.1.

Bolton, R.N., Parasuraman, A., Hoefnagels, A., Migchels, N., Kabadayi, S., Gruber, T., Loureiro, Y.K. and
Solnet, D. (2013), “Understanding generation Y and their use of social media: a review and research
agenda”, Journal of Service Management, Vol. 24 No. 3, pp. 245-267, doi: 10.1108/09564231311326987/
FULL/XML.

Bu, L., Wang, K.W. and Zipser, D. (2019), “China’s digital consumers: 2019 trends”, available at: www.
[Link]/featured-insights/china/china-digital-consumer-trends-in-2019
Campbell, D.R. (2022), “Nudging and social media: the choice architecture of online life”, Giornale Critico
Di Storia Delle Idee, Vol. 2, pp. 93-114, doi: 10.53129/gcsi_02-2022-06.
, L.V., Flavia
Casalo n, C. and Iba
ñez-Sa
 nchez, S. (2020), “Influencers on Instagram: antecedents and
consequences of opinion leadership”, Journal of Business Research, Vol. 117, pp. 510-519, doi: 10.1016/
[Link].2018.07.005.

Chen, J. and Shen, X.L. (2015), “Consumers’ decisions in social commerce context: an empirical
investigation”, Decision Support Systems, Vol. 79, pp. 55-64, doi: 10.1016/[Link].2015.07.012.
Chetioui, Y. and El Bouzidi, L. (2023), “An investigation of the nexus between online impulsive buying and
cognitive dissonance among gen Z shoppers: are female shoppers different?”, Young Consumers,
Vol. 24 No. 4, pp. 406-426, doi: 10.1108/YC-06-2022-1548/FULL/PDF.
Constantinides, E. (2014), “Foundations of social media marketing”, Procedia – Social and Behavioral
Sciences, Vol. 148, pp. 40-57, doi: 10.1016/[Link].2014.07.016.
Cui, Y., Zhu, J. and Liu, Y. (2022), “Exploring the social and systemic influencing factors of mobile short
video applications on the consumer urge to buy impulsively”, Journal of Global Information Management
(JGIM), Vol. 30 No. 1, pp. 1-23, doi: 10.4018/JGIM.301201.
Dai, X. and Liu, Q. (2024), “Impact of artificial intelligence on consumer buying behaviors: study about the
online retail purchase”, Journal of Infrastructure, Policy and Development, Vol. 8 No. 9, p. 7700, doi:
10.24294/JIPD.V8I9.7700.
Dodoo, N.A. and Wu, L. (2019), “Exploring the anteceding impact of personalised social media
advertising on online impulse buying tendency”, International Journal of Internet Marketing and
Advertising, Vol. 13 No. 1, pp. 73-95.
Duffett, R.G. and Charles, J.R. (2024), “Assessing antecedents of google shopping ads intention to
purchase: a multigroup analysis of generation Y and Z”, Young Consumers, Vol. 26 No. 7, pp. 1-21, doi:
10.1108/YC-12-2023-1923/FULL/PDF.
Dwivedi, Y.K., Ismagilova, E., Hughes, D.L., Carlson, J., Filieri, R., Jacobson, J., Jain, V., Karjaluoto, H.,
Kefi, H., Krishen, A.S., Kumar, V., Rahman, M.M., Raman, R., Rauschnabel, P.A., Rowley, J., Salo, J.,
Tran, G.A. and Wang, Y. (2021), “Setting the future of digital and social media marketing research:

j YOUNG CONSUMERS j
perspectives and research propositions”, International Journal of Information Management, Vol. 59,
p. 102168, doi: 10.1016/[Link].2020.102168.

El Alam, R. and Bitar, M.O. (2024), “Unveiling the influence of AI’s product recommendation on consumer
purchase intent: a quantitative study investigating the role of AI product recommendations on consumer
purchase intent”, Mälardalen University, available at: [Link]
diva-67008
Flynn, L.R. and Goldsmith, R.E. (1999), “A short, reliable measure of subjective knowledge”, Journal of
Business Research, Vol. 46 No. 1, pp. 57-66, doi: 10.1016/S0148-2963(98)00057-5.

