# The Most Powerful Influencer Post May Be the One That Isn't Selling Anything 2026-10-03 · Somerset County, New Jersey · Reported Feature A Rutgers Health and Harvard experiment suggests that influence can work most quietly when alcohol is simply part of the scenery, not the product being pitched. A Rutgers Health and Harvard experiment suggests that influence can work most quietly when alcohol is simply part of the scenery, not the product being pitched. --- A couple stands in a kitchen making dinner. They talk, move around the counter and drink something while the camera catches a version of ordinary domestic life polished just enough for social media. In one feed, the drink is wine. In the other, it is cocoa. That small substitution is the heart of a Rutgers Health and Harvard University experiment that is unusually useful because it does not merely ask whether people who see more alcohol online also drink more. The researchers changed the content, randomized who saw it and then measured what happened next. The study, published in JAMA Pediatrics, recruited 2,000 U.S. adults ages 18 to 24 through YouGov. Participants were randomly assigned to one of two simulated social-media feeds and shown 20 Instagram posts from the same lifestyle influencers. One group saw alcohol consumption or favorable alcohol imagery embedded in the posts. The other saw closely matched versions without alcohol. After accounting for factors including prior alcohol use, daily social-media use and earlier exposure to alcohol marketing, the alcohol-exposed group had 73 percent greater odds of reporting a higher desire to drink immediately after viewing the feed. The credibility effect was even larger: participants who regarded the influencers as especially trustworthy, honest and knowledgeable had more than five times the odds of reporting heightened desire to drink. The most interesting part is what the posts were not. They were not obvious commercials telling viewers to buy a bottle. They were not built around a product demonstration or a hard sell. Alcohol simply belonged inside the life being shown. That distinction matters because we have become relatively sophisticated at spotting advertising. We know what a sponsored post looks like. We notice discount codes, brand tags, polished product shots and the sudden enthusiasm for a mattress, protein powder or skin serum that happens to arrive with an affiliate link. We are much worse at spotting modeling. Modeling does not need to make an argument. It does not have to tell us what to buy, what to wear or how to behave. It can work by repeatedly answering a quieter question: what does normal look like here? The wine bottle beside the cutting board may not be the subject of the video. That may be precisely why it matters. When an object or behavior sits comfortably in the background, it can become part of the social scenery rather than something viewers stop to evaluate. The message is not necessarily 'drink this.' The message can be simpler: this is what people like this do while making dinner. Rutgers researcher Jon-Patrick Allem described the experiment as an effort to move beyond association and establish temporal order. That is an important methodological difference. Much of the existing research on social media and behavior can tell us that two things travel together. A randomized design lets researchers ask a stronger question: if we change what people see while keeping much of the surrounding content comparable, does the immediate response change too? In this case, it did. That does not mean a brief exposure caused anyone to pour a drink. The study measured self-reported desire immediately after viewing the posts, not whether participants actually consumed alcohol later that day or whether months of repeated exposure would change drinking behavior. The Rutgers release explicitly notes those limits, and they matter. But those limitations do not make the finding trivial. Desire is exactly the kind of small intermediate response that social environments are capable of shaping. A single kitchen video does not need to rewrite a person's behavior to be meaningful. It only has to make one behavior feel a little more available, familiar or expected. The credibility finding pushes the idea further. Influence is not simply about reach. It is about who has earned the right to appear ordinary in front of us. If viewers see a creator as knowledgeable, trustworthy and honest, the background details of that person's life may carry more weight because they arrive without the defenses that a direct advertisement can trigger. That may be one reason influencer culture is so difficult to describe using old advertising language. Traditional ads usually announce themselves as persuasion. Influencer content often works by collapsing persuasion into observation. Viewers are not just shown products. They are shown mornings, kitchens, vacations, workouts, relationships, purchases, meals, outfits and habits. The commercial and the personal can occupy the same frame. And while this experiment tested alcohol, the larger cultural question travels much farther. It is reasonable to ask how similar modeling might work around spending, clothing, body ideals, parenting, relationships, travel, status and countless other behaviors that appear incidentally inside lifestyle content. The Rutgers study does not prove those effects. It gives us a strong reason to pay closer attention to the mechanism. A fitness creator does not have to tell viewers what a body should look like for a visual standard to emerge. A travel creator does not have to say that a particular kind of vacation is necessary for it to become part of an aspirational life. A parent online does not have to issue instructions about family life for viewers to absorb a template. People learn from what is foregrounded, but they also learn from what is simply there. That is where the research collides with one of the weakest habits in the public conversation about social media: treating screen time as though every minute on a screen is psychologically interchangeable. Twenty minutes of messaging a friend, watching a cooking tutorial, scrolling political outrage and observing a trusted influencer casually drink wine are all counted as screen time. The clock can measure duration. It cannot tell us what the environment was teaching. Allem's team is explicitly interested in that problem. The researchers say future work will separate influencer posts from brand advertising and peer-generated content and follow people over time to better understand whether the immediate effects observed in the experiment connect to behavior. That next step matters because the digital environment does not expose users to one isolated post. It exposes them to patterns. Patterns are how normality gets built. A single glass of wine in a single video is scenery. A hundred versions of the same social cue, delivered by people a viewer likes and trusts, begin to look less like scenery and more like a map of how life is supposed to work. This is the larger lesson hiding inside a study about alcohol. Influence does not always arrive wearing a sales badge. Sometimes the most persuasive thing a creator can do is not make a pitch at all. Sometimes they only have to live on camera. SOURCE NOTES • Rutgers University. Andrew Smith. 'When Influencers Raise a Glass, Young Viewers Want to Join Them.' Feb. 24, 2026. https://www.rutgers.edu/news/when-influencers-raise-glass-young-viewers-want-join-them • Donaldson SI, Russell AM, Allem J. 'Exposure to Alcohol-Related Social Media Content and Desire to Drink Among Young Adults.' JAMA Pediatrics. 2026;180(4):456-458. Published online Feb. 23, 2026. doi:10.1001/jamapediatrics.2025.6335. https://jamanetwork.com/journals/jamapediatrics/fullarticle/2845350 • PubMed record: Exposure to Alcohol-Related Social Media Content and Desire to Drink Among Young Adults. PMID 41729531. https://pubmed.ncbi.nlm.nih.gov/41729531/ RESEARCH NOTE The official Rutgers release is dated Feb. 24, 2026, and the JAMA Pediatrics paper was published online Feb. 23, 2026. The study measured immediate self-reported desire to drink after exposure; it did not establish that participants subsequently consumed alcohol or that a single exposure produces long-term behavioral change. --- ProbleMattic is written and maintained by Matthew Kulcsar, a software engineer, project manager, technologist, platform builder, emergency-services-trained helper, grandfather, and lifelong collector of broken systems, odd behaviors, and useful nonsense.