OpenAI rolled out a slew of changes in March, many of them affecting their in-chat shopping experiences.
One of the biggest changes was the launch of product comparison tables and other visuals that pit ChatGPT’s recommended products against each other on price, materials, durability, style and other superlatives, part of an experience revamp that OpenAI announced March 24. The announcement came on the heels of another major change: a reversal on in-chat checkouts.
OpenAI launched “Instant Checkout” in September with great fanfare, but ChatGPT’s in-chat checkout option recorded lackluster sales generation, a Walmart executive told WIRED. The executive said Walmart saw a three-times higher sales conversion rate on products when ChatGPT users clicked out to Walmart’s website than when users were shown the in-chat checkout option. As a remedy, he said ChatGPT now embeds Walmart’s chatbot, Sparky, within its model, something Gemini will adopt as well in the near future.
“We’ve found that the initial version of Instant Checkout did not offer the level of flexibility that we aspire to provide, so we’re allowing merchants to use their own checkout experiences while we focus our efforts on product discovery,” OpenAI stated in its March 24 release.
"Ask a shopping question and ChatGPT will suggest shopping research automatically."
While in-chat checkouts have gone away (at least for now), ChatGPT’s shopping experiences certainly have not. We repeated 21,000 shopper-like prompts across 11 product categories multiple times throughout March on our AI marketing platform, and we found that almost 9 of 10 responses included a shopping experience - a widget with product image carousels, star ratings, buy links and other shopping-related details.
We also observed that a few retailers consistently dominate ChatGPT’s shopping experiences.
The Retailers Winning ChatGPT Shopping
When ChatGPT presents a product in its shopping experience, it typically displays three or more links to go buy that product on a retailers’ websites. All the product categories we reviewed had a wide array of retailers presented in these slots, totaling about 2,000 unique retailers per product category on average.
But three brands popped up again and again: Walmart, eBay and Target.
Of the responses that included a shopping experience, more than 80% included at least one link to buy a product on Walmart.com. eBay boasted a 75% share and Target a 63% share. All other retailers were far behind.
Percent of Responses that Include Retailer
Walmart had the highest count of purchase links in five of our 11 categories and was in the top three for every category but “Luxury & Lifestyle.” Target was top three for seven categories and No. 1 for two categories. eBay, though not the leader in any category, was top five in almost all categories.
Although some other retailers were dominant within their category - BestBuy for electronics, Chewy for pets - no other retailers came close to being as highly visible across a broad range of categories as Walmart, eBay and Target.
We also found that although more than 2,000 retailers made at least one appearance in the shopping responses ChatGPT gave us for each category, the lion’s share of purchase links were held by the top 5-10 retailers.
Percent of Purchase Links by Retailer
Best Price is a Visibility Advantage
When ChatGPT’s shopping experience presents retailers’ purchase links, we found a lot of variation in which retailer’s link is presented first. It isn’t always the cheapest price; it isn’t always a specific retailer, like Walmart; it isn’t always the maker of the product if that maker sells direct-to-consumer (e.g. Bose if ChatGPT is recommending Bose headphones).
Nothing guarantees ChatGPT will give a retailer the top spot in a given response. Price does seem to help, though. We found that about 60% of the time, regardless of product category, the first purchase link is the one with the lowest price for the product ChatGPT recommended.
Percent of Offer Links Where First Offer Shown Is the Cheapest
Having the lowest price is now even more advantageous in ChatGPT’s new product comparison tables. The most consistent feature on which the tables compare products is price, and the tables often highlight the product that is the “cheapest option,” “best for tight budgets,” etc.
The best way to ensure product visibility in ChatGPT, then, appears to be to have a competitively priced product and either be the lowest-price direct-to-consumer purchase option for that product or make sure your product is available in Walmart or Target or listed on eBay. But as March showed, things can change quickly.
Methodology
At Evertune Research we track hundreds of brands across 250 categories. For this analysis, we repeatedly ran 21,000 shopper-like prompts throughout February and March on ChatGPT. Our prompts spanned 11 product categories.
Evertune is the AI marketing platform for Generative Engine Optimization (GEO) that helps brands improve visibility in AI search by analyzing responses at scale and delivering actionable insights. Evertune works with leading brands across all verticals, including Finance, Retail and E-Commerce, Automotive, Pharma, Tech, Travel, Food and Beverage, Entertainment, CPG, and B2B. Founded by early team members of The Trade Desk, Evertune has raised $20M in funding from leading adtech and martech investors. Headquartered in New York City, the company has a growing team of more than 40 employees.
Don't forget to share this post
