Generative artificial intelligence (GenAI) is transforming marketing communication by enabling unprecedented personalization and targeting, automated and autonomous creative production, continuous optimization, and conversational interactions. Whereas prior research has largely focused on firm-level benefits or consumer reactions, this paper argues that GenAI-enabled marketing communication operates within a broader stakeholder ecosystem. We develop a conceptual stakeholder effects model that integrates four stakeholder groups: (1) companies, (2) the marketing communication industry, (3) consumers, and (4) society; our model also distinguishes between intended and extended effects for each stakeholder group. The model highlights that GenAI-driven effects are interconnected through direct and indirect pathways, producing networks of beneficial and detrimental outcomes rather than isolated effects. By theorizing company, consumer, industry, and societal effects and responses as mediators, the model provides a more accurate way of assessing the effects of GenAI-enabled marketing communication. The model contributes to marketing communication theory by extending traditional effects models, providing a new stakeholder perspective of GenAI-enabled marketing communication and the network of effects caused by it as well as a template for future research.
Digital data-driven advertising requires the collection, processing, and storage of personal data, which raises questions on its impact on individuals' perceptions of dataveillance. Yet it remains unclear which specific characteristics of the data drive such perceptions. This study examines how the sensitivity and amount of data used in digital data-driven advertisements influence perceived surveillance and its subsequent impacts on advertising effectiveness and privacy resignation. Through a 2 (data sensitivity: low vs. high) & times; 2 (data amount: single- vs. multiple-type) online experiment (N = 1,256), we find that ads using high-sensitivity data or multiple data types evoke stronger perceived surveillance than those using low-sensitivity data or a single data type, although increasing the amount of data when sensitivity is high does not further increase perceived surveillance. Consequently, perceived surveillance negatively impacts attitudes toward the ad, brand, and platform, but does not significantly influence privacy resignation. These findings extend our understanding about triggers of perceived surveillance by examining two key features in data-driven advertising: data sensitivity and data amount. They also illustrate the unintended consequences of data use on not only the ad and the brand, but also the platform, highlighting the important role of perceived surveillance in shaping consumer responses.
Animal adoption campaigns increasingly rely on visual imagery to attract attention and motivate helping behavior, yet little is known about how animals' emotional displays function as persuasive signals in advertising. Across five studies, this research examines how and why animals' facial expressions shape adoption-related responses, and when these effects attenuate or reverse. Drawing on theories of self-focused decision-making, emotional contagion, and cognitive inference, we propose that adoption is a high-commitment helping context that tends to invite self-focused evaluation, thereby privileging emotional cues that forecast positive personal outcomes. Using archival data and controlled experiments, we demonstrate that animals displaying happy (vs. sad) facial expressions attract more click-through interest and elicit higher adoption intention. We further show that this effect is jointly driven by two parallel psychological processes: positive affective responses and favorable cognitive inferences, which together enhance anticipated self-benefit from adoption. Importantly, we identify two boundary conditions. Conceptually, experimentally shifting attention away from the self toward the animal weakens the happy-expression advantage. Practically, helping mode moderates the effect: happy expressions are more persuasive in adoption contexts, whereas sad expressions are more effective in motivating monetary donation. Together, these findings advance emotional persuasion theory and offer guidance for aligning emotional cues with the motivational structure of helping contexts.
As digital systems increasingly mediate advertising, the field must move beyond message effects to engage with the structural and institutional dynamics that define the advertising ecosystem. This article argues that network analysis, when reimagined beyond its typical application to social media interactions, offers powerful tools for understanding the relationships among brands, consumers, media, influencers, and other intermediaries and stakeholders. Through two case demonstrations, one on influencer marketing and the other on media planning, we illustrate how bipartite projections and inferential network analysis uncover latent relational patterns such as brand competition and influencer heterophily. We further explore alternative strategies for constructing networks from advertising data, including sequence-based, co-occurrence-based, and influence-based approaches. These models enable a shift from descriptive to explanatory analyses and open new pathways for theorizing advertising as a relational, institutional, and dynamic system. Ultimately, we position network analysis as a methodological and theoretical catalyst for structural thinking in advertising research, facilitating deeper engagement with ecosystem-level complexity and interdisciplinary integration with computational social science.
