The researchers anchor their analysis in a striking natural experiment: the Coca-Cola Company, one of the world’s most recognizable advertisers, has produced two high-profile AI initiatives that met dramatically different fates. The first, the 2023 Create Real Magic platform, invited digital artists and fans to generate their own artwork using the company’s iconic logo through AI image generation. The platform was broadly embraced and enjoyed by consumers. The second, the company’s 2025 AI-generated “Holidays Are Coming” holiday commercial, was met with ridicule on social media and widespread customer criticism, with Forbes reporting on the backlash that erupted when the ad was released.

The difference between these two outcomes, Choi and Magee argue, lies in whether consumers can easily understand why the advertiser used AI in the first place. Create Real Magic offered something that would be difficult, if not impossible, to deliver without the technology: thousands of unique, personalized art pieces created by individual fans around the world. Producing that volume of bespoke creative work through human designers alone would have been impractical. In that context, AI was not a substitute for human creativity but an enabler of an experience that could not otherwise exist. Consumers, the researchers said, could readily grasp the purpose of the technology, and that transparency of purpose defused suspicion.

By contrast, the 2025 holiday commercial used AI to produce animation that human designers could plausibly have created themselves. When AI is deployed unnecessarily, the study finds, consumers can feel upset or unnerved, and in the case of the holiday ad those feelings spilled over into public ridicule. One YouTube commenter delivered a particularly cutting verdict, calling the spot “the most profitable commercial in Pepsi’s history” — a joke that cut to the heart of the problem, framing the ad as a self-inflicted wound. The researchers point to this reaction as evidence that audiences are not simply hostile to AI; they are hostile to AI that appears to replace human effort without offering anything new in return.

Part of the negative response, Magee explained, stems from a well-documented psychological phenomenon known as the uncanny valley effect, in which a machine or depiction that is almost, but not quite, human induces feelings of unease in observers. “We’re basically programmed to look at everything, whether humans or objects, and make snap judgments about them,” Magee said. “We make these quick decisions, but then our conscious mind reflects and thinks, ‘Machines do not feel, but this machine is telling me that it does.'” According to Magee, that is where the conflict originates: when something is almost human but not quite, the brain’s rapid social-cognitive machinery and its reflective reasoning pull in opposite directions, producing the discomfort that audiences experience as eeriness or distrust.

The uncanny valley becomes especially hazardous, the researchers said, when advertisers use AI to try to engender genuine emotion. A generated depiction of a hurt or hungry child, for example, can leave consumers feeling tricked, because the emotional appeal is being manufactured by a system that has no stake in the suffering it portrays. Magee emphasized that marketers need to be very careful to remind people that they are an organization of people, not just a brand and not just a company. “It’s easy to think of an organization as something impersonal, and to counteract that, you need to underline the human aspect,” he said. “For any organization, whether it’s for-profit or nonprofit, the problem comes when something they’ve created feels artificial or contrived. AI can be another instantiation of that disconnect.”

Even when the uncanny valley effect is not in play, the study identifies a second, more strategic pitfall: consumers can still perceive that marketers have taken a “shortcut” by using AI, substituting a cheaper production method for genuine creative effort. Choi describes the remedy as what the researchers call the AI-ad theme match. “A strong match is when AI is an important part of how the campaign works, like if the intention is to give each consumer a personal experience that would otherwise be impossible,” he said. “In that case, AI is not just a cheaper production tool; it’s an important part of the creative idea.” The distinction is subtle but consequential: when the technology is woven into the conceptual core of a campaign, its use feels purposeful; when it is bolted onto a conventional ad, it reads as cost-cutting at the audience’s expense.

This framework leads to practical guidance that Choi believes many marketers are currently ignoring. “Nowadays, many marketers are just saying, ‘Hey, why don’t we use AI?’ but I don’t recommend that,” he said. “Marketers should start with the campaign idea first, and ask ‘Is AI necessary?’ It’s a simple question. If the answer is yes, the reason why should be clearly understood.” In other words, the creative brief should precede the technology decision, not follow it. If AI is essential to delivering the campaign’s promise — personalization at scale, interactivity, or capabilities beyond human production — then its presence is justified and legible to consumers. If it is not, the study suggests, advertisers risk paying a reputational price for a tool that added nothing the audience values.

The researchers also caution against assuming that today’s consumer sensitivities are permanent. Magee noted that the extent to which people feel negatively about AI in advertising may shift as the technology becomes more widespread and familiar. “Right now, because AI is new, it’s better to err on the side of being careful,” he said. “But norms change. You don’t see as many people thinking about CGI in movies anymore. No one gets upset about it like they used to.” The trajectory of computer-generated imagery in film offers a useful analogy: a technique that once provoked complaints about authenticity gradually became an invisible, accepted part of visual storytelling as audiences learned its conventions and creators learned its limits. “Over time, the norms will probably change, and we’ll accept AI just like we accepted CGI, animation and other new technologies,” Magee said.

