AI shopping tools shift power from review sites to recommendation layers

Google, Reddit and Yelp are pushing consumer reviews into AI-generated shopping and local recommendations, changing how product research reaches buyers. PYMNTS Intelligence found that 48% of online shoppers used AI in their most recent purchase, while AI-referred traffic to U.S. retail sites rose 393% year over year in the first quarter of 2026.

The shift alters the economics of review platforms. If an AI tool can read reviews, forum posts and videos, then deliver a single answer with links or booking options, the platform controlling that answer may capture more of the commercial value than the site that hosted the original discussion.

Google adds community posts to AI search

Google updated AI Overviews and AI Mode in May to include an “Expert Advice” section, according to the company. The feature presents perspectives drawn from Reddit, web forums and social media alongside AI-generated summaries.

TechCrunch reported that AI Mode can show a Reddit quotation with the subreddit identified above traditional links when a user asks a product-related question. That placement puts community content inside Google’s AI response layer, where some users may get enough information without visiting the original forum.

Google already has a data licensing agreement with Reddit worth about $60 million a year, according to the reported terms cited by PYMNTS. The new search presentation extends the commercial importance of that content because Reddit discussions can appear directly in Google’s answer experience.

Reddit and Yelp build their own transaction paths

Reddit is also testing AI shopping inside its own search, according to a February company blog post. The test, available to a limited group of U.S. users, shows product carousels with prices, images and purchase links based on community conversations. Reddit said the trial focuses on electronics, video games, and health and beauty.

Jen Wong, Reddit’s chief operating officer, said at Cannes Lions 2026 that “honest, real-world experiences” have become a more important factor in purchase decisions than professional critics or influencers, according to TechWyse. Reddit has said 71% of surveyed users reported using the platform during the consideration stage of a purchase journey.

Yelp has taken a similar approach in local services. In April, the company launched Yelp Assistant, an AI chatbot that searches across 330 million local business reviews and gives tailored recommendations with explanations, according to Yelp.

The assistant connects with DoorDash, Grubhub, Vagaro, ZocDoc and RepairPal, allowing users to order food, book appointments or schedule repairs from within Yelp’s app. Yelp Chief Executive Jeremy Stoppelman told the Associated Press the chatbot can understand “500 reviews in a second,” compared with a consumer who may read only the first few reviews.

Traffic risk rises for review platforms

Yelp’s reliance on search traffic shows the commercial risk. The company depends on Google for more than 70% of its U.S. web traffic, according to PYMNTS. An in-app assistant gives Yelp a way to keep discovery and transactions closer to its own platform.

The Associated Press reported that an earlier data licensing arrangement between Google and Yelp ended after Google began summarizing restaurant recommendations in ways that gave consumers less reason to visit Yelp. Yelp has accused Google of improperly using its business reviews and favoring Google’s own services.

The broader pattern is visible in shopping research. PYMNTS Intelligence reported that ChatGPT’s share of product research increased from 2% to 30% over two years. For platforms built on user reviews, the contest is no longer only over hosting the content. It is increasingly over which AI system turns that content into a recommendation, a purchase link or a booking.

This story draws on original reporting from PYMNTS.