Yelp rolls out AI, video to enhance business searches
Yelp is improving its search and review functionality with a number of features based on two of-the-moment internet technologies: artificial intelligence and short-form video. It’s part of the online review platform’s “most significant update in recent years,” according to the company.
“Right now is a really exciting time for us, because we just see low-hanging fruit everywhere,” says Yelp cofounder and CEO Jeremy Stoppelman of the company’s work with large language model-based AI. “We have really trusted, unique content from consumers: millions and millions of reviewed local businesses, and if there’s one thing LLMs are good at, it’s summarizing and understanding content.”
Users often turn to Yelp simply to find nearby goods and services, but they occasionally also turn to the platform to search for and review faraway businesses, such as when they’re planning vacations—or when they’re back from a trip and looking to review spots they visited while out of town. Now, as users type, Yelp’s AI will attempt to intelligently determine whether they’re looking for a place around the corner or across the country—for example, if they enter the name of a celebrated restaurant in another town—as it suggests relevant results, without users needing to specify the name of a city.
“It’s smart enough to know that I’m actually looking for something very specific further away,” says Akhil Kuduvalli Ramesh, VP of consumer product at Yelp.
Yelp’s search engine will also use LLMs to better surface relevant review snippets in results, making it easier for users to figure out which businesses meet their needs without having to click or tap through to full reviews. In demonstrating the tool, Yelp shows a user who types in “tennis courts” into the text box and is then fed information about tennis facilities at nearby parks, or searches for “pixie cuts” and finds out details about hair salons offering the style, pulled from Yelp’s ever-growing database of reviews. The review snippets themselves aren’t a new feature, but Kuduvalli says the feature has been “supercharged” by the use of the new AI trained on Yelp data.
“Yelp’s big differentiator has always been its rich and broad set of customer reviews, and LLMs really allow us to harness the power of those reviews,” says Craig Saldanha, the company’s chief product officer. “Consumers have always had access to them, but LLMs allow us at speed and at scale to parse such vast amounts of data and essentially infer and succinctly summarize very compelling insights and to do that in a way that feels both personal as well as precise.”
Once Yelp results do pop up in certain categories including restaurants and nightlife, they’ll be marked with clickable category tags— like “cafés,” “seafood,” or “Mexican”—that users can tap to find other businesses in the same categories without having to go back to the search box. On individual restaurant and nightlife business pages, a new navigation bar makes it quicker to jump to particular sections of a listing, like the menu, photos, or, of course, the reviews. Yelp has also added AI-powered photo categorization to make it easier to find relevant images on business pages and a sort feature that can help find more recently posted pictures.
“In this way we are helping users, both who know what they want before they open the app and the ones who are not quite sure by guiding them and helping them crystallize their intent,” says Kuduvalli of the new search features.
It’s all part of an ongoing push by Yelp to focus on continued innovation, which Stoppelman says was first set in motion around 2018, when the company switched from a revenue strategy based around deploying sales personnel to one centered more on product evolution and development.
Yelp got its start around 2004 in San Francisco, where Stoppelman recalls, early reviews focused primarily on restaurants in the area.
While restaurants and other businesses sometimes take issue with Yelp’s power to drive traffic and with the contents of individual reviews—as early as 2009, a San Francisco pizzeria decked out staff in T-shirts printed with negative Yelp comments—Yelp had 6.3 million active business locations (along with more than 265 million cumulative reviews) as of the end of last year, according to a February investor presentation.
And though Yelp is still heavily used to find new dining experiences, the company can connect users with a wide variety of businesses. Since around 2016, it’s had an option to let users “Request a Quote” from local companies for services like home repair or moving. Now, it’s rolling out a Yelp Guaranteed program that provides consumers up to $2,500 to cover issues with covered businesses, including unsatisfactory work and property damage, when they’re hired through Request a Quote. (Angi, the contractor-matching service formerly known as Angie’s List, offers a similar guarantee).
Yelp Guaranteed covers businesses selected by Yelp from those advertising on the service, with businesses in the program marked by a badge. It’s currently active in a handful of cities and is on track to expand nationwide this summer, according to the company.
“What it’s saying is that we as Yelp are so confident in our ability to connect you with great local businesses that we will stand behind that connection,” says Stoppelman.
The company is also rolling out tools to enhance the reviewing experience, including letting users post videos up to 12 seconds long alongside traditional text reviews and photos in order to share more vivid details of their experiences, whether that means scenes from a restaurant meal or before-and-after shots of a haircut or home renovation.
“For example, people tend to take videos of food items being served or the ambience of
[a restaurant] inside,” says Kuduvalli. “This is rich information that anybody who’s planning a romantic dinner would want.”
Clients in the paid Yelp Connect program, which lets businesses share updates with consumers, will also be able to post videos of their own, and Kuduvalli says users can expect to soon see videos in search results and on the Yelp home page. That’s something that will likely be familiar to fans of the video reviews and restaurant posts that have lately become popular on platforms like TikTok and Instagram, driving especially younger consumers to try new eateries without necessarily needing to consult traditional text reviews.
The interface for posting written reviews is itself also getting an AI-powered revamp, starting with restaurant reviews, where topic buttons will remind users to consider writing about an eatery’s “food,” “service,” and “ambience,” with each topic automatically turning green and receiving a checkmark as the AI detects it’s addressed in a review. Reviews for other types of businesses will likely receive similar automated guidance in the near future, with appropriate types of topics listed.
“We need to pick the right nudges to push the reviewer towards, in order to scale them across all categories,” Kuduvalli says.
Yelp is not the only platform using AI to enhance restaurant recommendations: OpenAI announced last month a deal with reservation platform OpenTable to enable ChatGPT to make restaurant recommendations. (That’s also a use case Microsoft has suggested for its OpenAI-enabled Bing chat interface.)
Certainly, for Yelp, it’s unlikely this latest AI push is the company’s last.
“With the recent acceleration of LLMs, clearly there’s a lot more capabilities that have been unlocked,” says Stoppelman.
(18)