For an advertising industry racing to incorporate generative tools into everything from holiday spots to interactive platforms, the study’s central message is a tempering one. The technology’s power to create realistic depictions of just about anything does not mean advertisers should lean on it reflexively. Consumer reactions, the research suggests, hinge on a judgment audiences make quickly and intuitively: whether the machine was used to give them something genuinely new, or simply to do cheaply what a person could have done better. Advertisers who can answer that question convincingly — by matching AI to the intent and theme of the campaign — are far more likely to land on the side of Create Real Magic than on the side of a holiday commercial that handed its rival an accidental victory lap.

Subject of Research: Consumer reactions to AI-generated advertising and the match between AI use and ad campaign themes

Article Title: Study explores why some AI ads fail while others succeed

Article References: Study explores why some AI ads fail while others succeed. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: artificial intelligence, advertising, consumer behavior, uncanny valley, Coca-Cola, marketing, Journal of Interactive Advertising, AI-generated ads, integrated marketing communications, CGI, brand backlash, University of Mississippi

Wreetabrata Kar, assistant professor of marketing at Penn State and the study’s lead author, explains that marketers now deploy AI across almost every stage of their work, though content creation remains the most visible application. Teams rely on generative models to draft emails, social media copy, advertisements and product descriptions, and increasingly to produce images and short videos as well. The technology has moved well beyond writing, however. Marketers use AI to brainstorm campaign concepts, research markets and analyze customer data to divide audiences into groups for personalized messaging. Some firms employ predictive models to anticipate what customers will do next and to automate portions of the customer journey. The result is an unprecedented expansion of creative capacity, but also an unprecedented expansion of choices that must somehow be evaluated before a single message reaches a customer.

The core challenge, Kar argues, is that AI does not automatically know what works for a particular company. A message can sound creative and persuasive to a language model, yet that quality guarantees nothing about how a specific firm’s customers will respond. A general-purpose AI tool has never seen a company’s past campaigns and has no knowledge of how its customers reacted to them. That gap between generic fluency and firm-specific effectiveness is precisely what the new framework is designed to close. Rather than treating every piece of generated content as an unknown quantity requiring live experimentation, the approach leverages the rich, often underused record of historical campaign performance that most businesses already possess.

The technical heart of the method lies in how it connects new, untested messages to old, measured ones. Every company has a history of past campaigns, and each of those campaigns can be tied to measurable changes in customer behavior, such as purchases, clicks or conversions. A brand-new message, by contrast, has no history of its own, which is what makes evaluating it so difficult. The researchers solve this by having an AI model read each email message and convert its meaning into a set of numbers, effectively placing every message as a point on a high-dimensional map. Messages that mean similar things land close together on this map. For example, the phrases “Free Shipping on Orders $50+” and “Orders Over $50 Ship Free” end up side by side, while “Break Free from the Rules” lands far away despite sharing the word “free.” Older methods that simply count words cannot distinguish these cases, but the semantic representation generated by AI can.

Once every historical message occupies a position on this map, a newly generated message can draw insights from its nearest neighbors. If a candidate email sits close to past messages that performed well, the model predicts it will probably perform well too. If it sits far from anything the company has ever tried, the model flags that it cannot make a reliable prediction, and in that case the researchers recommend falling back on a traditional test. This is where the approach diverges sharply from conventional A/B testing, in which different versions of a message are sent to real customers and analysts wait to see which variant wins. Live testing takes time, consumes budget and, crucially, exposes some customers to the weaker message along the way. The new framework lets marketers skip the test entirely when their historical data already contains the answer, reserving genuine experimentation for truly novel ideas that lie outside the firm’s accumulated experience.

The framework’s second component addresses the generation side, and it is here that the study delivers one of its most striking findings about the limits of AI self-evaluation. The researchers gave an AI model the building blocks of a retailer’s past emails, elements such as discounts, free shipping offers and clearance sale announcements, and asked it to compose new messages. The model produced a flood of options within seconds, a substantial gain in productivity. But when the researchers asked the same model to select its own five best emails, those picks were predicted to perform poorly. Writing an email and judging it, the study concludes, are fundamentally different jobs. The solution was to split the labor: the AI model writes the candidate messages, while the company’s own historical data, interpreted through the AI-generated semantic map, judges them. Anything too distant from the company’s past experience is set aside, and the surviving candidates are ranked by their predicted performance.

