Scroll through the reviews for almost any restaurant, and you’ll notice something: The majority are positive. You might assume that reflects genuine satisfaction, or a selection effect where only happy customers bother to write. New research suggests there’s something else going on entirely, and it has less to do with the food than with the person holding the fork. Specifically, it concerns how the reviewer wants to be seen.
Francesca Valsesia, associate professor of marketing and Evert McCabe Endowed Fellow at the Foster School of Business, has been captivated by online self-presentation since long before she became a researcher.
“I was mostly observing other people’s behavior,” she says, “and I always found it fascinating because I would see my friends having these online personas that did not necessarily reflect what I saw offline.”
That early observation became the through line of her career. Valsesia has spent years studying a deceptively simple question: Why do people share what they share online? The answer, she keeps finding, is almost never purely informational. It’s social.
Why We’re Nice First and Honest Later
Her latest paper, forthcoming in the Journal of Consumer Research, brings that lens to one of the most economically consequential arenas of online sharing: consumer reviews.
Researchers have long documented that ratings for a given product tend to decline over time. Early reviewers tend to be enthusiasts; later ones grow more critical. But Valsesia and her co-authors noticed the same pattern playing out within individual reviewers’ own histories. A person’s early reviews, on average, skew more positive. The more reviews someone has written on a platform, the more willing they are to post a critical review.
The data they examined came from Yelp and TripAdvisor, where the pattern held across reviews of businesses and hotels. The question was why.
“When you are new to an online review platform, people really don’t know you,” Valsesia explains. “So it’s all about: What is the first impression I’m going to make? And people are very reluctant to put out negative information because they’re worried about how that is going to reflect on them.”
It’s Not About Being Right. It’s About Being Liked.
Valsesia points to two dimensions that shape how we size up strangers: whether they seem warm, and whether they seem competent. Prior research on online reviews had focused largely on the second of those. Her team found it’s actually the first that drives reviewers to initially hold back negative reviews.
In controlled experiments, participants who imagined reading a profile that posted a negative review early in its history expected that reviewer to be seen as less warm by others, less friendly, less likable. The same negative review posted after a longer history didn’t carry the same penalty. Competence worked differently: Regardless of when you post, readers assume you get more capable over time. But a critical review early in your history reads not as discernment, but as disposition.
One of the study’s most revealing findings came from a test of experienced reviewers switching platforms.
“Even if you have a long history of writing reviews, the moment you go to a new audience, you have to worry about self-presentation again,” Valsesia says. “And then the behavior starts again.”
Francesca Valsesia finds that online reviews aren’t just about products; they’re about perception. New reviewers stay positive to seem likable, only becoming more honest over time.
The Fix: Make It Safer to Be Honest
The stakes extend well beyond any individual reviewer. Platforms already have what researchers call a “positivity problem”: ratings skew high enough that they’re losing their usefulness as a signal.
The team tested several potential fixes. Some failed. Reminding reviewers that negative reviews help other consumers, or that many people post them, didn’t change behavior. The desire to appear warm proved stubborn.
“There is this ingrained desire to be liked,” Valsesia says, “and people have a very hard time letting go of that.”
What worked was changing the structural conditions so that a negative review didn’t threaten warmth perception in the first place. Allowing reviewers to post without tying the rating to their public profile eliminated the hesitation. So did prompting reviewers to write something positive about their experience before submitting a star rating, giving them a visible demonstration of warmth that made the critical rating feel less damning. Valsesia points to Amazon as a partial model: Reviewers can submit a star rating without a written review, and in that case, the rating isn’t associated with their profile.
“Change the way you ask for reviews,” she suggests. “Allow people to put out both the positive and the negative. Making people feel like they’re also putting out positive information could help.”
Francesca Valsesia and A Career Spent Reading Between the Lines
The Journal of Consumer Research paper is one piece of a longer research arc. Across her work, Valsesia has returned again and again to the same underlying tension: People want to share positive information about themselves online, but the moment an audience senses that’s the goal, it backfires. She has studied how posting about experiences rather than purchases signals authenticity in ways material goods don’t, and how tagging a friend in a public post can function as cover, making self-promotion look like a private exchange so the broader audience feels like bystanders rather than targets.
What connects all of it is a set of features Valsesia describes as unique to online communication: Posts are permanent, audiences are large and largely unknown, and there’s no taking back what you’ve put out. That makes the stakes of every sharing decision feel higher, and the social calculations more fraught, than most people consciously realize.
Valsesia brings her findings into her classroom, where students accustomed to mining social media and review platforms for consumer insight are often surprised by its implications.
“What is important is not just what is there, but also what is not in that data,” she tells them. “It’s truncated. And realizing this is not necessarily reflective of 100 percent of consumers’ experience is really important.”
Francesca Valsesia is Associate Professor of Marketing and Evert McCabe Endowed Fellow at the University of Washington Foster School of Business.