ChatGPT and Google AI favor large retailers in shopping answers

AI shopping recommendations heavily favor large national retailers over small and local stores, according to a Vaer AI study of nearly half a million AI responses conducted on behalf of Lightspeed Commerce.

National chains accounted for roughly two-thirds of store recommendations when AI answered without web search. Even with live search, large retailers were the top recommendation in 46% to 58% of responses.

Why we care. AI can find and cite smaller retailers without actually recommending them to shoppers. So visibility in AI search sources may not translate into being named when consumers ask where to buy a product.

Big retailers dominated recommendations. With web search turned off, ChatGPT and Google Gemini recommended large national chains 63% to 70% of the time. Small or local independent retailers appeared about 10% of the time.

• When given one large retailer and one smaller store without size labels, the models chose the larger retailer 90% to 94% of the time.
• With live search, large retailers still captured 46% to 58% of top recommendations across ChatGPT, Google AI Mode, and Google AI Overviews.
• Large and small retailers each accounted for roughly 38% of stores cited in AI. But large chains appeared about 2.5 times as often as small retailers in the top recommendation.

Specific products widened the gap. The more specific the shopping query, the more likely AI was to recommend a large retailer. Small stores captured roughly one-third of top recommendations for broad searches, such as toys for a 6-year-old. Their share fell to about 10% when shoppers named a specific product. Meanwhile, large retailers increased from roughly 40% to 60%.

One word changed the results. Asking AI for an “independent” store sharply changed both the recommendations and their sources.

• Adding “independent” more than doubled the share of small and local stores named. Their share rose from roughly one-third to nearly four-fifths in a randomized sample of the study’s prompts. On Google’s platforms, the share of retailer sources from large national chains fell from about 44% with neutral prompts to as low as 9%.
• “Local” and “near me” had much less effect. The study noted that AI systems could count a nearby branch of a national chain as local.

About the data. Tom Wells, co-founder and principal consultant at Vaer AI, conducted the study on behalf of Lightspeed Commerce. The study analyzed 20,000 shopping prompts across the U.S. and Canada. Those prompts produced 200,000 AI responses without web search and 260,000 with search enabled.