Gao, L. and Li, J. (2022), “E-Commerce personalized recommendation model based on semantic
sentiment”, Mobile Information Systems, Vol. 2022 No. 1, p. 7246802, doi: 10.1155/2022/7246802.
Garrido-Castro, E., Torres-Peña, F.J., Murgado-Armenteros, E.M. and Torres-Ruiz, F.J. (2024),
“Consumer knowledge in marketing: a critical review and research agenda”, Spanish Journal of
Marketing – ESIC, Vol. 29 No. 1, pp. 65-94, available at: [Link]
FULL/PDF

Gkikas, D.C. and Theodoridis, P.K. (2019), “Artificial intelligence (AI) impact on digital marketing
research”, Springer Proceedings in Business and Economics, Springer, Cham, pp. 1251-1259, doi:
10.1007/978-3-030-12453-3_143.

Gkikas, D.C. and Theodoridis, P.K. (2022), “AI in consumer behavior”, Learning and Analytics in
Intelligent Systems, Vol. 22, pp. 147-176, doi: 10.1007/978-3-030-80571-5_10.
Golzar, J., Noor, S. and Tajik, O. (2022), “Convenience sampling”, International Journal of Education &
Language Studies, Vol. 1 No. 2, pp. 72-77, doi: 10.22034/IJELS.2022.162981.
Gupta, A., Kumar, A. and Melese, E. (2023), “Young consumer engagement at the bottom of pyramid:
applying S-O-R framework in e-commerce context”, Young Consumers, Vol. 24 No. 6, pp. 786-806, doi:
10.1108/YC-06-2022-1542/FULL/PDF.
Gyulai, Z. and Revesz, B. (2023), “Nudging in the digital world: an up-to-date systematic literature
review”, Proceedings of the European Marketing Academy, available at: [Link]
[Link]/pdfs/[Link]
Hayes, A.F. (2013), Introduction to Mediation, Moderation, and Conditional Process Analysis: A
Regression-Based Approach, Guilford Press, New York, NY, available at: [Link]
record/2013-21121-000
Hayes, A.F. and Rockwood, N.J. (2020), “Conditional process analysis: concepts, computation, and advances
in the modeling of the contingencies of mechanisms”, American Behavioral Scientist, Vol. 64 No. 1, pp. 19-54,
doi: 10.1177/0002764219859633/ASSET/IMAGES/LARGE/10.1177_0002764219859633-[Link].
Hayes, A.F., Montoya, A.K. and Rockwood, N.J. (2017), “The analysis of mechanisms and their
contingencies: process versus structural equation modeling”, Australasian Marketing Journal, Vol. 25
No. 1, pp. 76-81, doi: 10.1016/[Link].2017.02.001.
Hazzam, J. (2022), “The moderating role of age on social media marketing activities and customer brand
engagement on Instagram social network”, Young Consumers, Vol. 23 No. 2, pp. 197-212, doi: 10.1108/
YC-03-2021-1296/FULL/PDF.
Husada, J.N., Prawiyadi, C. and Andreani, F. (2023), “The influence of hedonic browsing and flow
experience of Instagram on food and beverages online impulsive buying”, Petra International Journal of
Business Studies, Vol. 6 No. 2, pp. 167-174, doi: 10.9744/petraijbs.6.2.167-174.
Ivankovic, V. and Engelen, B. (2024), “Market nudges and autonomy”, Economics and Philosophy,
Vol. 40 No. 1, pp. 138-165, doi: 10.1017/S0266267122000347.
Koch, C., Pförtsch, W. and Brüggemann, P. (2024), “Guiding customer choices: the impact of digital
nudging on preferred decisions throughout the customer journey”, Springer Proceedings in Business
and Economics, Springer, Cham, pp. 186-193, doi: 10.1007/978-3-031-62135-2_20.
Kolyesnikova, N., Laverie, D.A., Duhan, D.F., Wilcox, J.B. and Dodd, T.H. (2010), “The influence of
product knowledge on purchase venue choice: does knowing more lead from bricks to clicks?”, Supply
Chain Forum: An International Journal, Vol. 11 No. 1, pp. 28-40, doi: 10.1080/16258312.2010.11517223.
Kumar, V., Ashraf, A.R. and Nadeem, W. (2024), “AI-powered marketing: what, where, and how? ”,
International Journal of Information Management, Vol. 77, p. 102783, doi: 10.1016/[Link].2024.102783.

