As AI agents proliferate across marketing, commerce, and service, Salesforce is anticipating that competitive advantage will come from smarter agents that share context rather than their sheer number.
That premise shapes the flagship capabilities around agentic commerce in Agentforce 360 arriving with the vendor’s Spring ’26 Release. They signal a shift away from siloed AI deployments toward a unified intelligence layer designed to keep customer experiences coherent from first click to post-purchase support.
As more AI agents come online, many enterprise are discovering that deploying them one function at a time results in agents being optimized locally based on the data, APIs and workflows for their own systems. As Nitin Mangtani, Salesforce SVP/GM, Salesforce Commerce and Retail Clouds, put in a blog post:
“So while context exists across the enterprise, you may end up designing a solution where no single agent can see or reason across it all.”
The result is a fragmented experience where marketing, commerce, and service interactions may work individually, but fail to connect. Customers run into broken promises, get asked the same questions again and again, and feel like every interaction starts from scratch.
Salesforce’s take on fixing this is to add a new set of commerce and messaging capabilities to Agentforce 360 designed to connect scattered touchpoints into continuous conversations.
Agentforce 360 allows companies to deploy AI agents “on a single shared intelligence layer, where context is understood, interpreted and acted on consistently,” according to Mangtani, avoiding the fragile, hard-to-maintain integrations that tend to bog down DIY setups.
There are three new agentic commerce features coming to Agentforce 360 in February.
Two-way messaging across email, SMS, and WhatsApp turns basic notifications into interactive channels. Rather than pushing customers back to a website, the same agent intelligence that powers guided shopping can answer questions, recommend products, and even complete transactions directly within a conversation, moving the interaction away from one-way messaging. Salesforce positions this as a move away from broadcast marketing toward something closer to “texting a personal shopper.”
Contextual Search replaces traditional keyword-based queries with intent-aware responses, allowing customers to ask questions in natural language. They receive answers informed by factors such as inventory, shipping availability, and previous interactions, whether through the storefront search bar or a shopping agent.
Agentforce Commerce Guided Shopping also takes a more proactive approach. Agents can step in at key points in the customer journey to provide navigation help, confirm product availability, calculate delivery timelines, and answer questions within a single, continuous conversation.
Because these experiences are built on shared context, conversations can move seamlessly across channels. A question asked in an email can continue on the storefront. A service interaction can pick up where commerce left off. As Salesforce puts it:
“One customer. One conversation. One shared context.”
Agentic Commerce Inside the Spring ’26 Release
The new agentic commerce capabilities arrive alongside Salesforce’s broader Spring ’26 Release, which the company says is designed to help organizations become Agentic Enterprises, environments where “human expertise and AI agents drive customer success together.”
The updates include a range of tools designed to help companies use AI and data more effectively across sales, customer service, and commerce.
In commerce operations, Agentic Order Routing can flag orders that a location can’t fulfill and automatically reroute them, helping prevent delays before they affect customers.
On the engagement side, Two-Way Email brings the same conversational approach to inboxes, letting AI agents respond to questions, suggest products, and handle support cases, reducing reliance on one-way messages.