When Yelp was first launched in 2004, it flopped. The company’s initial model of an email-based referral service failed to take off until its creators realized users were taking the platform into their own hands; in lieu of sending one another recommendations via email chain, they began using a “hidden” link at the bottom of the startup’s website to write unsolicited reviews of everything from doctors to restaurants. Yelp 2.0 launched in 2005, and gave those reviews a more visible place to live online — one that emboldened users to feel more authentic stewardship, by way of direct authorship, over hyping up the businesses they enjoyed (or didn’t). The startup bought a whopping 20-million-business database, and amplified a world of reviews outside of the restaurant- critic echo chamber. It was a fresh democratization of information, which was itself one of the more optimistic legacies of the Y2K internet.
It’s been over 20 years since Yelp was founded. At its best, the platform still feels like a place where the voices of hungry, curious diners can share their personal experiences and offer recommendations and tips; at its worst, it feels like a bloated repository for reviews filled with misdirected anger, unreasonable expectations, or else completely fake customer reviews. Initially, unsubstantiated and unreliable reviews were more often written by a person — either someone from the business itself, or paid by the business; now, a fake review can be penned by a paid influencer, a detailed generative AI prompt, or purchased from third-party agencies that will produce fake bot reviews in bulk.
Fake reviews were perhaps inevitable. Bot-driven comments became ubiquitous during an overripe moment in the untamed internet’s lifespan. But how does a restaurant shamed for gamifying its reviews — even if it was, presumably, out of desperation in a tough economic landscape — regain customers’ broken trust?
… The offending restaurant’s foot traffic went down by 8 percent, and dipped even lower around the holidays, a time when would-be diners rely even more on Yelp.
In 2026’s AI-addled digital ecosystem, researchers are starting to understand even more about how paid-for, planted reviews impact the restaurants behind them, and whether trying to gamify platforms to achieve high ratings is worth it at all for businesses. A recently published working paper on the consequences of restaurants flagged by Yelp for fake (planted, incentivized) reviews found that since 2023, more than 4,900 businesses have been flagged with Yelp’s “Consumer Alert” banner, which exists on the penalized business’ page for a 90-day period and lets Yelp users know that the business has been caught peddling fake reviews. In 2020, the fast-casual chain Burgerim was found guilty of planting fake reviews, and received over 200 digital warning banners across its various locations’ Yelp pages.
The impacts of that digital pillory extend far beyond its 90-day duration. Yi Cao is an assistant professor of accounting at George Mason University, and the co-author of the paper; almost immediately, Cao noticed the offending restaurant’s foot traffic went down by 8 percent, and dipped even lower around the holidays, a time when would-be diners rely even more on Yelp. Cao’s team cross-analyzed the business’ performance on Yelp with real-life foot traffic reports (from global location data firm SafeGraph) and credit card transaction data (courtesy of Consumer Edge), and discovered that most of the penalized restaurants struggled to recover to pre-alert traffic even after a year and a half. The stink of chicanery lingered.
While Cao says his team doesn’t know every signal that Yelp uses to identify fake reviews, they do know that it relies “substantially on reports from consumers and business owners,” and that “they also have an algorithm running in the background to screen reviews for AI-generated language — which is a sign of possible review inflation.” At this point, the cost of planting a good review is only getting steeper as the technology capable of detecting it improves. “This is a classical theory of a trade-off between the benefits of manipulating reputation and the risks of getting caught,” Cao explains, “When the risk is relatively low, the business owners decide the incremental benefits are worth the risk.”