j YOUNG CONSUMERS j
 Lee, J. and Chu, W. (2023), “Nudge design strategies for experiential marketing and retail space
branding”, Journal of Interior Design, Vol. 48 No. 3, pp. 174-190, doi: 10.1177/10717641231177463.
Lervik-Olsen, L., Andreassen, T.W. and Fennis, B.M. (2024), “When enough is not enough: behavioral
and motivational paths to compulsive social media consumption”, European Journal of Marketing, Vol. 58
No. 2, pp. 519-547, doi: 10.1108/EJM-12-2022-0898/FULL/PDF.
Mandolfo, M. and Lamberti, L. (2021), “Past, present, and future of impulse buying research methods: a
systematic literature review”, Frontiers in Psychology, Vol. 12, doi: 10.3389/FPSYG.2021.687404.

Manoharan, A. (2024), “Enhancing audience engagement through ai-powered social media automation”,
World Journal of Advanced Engineering Technology and Sciences, Vol. 2024 No. 02, pp. 150-157, doi:
10.30574/wjaets.2024.11.2.0084.
Mohamed, E.A.S., Osman, M.E. and Mohamed, B.A. (2024), “The impact of artificial intelligence on social
media content”, Journal of Social Sciences, Vol. 20 No. 1, pp. 12-16, doi: 10.3844/JSSP.2024.12.16.
Mohan, G. (2020), “The role of retargeted advertisements in dealing with deflecting customers and its
impact on the online buying process”, International Journal of Internet Marketing and Advertising, Vol. 14
No. 4, pp. 417-432, doi: 10.1504/IJIMA.2020.111050.
Mude, G. and Undale, S. (2023), “Social media usage: a comparison between generation Y and
generation Z in India”, International Journal of E-Business Research (IJEBR), Vol. 19 No. 1, pp. 1-20, doi:
10.4018/IJEBR.317889.
Muhammad, A.S., Adeshola, I. and Isiaku, L. (2024), “A mixed study on the ‘wow’ of impulse purchase on
Instagram: insights from Gen-Z in a collectivistic environment”, Young Consumers, Vol. 25 No. 1,
pp. 128-148, doi: 10.1108/YC-04-2023-1728/FULL/XML.
Nair, N. and Trehan, G. (2024), “Examining the use of Ai-powered social media analytics for target
customer segmentation: a systematic review In retail industry”, Educational Administration: Theory and
Practice, Vol. 30 No. 4, pp. 1798-1805, doi: 10.53555/KUEY.V30I4.1753.

Olan, F., Suklan, J., Arakpogun, E.O. and Robson, A. (2021), “Advancing consumer behavior: the role of
artificial intelligence technologies and knowledge sharing”, IEEE Transactions on Engineering
Management, Vol. 71, doi: 10.1109/TEM.2021.3083536.
Osei-Frimpong, K., McLean, G. and Famiyeh, S. (2020), “Social media brand engagement practices:
examining the role of consumer brand knowledge, social pressure, social relatedness, and brand trust”,
Information Technology and People, Vol. 33 No. 4, pp. 1235-1254, available at: [Link]
ITP-05-2018-0220/FULL/XML
Oyedele, A. and Goenner, E. (2024), “Exploring social influence effect on mobile marketing behaviors of
young consumers in the USA and Mexico”, Young Consumers, Vol. 25 No. 2, pp. 250-272, doi: 10.1108/
YC-05-2023-1747/FULL/PDF.
Rashid, A.B. and Kausik, M.A.K. (2024), “AI revolutionizing industries worldwide: a comprehensive
overview of its diverse applications”, Hybrid Advances, Vol. 7, p. 100277, doi: 10.1016/J.
HYBADV.2024.100277.
Rook, D.W. (1987), “The buying impulse”, Journal of Consumer Research, Vol. 14 No. 2, pp. 189-199, doi:
10.1086/209105.
Ruel, E., Wagner, W.E. and Gillespie, B.J. (2016), “Pretesting and pilot testing”, The Practice of Survey
Research: Theory and Applications, pp. 101-119, doi: 10.4135/9781483391700.
Shah, A.M., Zahoor, S.Z. and Qureshi, I.H. (2019), “Social media and purchasing behavior: a study of the
mediating effect of customer relationships”, Journal of Global Marketing, Vol. 32 No. 2, pp. 93-115, doi:
10.1080/08911762.2018.1497243.
Singh, P., Sharma, B.K., Arora, L. and Bhatt, V. (2023), “Measuring social media impact on impulse buying
behavior”, Cogent Business & Management, Vol. 10 No. 3, doi: 10.1080/23311975.2023.2262371.
Starke, A.D. and Lee, M. (2022), “Unifying recommender systems and conversational user interfaces”,
4th Conference on Conversational User Interfaces (CUI 2022), doi: 10.1145/3543829.3544524.
Taylor, B., Henshall, C., Kenyon, S., Litchfield, I. and Greenfield, S. (2018), “Can rapid approaches to
qualitative analysis deliver timely, valid findings to clinical leaders? A mixed methods study comparing
rapid and thematic analysis”, BMJ Open, Vol. 8 No. 10, doi: 10.1136/BMJOPEN-2017-019993.
Thaler, R.H. and Sunstein, C.R. (2008), Nudge: Improving Decisions about Health, Wealth, and
Happiness, Yale University Press, CT, available at: [Link]