Artificial intelligence (AI) has transformed the advertising industry, yet important research gaps remain in understanding: (a) the cultural dimensions lagging behind AI-enabled advertising, particularly in ad creation and ad placement; (b) consumer literacy and sensemaking of AI-enabled advertising; and (c) how consumers cope with AI-enabled ads and what solutions they recommend. Our research aims to fill these gaps, guided by the culture lag theory and persuasion knowledge model (PKM). We conducted 12 focus group discussions across different generational cohorts (Gen Z, Millennials, Gen X, Baby Boomers) and found four different cultural aspects that lag behind AI-enabled advertising: consumer beliefs; consumer knowledge and literacy; public policy; and public education. Most participants lack the literacy to differentiate AI-enabled ads from traditional ads in absence of AI disclosure, and they employ various coping strategies to address concerns over AI-enabled ad creation and data privacy risks from AI-enabled ad placement. Theoretically, we enrich the investigation leveraging on the theory of culture lag in explaining the discrepancy between technological development and social structures in context of AI and advertising. We also contribute to the application and theoretical understanding of PKM in AI-enabled advertising.
The scientific literature offers crucial insights into ad effectiveness; however, we argue that research on political advertising should extend beyond measuring success in terms of voting preferences by examining its influence on public opinion regarding broader societal issues. In this paper, we used two experimental designs to examine how politically targeted ads influence people's attitudes and their stance on the featured issue (i.e., a sugar tax). More specifically, we investigated how people responded to political ads that matched their views or supported their favorite political party. We deployed two large-scale experiments in the Netherlands (N-study 1 = 1,182 and N-study 2 = 707). Our results indicate that people who are exposed to an ad that aligns with their stance on a given topic or their preferred party will like it more. Ad liking, in turn, is related to the higher importance allocated to the topic shown in the ad. The findings of this work support previous research on the impact of political ads on ad evaluation and perceived issue importance, adding novel insights into changes in issue position. Overall, this study offers new insights into the extent to which targeted communication can be persuasive.
Incidental similarity arises when two individuals share a trivial connection. Marketing research demonstrates that similarity between a consumer and a brand representative (e.g., salesperson, service associate) increases brand beneficial outcomes, specifically attitudes and purchase intentions. We propose that the elicitation of incidental similarity and its implications may not be exclusive to interpersonal relationships. Drawing on brand relationship theory, we contend that trivial connections that arise between a consumer and a brand may also result in similarly favorable outcomes as those observed when studying incidental similarities between people, such as feelings of connection and interest. Across a series of studies, we demonstrate that eliciting consumer-brand incidental similarity through advertisements increases consumers' brand interest, an effect that is mediated by self-brand connection. Further, we demonstrate that the perceived novelty of the brand characteristic moderates this effect. Implications for advertising theory and practice are provided.
Brands increasingly use artificial intelligence (AI) in their advertising efforts. Although extant literature has assessed AI disclosure effects, a research gap exists in disclosures of AI-generated elements in inclusive advertising. Across three main studies, AI-generated plus-size models were used to promote body positivity for a fictitious fashion brand. We recruited female participants because they remain the primary audience for body positivity messaging. This research developed a holistic theoretical model that does the following: (1) confirms the negative impact of disclosing AI-generated models in inclusive advertising (Study 1); (2) uncovers the underlying serial mechanism of social presence and brand hypocrisy (Study 2); and (3) identifies two distinct, theoretically grounded, intervention strategies to mitigate the negative effects. Study 2 found that cause concreteness (a message-based intervention) played a moderating role, whereby concrete messages mitigated the negative effects of reduced social presence on brand hypocrisy. Study 3 revealed that, in ads featuring abstract cause messaging, the addition of an AI transparency disclosure (a source-based intervention) together with the simple AI disclosure enhanced brand evaluations.
In charity advertising, depictions of donor and donee hands are pervasive, yet their relative positions and gestures vary widely. Drawing on the stereotype content model (SCM) and processing fluency theory, this research examines how hand position (donor's hand higher vs. lower), hand gesture (grasping vs. holding), and advertising appeal (competence vs. warmth) jointly influence charitable giving. Across four experiments, we find that when no gesture is shown, a donor's-hand-higher image increases willingness to donate and actual donation amounts (vs. a donor's-hand-lower image). This effect occurs because higher donor hands evoke perceived competence, a key dimension of the SCM. Furthermore, a grasping (vs. holding) gesture enhances advertising persuasiveness when paired with a donor's-hand-higher (vs. lower) image because congruent cues improve processing fluency. The same fluency-based congruence effect emerges when advertising appeal matches the dominant dimension-competence (warmth) appeal with donor's-hand-higher (lower) imagery. Theoretically, this research extends the SCM to visual design by showing how simple hand depictions communicate competence and warmth. Practically, it offers actionable guidance for nonprofits and marketers to combine hand position, gesture, and advertising appeal to create visually and emotionally persuasive charity campaigns.
