A recent report from D.A. Davidson & Co. suggests that while the rise of artificial intelligence-powered shopping agents could significantly boost online sales, it may also pose increased risks for many retailers. The firm pointed to the September 29 launch of OpenAI’s “Instant Checkout” for ChatGPT users as the dawn of “agentic commerce, ” a new phase in retail where AI agents handle purchases on behalf of consumers. This could drastically alter the competitive landscape for retailers.
At present, the feature is only available to U.S.-based ChatGPT users, allowing them to purchase directly from Etsy, with Shopify merchants expected to be included soon. Although the immediate impact appears limited, analysts caution that the long-term effects of this technology could be revolutionary.
Despite e-commerce being around for almost three decades, “non-store retail” still represents just 28% of total retail sales in the U.S. (excluding autos and gas), and its annual growth rate is less than 1%. However, if agentic commerce accelerates that growth, D.A. Davidson predicts online sales could account for as much as 40% of total retail by 2035.
The potential benefits of this shift are clear: AI agents can make online shopping more efficient by refining product searches, providing personalized recommendations, and streamlining checkout, which could lower cart abandonment rates. They could also help manage auto-replenishment for household goods, more accurately predicting when products need to be restocked, leading to increased sales.
However, the report also highlights several risks. The most significant concern is the downward pressure on pricing. Because AI agents will likely be programmed to scour the web for the best deals, prices and margins could shrink across the board. Retailers will also face the need for more advanced optimization of product listings to ensure AI agents can find them—an effort that could be even more complex than current search engine optimization (SEO) strategies.
Another risk is that retail media networks, which depend on human interaction with websites, could lose value, leading to lower advertising returns. The rise of AI agents might also reduce impulse buying, as these agents are more likely to purchase only what is necessary, though some companies may adapt by teaching agents to suggest additional items. Moreover, the dominance of direct-to-consumer brands could increase, while traditional multi-brand retailers might find their influence waning.
D.A. Davidson also noted that many brands still rely heavily on wholesale channels due to the scalability and reach offered by large retailers. In the battle for dominance, scale will be a key differentiator. Companies like Amazon and Walmart, with their technological resources and vast customer data, are particularly well-positioned to thrive in this new landscape. Walmart’s own AI shopping agent, “Sparky, ” was mentioned as an example of early adoption.
Retailers that offer exclusive access to popular brands—such as Dick’s Sporting Goods and Ulta Beauty—could also see benefits, though having a broad selection of products may become less critical as AI agents can quickly search across multiple sites. On the other hand, project-based and membership-based retailers may face less disruption.
Finally, certain sectors, like home improvement or auto parts, may remain less affected by AI-driven commerce for the time being, as these categories often involve complex, multi-item purchases that still require human input. Companies like Home Depot and Lowe’s, where online sales make up around 15% and 10% of total sales, respectively, are less vulnerable. Warehouse clubs like Costco, BJ’s, and Sam’s Club could also stand to gain, as AI agents can more efficiently spot lower prices, though human memberships will still be required to complete transactions.
About The Author
