That shift changes the implementation question too. Commerce teams once focused mainly on storefront configuration, catalog management, checkout, and order processing. Agent-based shopping adds another layer because an AI agent may search products, answer questions, retrieve order information, or help complete a transaction. Companies considering Agentforce Commerce therefore need to understand how Salesforce reached this point before deciding what their own implementation should cover.
Agentforce began as a wider Salesforce AI layer in 2024
The starting point came on September 12, 2024, when Salesforce introduced Agentforce as a suite of autonomous AI agents. Salesforce described agents that could analyze information, make decisions, and perform tasks across areas including sales, service, marketing, and commerce. The announcement also introduced the Atlas Reasoning Engine and positioned low-code agent creation as part of the platform.
Commerce was therefore part of the Agentforce direction from the beginning, but the operating model was still developing. Traditional Commerce Cloud projects had usually centered on the store itself and the systems directly supporting it. Agentforce added a new question: what actions should software be allowed to take for a shopper or employee, and what information should it be permitted to use?
That question now sits near the center of Agentforce Commerce Implementation & Consulting Services. Implementation planning has to account for agent permissions, customer and product context, integrations, fulfillment rules, testing, and the points where human review remains necessary. The project scope can therefore extend beyond adding an AI interface to an existing storefront.
Commerce moved toward guided shopping during 2025
Salesforce made the commerce direction clearer during 2025. Guided Shopping capabilities were designed to help B2B and D2C customers find products, review previous orders, reorder items, and check order status through conversational interactions. Salesforce published setup guidance for Guided Shopping on October 3, 2025, showing that agent-assisted product discovery had moved from a broad Agentforce concept into a defined commerce use case.
The next major milestone came in November. Salesforce announced new Agentforce Commerce capabilities on November 17, 2025, reporting that traffic from AI assistants to retail sites had risen 119%. The announcement covered guided shopping, order routing, unified commerce operations, and future connections between merchant catalogs and consumer AI channels.
This period explains why an Agentforce Commerce project needs a wider system view. A useful deployment depends on the quality of catalog information, inventory availability, order data, customer permissions, and the actions exposed to the agent. Weaknesses in those areas can surface quickly once conversational access sits in front of them.
The 2025 holiday season supplied evidence of changing shopper behavior
The 2025 holiday period gave businesses a measurable view of how AI was affecting online shopping. Salesforce reported $1.29 trillion in global online holiday sales between November 1 and December 31, with AI and agents influencing $262 billion of that spending. Shoppers arriving through AI-powered search channels converted 9 times more often than shoppers referred through social media, according to Salesforce’s holiday analysis.
Retailers using their own AI agents also recorded a 59% higher sales growth rate in Salesforce’s dataset, averaging 6.2% year-over-year growth compared with 3.9% among retailers without those agents. These figures don’t guarantee the same outcome for every company because retailer size, product category, implementation quality, and customer behaviour differ. They do show why commerce teams are paying closer attention to AI-assisted discovery and transaction paths.
For teams planning an Agentforce Commerce implementation, the lesson is operational. Agent performance depends on the systems behind the conversation. Product information has to be accurate, inventory needs dependable updates, order actions require clear rules, and customer information needs appropriate access controls before an agent is allowed to act.
January 2026 pushed commerce beyond the retailer’s own storefront
A further change arrived on January 14, 2026. Salesforce announced support for Google’s Universal Commerce Protocol, with plans to support product discovery and shopping experiences through Google surfaces including AI Mode in Search and the Gemini app.
That development matters because the storefront is no longer the only possible starting point for a digital purchase. A shopper may begin with an AI conversation and expect current product information, availability, or transaction support without following the older browse-search-checkout sequence. Commerce architecture must therefore consider how product and transaction information can be exposed safely to external AI environments.
This changes the role of Salesforce Commerce Cloud consulting as well. The work may include catalog readiness, commerce APIs, order workflows, customer data, agent permissions, and testing across several buying paths. Planning those dependencies before deployment reduces the risk of building an agent experience on top of inconsistent commerce data.
June 2026 turned separate capabilities into a broader commerce model
Salesforce described its June 2026 update as its biggest Agentforce Commerce release in 5 years. The release made Shopper Agent, Buyer Agent, and Merchant Agent generally available and connected the commerce direction across B2C, B2B, point of sale, and order management. Salesforce also announced connections with external AI environments, extending the idea that product discovery can happen beyond a merchant-owned website.
The company’s June 2026 Agentforce Commerce release also cited the holiday evidence behind the shift, including the 20% share of global online sales influenced by AI and agents. Salesforce reported that AI-referred traffic converted at 8 times the rate of social traffic in the data used for the June announcement. Different Salesforce publications use different comparison windows and methodologies, so these figures should be read within the context supplied by each source.
For implementation teams, the current position is clearer than it was in 2024. Agentforce Commerce now covers shopper-facing assistance and merchant work while connecting with established commerce functions. That makes architecture, permissions, data quality, integration testing, and measurable use cases central implementation concerns.
What the timeline means for an implementation decision now
The 2024 to 2026 timel
