Salesforce Agentforce Commerce Goes GA With Seven Named Agents: What Changes for Enterprise Commerce Teams Before Peak Season

Salesforce introduced seven named Agentforce agents on September 11, 2026, each scoped to a specific business function rather than shipped as one general purpose assistant. Casey, Paige, Carter, Marshall, Piper, and Fin are generally available now, covering sales, service, commerce, supply chain, and customer experience work. Hunter, the seventh, stays in pilot, and it is the first Agentforce agent running on a new long horizon runtime built to pursue a goal over the course of weeks instead of resolving inside a single chat session, a meaningfully different execution model from the rest of the lineup.

Carter is the one that matters for commerce teams. It is described as a shopper agent, and it rolls up work Salesforce had already been building toward earlier in 2026 under the Agentforce Commerce umbrella: a Shopper Agent for retailer websites and agentic commerce search capabilities aimed at both B2B and B2C brands. The general availability date is not incidental. Salesforce is explicitly positioning this GA as ready ahead of peak season, the run up to Black Friday and the holiday shopping window that most retail commerce teams spend Q3 preparing for, which puts a real deadline on any evaluation decision rather than leaving it as an open ended roadmap item.

What actually shipped

Salesforce's own material cites AI assistants driving a 119 percent increase in retail traffic volumes as part of the launch framing. Treat that figure the way you would treat any vendor reported adoption statistic released alongside a product announcement: directionally interesting as evidence that AI driven shopping traffic is growing fast, but not a number to plan a specific ROI case around without knowing the underlying methodology, comparison baseline, or which subset of Salesforce customers it was measured against.

Why Carter is a data problem before it is an integration problem

Carter's job is to represent your product catalog inside a conversational, agent mediated shopping flow, which means its usefulness is capped by the same constraint every AI shopping agent runs into regardless of vendor: the quality and completeness of the structured product data underneath it. We covered this in depth when the Universal Commerce Protocol and Agentic Commerce Protocol data requirements came into focus, structured, consistent product data covering every attribute a shopper might reasonably query, live and accurate pricing and availability, and identifiers that resolve cleanly for a transaction rather than just a search result. Carter reads from the same category of underlying data layer that both of those protocols depend on, and a catalog with merchandising detail hardcoded into frontend templates rather than the actual product record will limit what Carter can do regardless of how well Salesforce built the agent itself.

This is the practical reason enterprise commerce teams should not treat Carter's GA date as the moment to flip a switch. The honest first step is a catalog and data audit: does every product have the structured attributes a shopper actually asks about, is pricing and availability synced in near real time rather than on a batch delay, and is there a single source of truth for product data rather than multiple systems that can drift out of sync. Teams that already did this work for search or for other agentic commerce protocols are close to ready for Carter specifically. Teams that have not are looking at a data project, not a rollout project, and starting that work in late September leaves meaningfully less runway before peak season than starting it in the spring would have.

Hunter is the one to watch, not the one to deploy yet

Hunter's pilot status and its long horizon runtime are worth tracking even though it is not commerce specific and not ready for production use. A runtime built to pursue a goal over weeks rather than resolve in a single session is a different operating model from every other agent in this lineup, and if it graduates out of pilot successfully, it signals where Salesforce intends to take Agentforce next: agents that manage an ongoing objective, inventory rebalancing across a season, a multi step vendor negotiation, a gradual personalization rollout, rather than agents that resolve a single customer interaction and stop. Commerce teams evaluating Agentforce now are evaluating the GA lineup, but the roadmap signal sitting in Hunter's pilot status is worth factoring into a multi year platform decision, not just a peak season one.

What to actually evaluate before turning Carter on

Start with the data audit above before anything else, since it determines the ceiling on what Carter can represent regardless of configuration. Then scope a narrow pilot against a subset of your catalog, ideally the categories where you already have the most complete structured data, rather than a full catalog rollout timed against a hard peak season deadline, since a narrow pilot surfaces data gaps and integration issues while there is still time to fix them. Confirm how Carter's shopper interactions integrate with your existing analytics and attribution, since agent mediated shopping sessions do not always flow through the same tracking as a traditional browsing session, and a gap there can make it hard to actually measure whether Carter is driving the traffic or conversion lift Salesforce's own numbers suggest is possible. And treat the 119 percent traffic figure as a hypothesis to test against your own data, not a result to assume you will replicate.

FAQ

Which Agentforce agents are generally available as of September 2026? Six of the seven: Casey, Paige, Carter, Marshall, Piper, and Fin. Hunter remains in pilot.

What does Carter actually do? Carter is Salesforce's shopper facing commerce agent, built on top of Agentforce Commerce capabilities for B2B and B2C brands, including a shopper agent for retailer websites and agentic commerce search.

Why did Salesforce time this GA for September? Salesforce is explicitly positioning the release as ready ahead of peak holiday shopping season, giving commerce teams a defined evaluation and rollout window before the highest traffic and revenue period of the year.

What does a commerce team need before deploying Carter? Complete, consistent structured product data, live pricing and availability, and a single source of truth for product records. Carter's usefulness is capped by the same data quality constraints as any AI shopping agent, regardless of the platform behind it.

When this applies to your stack

If your team is evaluating Carter or any Agentforce Commerce capability ahead of peak season, the highest leverage work before the rollout itself is a structured product data audit against what an agent actually needs to read. That data architecture and platform integration work, scoped against a real peak season deadline, is what we do at Contra Collective.