Importantly, the framework does not sideline human marketers; it repositions them at the points of the process where judgment matters most. Marketers choose the building blocks from which new messages are composed, deciding which offers, themes and tones are appropriate for the brand. They also calibrate the acceptable level of risk, determining how far a candidate message may stray from past experience before it must be validated with a real test rather than a model prediction. Finally, the marketer makes the final selection from the shortlist, drawing on deep knowledge of the business and the nuanced context that no model can fully capture. AI performs the heavy lifting on both writing and reading, Kar notes, but people make the decisions, and with far more options on the table than ever before, that human judgment has become more valuable, not less.

The study, titled “EXPRESS: Evaluating Novel Unstructured Treatments with Generative AI: A Causal Prediction Framework,” was conducted using data and statistical analysis and published on 3 August 2026. Kar’s co-authors are Paul Ellickson and Guang Zeng of the University of Rochester and James Reeder of the University of Tennessee, Knoxville. Their combined expertise spans marketing, econometrics and machine learning, reflecting the interdisciplinary nature of a problem that sits at the intersection of causal inference and generative modeling. By framing content evaluation as a causal prediction task grounded in historical treatment effects, the researchers move beyond the superficial quality signals that language models excel at producing and toward measures tied to actual business outcomes.

The implications extend well beyond email marketing. Any organization that generates content at scale, from retail promotions to public health messaging, faces the same evaluation bottleneck: generative models can produce nearly unlimited variations, but live testing each one is expensive, slow and sometimes harmful to the audience receiving inferior variants. A framework that uses a firm’s own campaign history as a training signal for automated screening offers a scalable middle path between untested generation and exhaustive experimentation. It also carries a broader lesson for the adoption of AI across knowledge work. The technology’s greatest value emerges not when it replaces human judgment wholesale, but when it is paired with domain-specific data and human oversight, allowing machines to generate and rank options while people retain control over strategy, risk and the final call.

For marketers contemplating the next wave of AI tools, the message from University Park is clear and pragmatic. Speed and volume are no longer the binding constraints; discernment is. Companies that invest in connecting their generative systems to their own historical performance data, and that preserve meaningful human decision points in the workflow, stand to capture the productivity gains of AI without surrendering the contextual knowledge that makes marketing effective. Those that rely on the models’ own assessments of their output, the study suggests, may find that the most confident machine recommendations are also the least reliable.

Subject of Research: A causal prediction framework using generative AI to evaluate and rank AI-generated marketing content against historical campaign performance data.

Article Title: Q&A: How can marketers use AI effectively — without losing human judgment?

Article References: Q&A: How can marketers use AI effectively — without losing human judgment?. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: artificial intelligence, marketing, generative AI, A/B testing, Journal of Marketing Research, Penn State, email marketing, causal prediction, consumer behavior, human judgment, content evaluation, machine learning

The research team, led by Neda Naqdi Babaee and Saeedeh Shamsaee of Mazandaran University of Science and Technology together with Mohammad Ali Shams of Babol University of Medical Sciences, adopted a qualitative design centered on the perspectives of the audience itself. Rather than analyzing advertisements from a purely rhetorical distance, the researchers asked the people who actually watch these clips what makes them work. Forty Iranian English teachers were purposively selected for semi-structured interviews, a sampling strategy that ensured participants had direct professional familiarity with both language teaching and the social media ecosystem in which teacherpreneurs operate. Each participant viewed and discussed ten highly viewed Instagram clips produced by prominent entrepreneurial teachers, providing a shared reference point that grounded the conversations in concrete, comparable material.

The technical backbone of the analysis was MAXQDA, version 2022, a widely used software package for qualitative data analysis. Interview transcripts were systematically coded, allowing the researchers to identify recurring patterns across forty distinct sets of perceptions and to move from individual impressions toward robust thematic categories. This kind of computer-assisted thematic coding is increasingly standard in applied linguistics and educational research because it imposes discipline on what can otherwise be a subjective interpretive process. By tagging segments of speech and clustering them into themes, the team could trace how often particular persuasive features were mentioned, in what contexts, and with what evaluative weight, producing a structured map of audience judgment rather than a loose collection of anecdotes.

Three factors emerged with particular force. Speaker credibility, emotional appeals, and varied sentence structures were identified as the elements that most significantly influenced the success of the advertisements. Credibility, in this context, was not simply a matter of credentials displayed in a bio. Teachers in the study judged teacherpreneurs by the confidence and professionalism of their delivery, the coherence of their message, and the sense that the person on screen genuinely knew the craft of language teaching. Emotional appeals operated alongside this expertise, with successful clips tapping into learners’ anxieties, aspirations, and desire for progress. Varied sentence structures, meanwhile, kept the delivery dynamic and engaging, preventing the monotony that quickly loses viewers in a fast-paced visual medium.

