For years, retail has talked about AI in terms of features that can help with better recommendations, smarter forecasts, and faster user services. These were all incremental, bolted-on improvements layered onto an already complex stack.
NRF 2026 made it clear that this phase is coming to a close.
What emerged instead was a shared recognition that retail is moving toward a new operating model shaped by agentic systems, open standards and platforms that span across the entire commerce journey. This is not just another technology cycle. It is a shift in where decisions are made, how value flows and who ultimately controls the customer experience.
Google’s strategy is to not to position AI as a capability retailers consume, but as shared infrastructure retailers can build on for every step of what remains of the future of the existing funnel.
From optimisation to orchestration
Most AI deployments in retail today are still optimised for local impact. A model improves conversion on a product page. Another reduces service costs. Each delivers value, but largely in isolation.
Agentic AI changes the equation by connecting these moments.
Rather than responding to single prompts, agentic systems can reason across contexts, plan actions and execute them – within defined guardrails – across discovery, selection, checkout, fulfilment and service. Commerce stops behaving like a sequence of handoffs and becomes a coordinated system.
When AI becomes the infrastructure, it doesn’t just improve outcomes: it reshapes control. Decisions about what is shown, what is recommended and what action is taken move upstream. For retailers, that raises a critical question: if AI is orchestrating the journey, who defines how those decisions are made?
Why decisions matter more than transactions
Transactions are already efficient. In many cases, they are commoditised.
What remains scarce is confidence. Helping customers decide quickly, clearly and with minimal friction has always been retail’s hardest problem. Agentic systems move directly into this space.
When an AI agent filters options, compares alternatives or completes a purchase on a customer’s behalf, it becomes part of the decision itself, not just the interface. That has profound implications for brand visibility, pricing dynamics and loyalty.
Retail advantage increasingly depends on influence at this decision layer. Those who shape how decisions are made will capture more value than those who simply execute transactions, regardless of where checkout ultimately happens.
Interoperability unlocks scale
None of this works in isolation.
Retail has spent decades stitching together platforms, tools and partners. Agentic systems immediately expose the limits of that fragmentation. Without shared standards, agents remain trapped inside individual ecosystems, unable to reason or act across brands and channels.
This is why open protocols such as the Universal Commerce Protocol (UCP) matter. By giving agents and commerce systems a common language across discovery, decision and checkout, interoperability becomes the foundation, not an afterthought.
Every major digital shift has followed this pattern. Payments, the web and mobile all scaled because shared standards came first. Agentic commerce is no different.
When discovery becomes the moment of purchase
One of the most visible consequences of this shift is the collapse of the retail funnel.
As AI-driven discovery surfaces begin to support native checkout, the distance between intent and transaction shrinks dramatically. Customers no longer move through clearly defined stages. Discovery, decision and purchase increasingly happen at the same time.
For retailers, this changes the rules.
Data architectures built around channels begin to break down. Attribution models lose clarity. Organisational boundaries between marketing, commerce and service start to blur.
At the same time, the opportunity is significant: higher conversion, faster decision-making and a more seamless customer experience. The retailers that succeed will be those that treat commerce as a continuous, intent-led system.
What this means for the next 12–18 months
This transition will not happen overnight, but the most important control points will be claimed quickly.
The next 12–18 months represent a narrow window in which retailers can actively shape their role in the emerging agentic commerce ecosystem. Decisions made now about architecture, interoperability and governance will determine whether retailers influence how commerce works or simply plug into systems defined by others.
Read our whitepaper to understand how Google Cloud can help retailers move beyond experimentation and remain in control as agentic commerce becomes operational.