Agentic commerce has quickly become one of the most discussed emerging trends in retail and technology, raising questions about whether enthusiasm is outpacing real-world progress. 

A recent industry commentary from Mizuho Securities centered on this topic, featuring insights from tech industry analyst Juozas Kaziukėnas. 

The brokerage addressed whether the concept is becoming overstated, concluding that the “simple answer: No,” while emphasizing that the technology remains early and exploratory.

According to the brokerage, agentic commerce currently functions as “a massive umbrella of various shopping experiments,” attempting to bring AI-driven shopping interactions to everyday consumers. 

The analyst indicated it may take several years before widespread commercial adoption or highly interactive interfaces develop, despite early signs of progress.

Signs of consumer behavior shifts are emerging. The brokerage notes that approximately 1% of all Walmart website traffic is now being referred from ChatGPT, growing from “essentially nothing 3–6 months ago.” 

Referral traffic overall represents about 4% of Walmart site activity, and more than one-third of that now originates from ChatGPT.

The brokerage suggests this referred traffic could prove incremental rather than simply diverted from search engines, though it does not provide quantified evidence at this stage.

Retailers and platforms are accelerating involvement through partnerships, including agreements with Etsy, Shopify and Walmart that expand product access within ChatGPT.

Target recently announced a collaboration that introduces conversational AI tools such as “Shopping Assistant” and “Gift Finder,” including the ability to shop fresh food categories.

Google also announced agentic shopping capabilities that include conversational search, price tracking and local inventory visibility, with agentic checkout beginning to roll out through Chewy and Wayfair.

However, the brokerage highlights structural challenges that could temper expectations. The process of how agents choose product recommendations was described by the expert as “a mess” and “a massive puzzle to solve,” with Reddit identified as an unusually influential source for product visibility.

The commentary noted that this visibility strategy is easily manipulated, which may complicate reliability and consumer trust.

The industry remains divided on whether the rush to adopt agentic systems will lead to meaningful near-term gains. 

The brokerage describes broad movement into partnerships and product catalog integrations, while noting that Amazon “may be able to wait this all out and adapt once the dust settles.”

Tractor Supply is testing its own AI “Scout” agent to support faster product recommendations through beta deployment, demonstrating experimentation even among specialty retailers.

While momentum is clear, the economics remain uncertain. In the Walmart partnership with OpenAI, the brokerage notes that the revenue take rate could be “as low as 0% initially,” signaling early-stage positioning rather than proven commercial models.

Taken together, the evidence presented shows enthusiasm backed by measurable early activity but tempered by technical hurdles and unclear financial outcomes. 

Adoption data remains narrow, product recommendation systems are underdeveloped and traffic gains, while rapid, are still small in scale.