Underlying these surface features, the thematic analysis identified five main themes that organize how audiences make sense of teacherpreneur advertising: audience engagement strategies, the link between marketing and education, authenticity and effective communication, teacher credibility in the eyes of the audience, and educational effectiveness. The first theme captured the tactics teacherpreneurs use to hook and hold attention, from direct address to relatable classroom scenarios. The second highlighted a tension and a synergy: the best advertisements were seen not as pure sales pitches but as hybrid artifacts in which a genuine teaching moment doubles as a demonstration of the seller’s competence. When a short clip actually teaches something useful, it simultaneously advertises the teacher’s ability to teach, collapsing the traditional boundary between content marketing and instruction.

Authenticity proved to be perhaps the most delicate ingredient. Participants repeatedly emphasized that effective communication depended on an authentic, believable presentation, and that audiences appear remarkably sensitive to performances that feel staged or exaggerated. Clarity of message delivery was repeatedly named as a key factor affecting engagement, alongside the presenter’s credibility and the believability of the overall performance. This finding resonates with broader research on parasocial trust in influencer marketing, but it carries a distinctive educational twist. A language teacher who exaggerates results or oversells a method risks not only a lost sale but damage to the professional trust on which all teaching depends. The teachers in the study appeared to reward restraint and transparency as much as charisma.

The fifth theme, educational effectiveness, connects the study to larger debates about the commercialization of language education. Teacherpreneurship, in which educators build businesses around their personal brands, has grown explosively on platforms like Instagram, particularly in contexts such as Iran where private English education is a major industry and social media offers a direct channel to prospective students. The study suggests that this commercial ecosystem is not inherently at odds with pedagogical quality. On the contrary, the perceptions gathered here indicate that audiences evaluate teacherpreneur advertisements through an educational lens, asking whether the person behind the advertisement can actually deliver the learning outcomes being promised. Marketing that demonstrates real teaching skill earns approval; marketing that substitutes hype for substance invites skepticism.

The methodological choices of the study deserve attention for what they reveal about how such questions can be investigated. Purposive sampling of forty teachers ensured that every participant was positioned to comment meaningfully on both the linguistic and commercial dimensions of the clips. Semi-structured interviews allowed the researchers to follow unexpected lines of thought while still covering a consistent set of prompts across participants, and the use of ten shared stimulus clips reduced the variability that plagues studies of social media content. The thematic coding in MAXQDA 2022 then transformed a large volume of conversational data into five coherent themes, a pipeline that other researchers studying educational influencers on TikTok, YouTube, or other platforms could readily adapt. The study was approved by the Mazandaran University of Science and Technology Ethics Committee, with informed consent obtained from all participants, and the authors report no competing interests and no external funding.

For teacherpreneurs themselves, the practical implications are direct. Credibility must be built and signaled, not merely claimed. Emotional connection matters, but it works best when anchored to genuine pedagogical substance. Delivery should be varied and clear, avoiding both monotony and theatrical excess. And authenticity, that most overused of marketing buzzwords, emerges here as a measurable audience expectation: viewers judge whether the presentation is believable, and their willingness to engage follows from that judgment. For teacher educators and researchers, the study opens questions about how prospective teachers should be prepared for a professional landscape in which self-presentation and marketing literacy are increasingly part of the job. For platforms and regulators, it hints at the standards by which educational advertising is informally policed by audiences themselves.

The broader significance of the work lies in its reframing of social media advertising as a site of educational perception. The teachers who watched these Instagram clips were not passive consumers but active evaluators, applying professional criteria to commercial messages. Their collective verdict, distilled through qualitative analysis, is that persuasion in educational marketing rests on a tripod of credibility, emotion, and craft, with authenticity as the load-bearing joint between them. As language teaching continues its migration toward digital and hybrid formats, understanding how teachers and learners judge the people who sell it online will only grow in importance. This study provides an early, carefully documented map of that judgment, drawn directly from the professionals who know best what good teaching looks like, even when it arrives in a sixty-second clip between advertisements for everything else.

Subject of Research: EFL teachers' perceptions of teacherpreneurs' persuasive techniques in Instagram advertisements

Article Title: EFL teachers’ perceptions of teacherpreneurs’ persuasive techniques on Instagram

Article References: Naqdi Babaee, N., Shamsaee, S., & Shams, M. A. (2026). EFL teachers’ perceptions of teacherpreneurs’ persuasive techniques on Instagram. Discover Psychology. https://doi.org/10.1007/s44202-026-00899-x

Image Credits: AI Generated

DOI: 10.1007/s44202-026-00899-x

Keywords: teacherpreneurs, TEFL, Instagram, persuasive techniques, marketing, EFL teachers, credibility, emotional appeals, authenticity, social media advertising, qualitative research, thematic analysis