Transformative advertising research uses institutional theory to position advertising as a site for transformative well-being, highlighting the role of ideology in shaping industry practice and output. Our study into discourses of successful aging in three films addressing the lives of older women demonstrates how film uses cinematography, dialogue, and mise-en-sc & egrave;ne to discursively create film templates of successful aging. Templates provide cultural ideological frameworks for film audiences in their materialized representations of subject positions, thereby creating consent for the conditions and consumption that symbolize those positions. Although film can offer advertising and its audiences potentially transformative ideological templates of aging and gender, our findings reveal that templates rest on existing power hierarchies, limiting their potential use as a tool for transformative well-being in industry that could contribute to individual and societal well-being.
This study applies situational crisis communication theory (SCCT) to examine how deny, diminish, and bolstering strategies employed in advertising influence consumer responses in the context of the Fukushima wastewater release. Focusing on the seafood industry, which has been significantly affected by the disaster, we conducted experiments in China, Japan, and South Korea to test how these strategies influence consumer-based brand reputation (CBR) and attitude toward ads, with individual-level uncertainty avoidance (UA) as a moderator. Whereas the pooled data show that diminish and bolstering significantly improve ad attitude through enhanced CBR (with deny showing no effect), multi-group structural equation modeling reveals substantial cross-country variation: diminish and bolstering are both effective in China, only diminish works in Japan, and only bolstering works in South Korea. In addition, CBR consistently mediates the effects of effective strategies on ad attitude, and UA moderates this process such that as UA increases, the positive indirect effects of diminish and bolstering through CBR weaken. These findings extend SCCT to the context of disaster-related advertising and highlight that strategy effectiveness is culturally contingent. Practically, the study underscores the need for brands to localize crisis strategies based on both psychological and cultural profiles to maintain favorable consumer responses during external crises.
The surge in popularity of glucagon-like peptide-1 receptor agonist (GLP-1) medications, such as Ozempic, for weight loss has been amplified by direct-to-consumer advertising on social media platforms like Meta. With documented risks of such messages to consumer well-being in mind, this study applies a transformative advertising research (TAR) framework together with objectification theory and the health belief model to examine the content and thematic features of prescription weight loss advertisements. Findings from our content analysis (N = 201) and qualitative thematic analysis (N = 109) reveal that although the advertisements do not often target a specific gender, when they do target a gender, they predominantly target women (95.6%). The advertisements also emphasize thin-ideal body types, and prioritize benefits (92.5%) over risks, with few advertisements (12.9%) including important safety information. Thematic analysis identified three key themes: benefits without balance, highlighting the omission of risk information; feminized fixes, reflecting gendered messaging tied to beauty standards; and engineered urgency, promoting quick action through personalized calls to action. These strategies often normalize off-label use for aesthetic goals, raising ethical concerns about health equity and informed decision-making. By integrating TAR, this study emphasizes the need for advertising practices that prioritize consumer well-being, offering insights into the sociocultural and institutional impacts of digital pharmaceutical promotion.
This study explores early adolescents' understanding of and perceptions about the digital advertising system and the types of responses they feel are available as they engage with digital advertising, particularly personalized online ads and influencer marketing, two forms of digital advertising that introduce unique challenges as users must navigate platform dynamics and processes of datafication. To answer these questions, we facilitated an advertising literacy program for 11- and 12-year-old students. We designed the program using what we call the EASY (Ethics of Actors in SYstems) approach, which engages students in role-play activities and reflective discussion, whereby creating an opportunity for students to grapple with the range of actors and the ethical dilemmas that are introduced by and embedded within the multiple components of the digital advertising system. Here we present findings from this program, including how participating students articulated their understandings of and strategies for negotiating the digital advertising system, as well as their ideas for how the system could work more ethically. This research can inform the further development of advertising literacy education that helps adolescents explore how various actors might work toward a transformed digital advertising system.
The prevalence of artificial intelligence (AI)-generated models in advertising raises questions about their impact on consumer perceptions and well-being. This research examines how advertising disclaimers stating the model was "AI-generated" versus "real" influence consumers' evaluation of the model's attractiveness and their attitude toward the advertisement. Drawing on reactance theory and construal level theory, we propose that both psychological reactance and psychological distance shape consumers' evaluations of the model and the ad. In three experimental studies, followed by a qualitative post-hoc analysis, our findings suggest that female consumers perceive the model as less attractive and hold a more negative attitude toward the advertisement when it states the model was AI-generated (vs. real). This effect is explained by the AI disclaimer triggering more psychological reactance and more psychological distance. We extend these findings from Gen Z female consumers to more mature, and often ignored, Gen X female consumers and provide some qualitative insights in support of our findings and the consumer well-being discourse. This study gives recommendations for policymakers and advertisers, contributing to the growing discourse on responsible use of AI in marketing.