j YOUNG CONSUMERS j
Ullah, I., Boreli, R. and Kanhere, S.S. (2023), “Privacy in targeted advertising on mobile devices: a
survey”, International Journal of Information Security, Vol. 22 No. 3, pp. 647-678, doi: 10.1007/S10207-
022-00655-X/TABLES/2.
Wang, Y. and Yu, C. (2017), “Social interaction-based consumer decision-making model in social
commerce: the role of word of mouth and observational learning”, International Journal of Information
Management, Vol. 37 No. 3, pp. 179-189, doi: 10.1016/[Link].2015.11.005.
Weinmann, M., Schneider, C. and Brocke, J. V (2016), “Digital nudging”, Business & Information Systems
Engineering, Vol. 58 No. 6, pp. 433-436, doi: 10.1007/S12599-016-0453-1/TABLES/2.

Yang, P., Sheng, H., Yang, C. and Feng, Y. (2024), “How social media promotes impulsive buying:
examining the role of customer inspiration”, Industrial Management & Data Systems, Vol. 124 No. 2,
pp. 698-723, doi: 10.1108/IMDS-05-2023-0343/FULL/PDF.

Yoon, N. and Lee, H.K. (2021), “AI recommendation service acceptance: assessing the effects of
perceived empathy and need for cognition”, Journal of Theoretical and Applied Electronic Commerce
Research, Vol. 16 No. 5, pp. 1912-1928, doi: 10.3390/JTAER16050107.

Yoon, V.Y., Hostler, R.E., Guo, Z. and Guimaraes, T. (2013), “Assessing the moderating effect of
consumer product knowledge and online shopping experience on using recommendation agents for
customer loyalty”, Decision Support Systems, Vol. 55 No. 4, pp. 883-893, doi: 10.1016/J.
DSS.2012.12.024.
Zafar, A.U., Qiu, J., Li, Y., Wang, J. and Shahzad, M. (2021), “The impact of social media celebrities’
posts and contextual interactions on impulse buying in social commerce”, Computers in Human
Behavior, Vol. 115, p. 106178, doi: 10.1016/[Link].2019.106178.
Zhang, Z., Zhang, N. and Wang, J. (2022), “The influencing factors on impulse buying behavior of
consumers under the mode of hunger marketing in live commerce”, Sustainability, Vol. 14 No. 4, p. 2122,
doi: 10.3390/SU14042122.
Zhen, Y., Liu, H., Sun, M., Yang, B. and Zhang, P. (2022), “Adaptive preference transfer for personalized
IoT entity recommendation”, Pattern Recognition Letters, Vol. 162, pp. 40-46, doi: 10.1016/J.
PATREC.2022.08.011.

Corresponding author
Afiqah Amin can be contacted at: [Link]@[Link]

For instructions on how to order reprints of this article, please visit our website:
[Link]/licensing/[Link]
Or contact us for further details: permissions@[Link]

j YOUNG CONSUMERS j