Agentic Commerce is Reshaping Retail … and Most Operating Models Aren't Ready

Retailers have seen this movie before.

When mobile commerce first emerged, many brands treated it as a lightweight extension of e-commerce — a smaller screen layered onto existing digital infrastructure. Over time, it became clear that mobile was not simply a new interface. It fundamentally changed customer behavior, expectations, operating models, and the economics of digital commerce itself.

Retailers are now entering a similar phase of agentic commerce adoption, one where the focus is shifting from experimentation to operational readiness.

Over the past year, much of the industry conversation has centered on AI shopping assistants and conversational interfaces. Increasingly, however, retailers are realizing that agentic commerce is beginning to reshape how products are discovered, evaluated, and potentially purchased.

The implications extend far beyond the customer interface. Participating effectively in agentic commerce requires retailers to rethink how their commerce ecosystems are structured, how product data flows across platforms, and how customer relationships are maintained when discovery is mediated externally.

Retailers cannot approach this as a bolt-on innovation initiative. They need a new channel strategy.

The reality is that AI agents do not behave like human shoppers. They do not browse category pages, respond to visual merchandising in the same way, or navigate digital storefronts sequentially. Instead, they interpret structured data such as attributes, taxonomy, metadata, availability, pricing, reviews, and fulfillment signals to resolve intent-based prompts like: “Find me a lightweight running shoe under $150 with strong arch support.”

That creates an entirely different optimization challenge.

For years, retailers optimized digital commerce around human behavior and search engine visibility. Increasingly, they will also need to optimize for machine interpretability.

What many retailers are now discovering is that agentic commerce is ultimately an enterprise architecture challenge. Many still operate with fragmented product information management systems, inconsistent metadata standards, siloed inventory visibility, and complex commerce platforms. In agentic commerce environments, those limitations become far more consequential because structured product data quality and ability to continuously evolve product data attributes directly influences discoverability, recommendation outcomes, and transaction visibility.

The retailers best positioned for agentic commerce will be the organizations with the strongest and most flexible underlying product data foundations, the clearest operational visibility, and the ability to expose structured commerce information across systems in real time.

In practice, this means retailers must begin modernizing foundational capabilities that historically sat behind the scenes:

• product information management and governance;
• taxonomy and metadata standardization;
• real-time inventory and fulfillment visibility;
• semantic data layers and interoperability; and
• flexible and adaptable commerce infrastructure.

These capabilities are no longer operational nice-to-haves. They're becoming prerequisites for participation.

As AI agents increasingly mediate discovery and purchasing decisions, retailers will also need to rethink long-standing assumptions about customer ownership, attribution, and loyalty. In traditional e-commerce, retailers controlled the digital storefront and customer journey. In agentic commerce environments, portions of that journey may increasingly occur outside owned channels.

That creates new strategic questions: Where does brand differentiation occur when AI systems summarize product options? How should retailers think about loyalty integration when transactions are initiated externally? What happens to attribution models when AI agents, rather than consumers, increasingly shape purchase decisions?

These are not hypothetical concerns. They represent the early stages of a broader shift in how digital commerce ecosystems may function over the next decade.

That doesn't mean retailers should resist participation. Quite the opposite. History shows that new channels tend to reward organizations willing to engage early and adapt deliberately. But it does mean retailers need to enter agentic commerce intentionally, with a clear understanding of the operational and strategic implications.

Mobile commerce ultimately reshaped retail because it forced companies to redesign experiences, workflows, and operating models around new consumer behaviors. Agentic commerce has the potential to do the same.

The retailers that succeed in agentic commerce will not simply deploy better AI interfaces. They will redesign their operating models, data foundations, and commerce architectures for a world where AI systems increasingly participate in the customer journey itself.

Sudip Mazumder is senior vice president, retail industry lead, North America at Publicis Sapient, a technology company that provides enterprise AI platforms and services.