For more than two decades, ecommerce has been built around one core mechanic: humans browse, click and buy.

Online retail operating models were designed around that behaviour. Search visibility, paid acquisition, conversion optimisation and UX all revolved around traffic and clicks as the primary economic drivers of growth. Platforms such as Salesforce Commerce Cloud (SFCC) became powerful engines for managing catalogue, promotions, checkout and fulfilment optimised for human-led journeys.

That assumption is now being challenged.

Agentic commerce has emerged as a new operating model where AI systems do not simply recommend products but participate directly in transactions, providing intelligent and goal-oriented interactions. Customers delegate intent and constraints to an autonomous agent that evaluates options, selects offers and can increasingly execute the purchase itself.

The AI agent is the mechanism. Agentic commerce is the operating model.

Global research supports this shift. Morgan Stanley estimates AI-driven shopping assistants could represent 10% to 20% of US ecommerce by 2030, equivalent to $190 billion to $385 billion in annual online sales. In the UK, Google notes that agentic commerce is already moving from concept to execution, with retailers actively building toward AI-mediated discovery and checkout flows. Independent industry reporting suggests more than 70 percent of retailers are already piloting or partially deploying AI agents, with automation and operational integration delivering early operational gains.

Platform signals reinforce that direction. Google’s Universal Commerce Protocol (UCP) is an open standard backed by major commerce partners, designed to allow AI agents to interact with retailer systems from discovery through checkout and post purchase without bespoke integrations.

For agentic commerce to scale, commerce itself must become machine operable. This is where enterprise platforms such as Salesforce Commerce Cloud move from being storefront engines to structured transaction authorities.

The shift beyond the click is not about interface innovation. It is about infrastructure readiness.

‍

Winning in a World Where Agents Act, Not Clicks

In traditional ecommerce, success was measured through clicks, views and conversion funnels. Retailers optimised for attention and persuasion because human behaviour drove the journey.

As AI agents move from recommendation to execution, the unit of competition shifts from interaction to execution certainty.

Agents do not scroll category pages or absorb brand storytelling. They interrogate structured product data, validate inventory through APIs, assess promotion logic and execute transactions based on deterministic business rules. Salesforce Commerce Cloud, with its pricing engine, promotion framework, OCAPI/SCAPI layers and order management capabilities, becomes the system agents must trust to transact reliably.

Retailers running SFCC already possess many of the structural components required for agentic commerce:

• Centralised catalogue and taxonomy
• Promotion and pricing governance
• Inventory integration
• Secure checkout and payment orchestration
• Order lifecycle management

However, possessing these capabilities is not the same as exposing them in a machine-readable, consistent and real-time way.

Retailers with fragmented integrations, inconsistent feeds or delayed inventory synchronisation are structurally disadvantaged because AI agents prioritise reliability and completeness over persuasion.

This requires a mindset shift: from optimising for visibility to optimising for eligibility.

Brands that present machine-operable offers consistently and at scale via structured feeds, governed APIs and clean integration layers will be selected more often by autonomous systems.

‍

UCP and the Standardisation of Machine-Operable Commerce

Agentic commerce cannot scale without standardisation. Without common frameworks, every AI system would require bespoke integrations into each retailer’s commerce stack, limiting adoption.

Google’s Universal Commerce Protocol addresses this by establishing a structured way for AI agents to interact with retailer systems across discovery, checkout and post purchase. It enables agents to interpret product data, validate pricing and availability, initiate transactions and support fulfilment without requiring custom integrations for every platform.

This matters because most enterprise retailers already operate on established commerce ecosystems such as Salesforce Commerce Cloud. UCP does not replace SFCC. It enables structured interoperability across it.

For retailers running Salesforce Commerce Cloud, the strategic question is not whether SFCC can theoretically support agentic commerce. It is whether the implementation is architected to expose machine-readable offers and transactional workflows with clarity and reliability.

That requires:

• Structured and attribute-rich product data
• Governed OCAPI / SCAPI endpoints
• Real-time inventory visibility
• Consistent pricing and promotion logic across regions
• Clean integration between SFCC, OMS and payment providers

SFCC provides the transactional authority. UCP provides the interoperability layer. The competitive advantage lies in how well those layers are orchestrated.

Platforms alone are not enough if they operate in silos. The real differentiator is how well commerce architecture, integrations and data governance support consistent machine operability.

‍

Beyond Front-End Experience: Infrastructure Becomes the Strategic Layer

In the click era, brands competed through creative, UX and campaign optimisation. In the agentic era, infrastructure competes.

AI agents require consistent access to operational truth. Within Salesforce Commerce Cloud that means pricing logic must be deterministic, promotions must resolve predictably, inventory must reflect real availability, and checkout must complete without ambiguity.

When operational truth is fragmented across systems ERP, OMS, payments, promotions engines agents detect inconsistency and route elsewhere. When it is governed and reliable, agents can transact with confidence.

Agentic execution also raises the bar for checkout and payments. When transactions are initiated or mediated by AI agents, retailers need secure payment orchestration, tokenised flows and predictable handoffs between commerce platforms and payment providers.

Within an SFCC ecosystem, this means tight integration between:

• SFCC storefront and APIs
• Order Management
• Payment service providers such as Stripe or Adyen
• Fraud and risk engines

Agentic commerce is not a front-end innovation project layered on top of SFCC. It is a commerce architecture and operating model evolution that depends on how well SFCC is implemented, integrated and governed. The retailers who respond well will not be the ones who experiment fastest, but the ones who execute most reliably.

‍

Customer Experience Becomes Dual-Layered

Agentic commerce does not diminish the importance of human experience. Brand storytelling, emotional resonance and curated journeys still shape preference and loyalty.

However, a second layer now matters equally: the machine experience.

For humans, experience is shaped by design, relevance and trust.

For humans, experience is shaped by design, relevance and trust. For AI agents, experience is shaped by structured completeness, policy clarity, data quality and system reliability. The organisations that orchestrate both effectively will outperform.

In an SFCC environment, this means treating APIs, feeds and integrations with the same strategic importance as front-end UX.

This shift is inherently cross-functional. Marketing, digital commerce, product, engineering and operations must align because the systems that power execution increasingly shape the customer outcome.

‍

Where This Is Heading

Agentic commerce is moving from concept to execution. UCP is one of the clearest signals that the industry is standardising how AI agents interact with retailer systems.

For Salesforce Commerce Cloud retailers, the opportunity is significant. SFCC already contains the core components required to act as a trusted transaction authority in an agent-mediated world. The competitive question is no longer simply how well it converts human traffic but how reliably it can be interpreted and executed by machines.

Retailers will increasingly compete not only through what customers see but through how consistently and confidently their SFCC infrastructure can transact on behalf of autonomous agents.

If you are assessing what this shift means for your Salesforce Commerce Cloud implementation, integrations or global commerce architecture, we are happy to share a structured readiness lens and practical next steps.

‍

You can reach us at info@thecommerceteam.com

‍