AI Agents for Salesforce Commerce Cloud Stores: What They Do and Where They Fit
AI agents for Salesforce Commerce Cloud stores fall into two groups. Salesforce's Agentforce Commerce agents work with shoppers, B2B buyers, and merchandisers. Operations agents, such as Noibu's, work on the storefront itself: finding errors, slow pages, and conversion leaks, drafting the changes, and measuring results after a person approves them.
TL;DR
- Salesforce's Shopper, Buyer, and Merchant agents serve shoppers, B2B buyers, and merchandisers.
- Operations agents work on the storefront: errors, speed, conversion, tests, and paid landing pages.
- SFCC storefronts are hard to keep healthy because of mixed architectures, cartridges, and configuration changes.
- Roll out agents read-only first, log Business Manager changes as releases, and route code through review.
Salesforce Commerce Cloud teams heard a lot about AI agents in 2026. Most of it was about agents that talk to shoppers. Fewer people asked what happens when those shoppers, human or agent-assisted, hit a storefront that is slow or broken. That is where the second kind of agent comes in.
Noibu is AI for ecommerce operations, the always-on optimization system that keeps online stores running at their best. A team of supervised AI agents finds what's broken, slow, or leaking revenue, does the work, and hands results back for human approval.
What changed for Salesforce Commerce Cloud in 2026?
Salesforce now markets its commerce products as Agentforce Commerce. In mid-2026 its Shopper, Buyer, and Merchant agents became generally available, alongside Agentic Commerce Search, a new front-end framework called Storefront Next, and developer tooling that plugs B2C Commerce into AI coding tools.
What do Salesforce's commerce agents do?
| Agent | Works with | What it does |
|---|---|---|
| Shopper Agent | Consumers on the storefront | Guides discovery, product questions, and checkout in the brand's voice |
| Buyer Agent | B2B buyers | Handles buying conversations in channels such as WhatsApp and SMS |
| Merchant Agent | Merchandising teams | Organizes catalogs, sorts products, and sets storefront rules in plain language |
All three work with shoppers, buyers, or merchandisers on commerce tasks. None of them is designed to find a JavaScript error on the payment step or a slow product template on mobile.
What do operations agents add on an SFCC store?
Operations agents add continuous work on storefront health and conversion. They watch sessions, errors, and Core Web Vitals across every site and locale, find what is broken, slow, or leaking revenue, trace it to a cause, and draft the change: a code change, a test, or a landing page improvement. A person approves before anything ships.
1.58s · 80–85ms · 0.02
LCP, INP, and CLS held by the top 10% of ecommerce performers under Holiday 2025 peak load, measured on real users.
Noibu Holiday 2025 Ecommerce Benchmark Report
Why are SFCC storefronts hard to keep healthy?
- Mixed architectures. Many organizations run SiteGenesis, SFRA, and headless builds on Composable Storefront at the same time, each failing differently.
- Cartridge stacks. Payment, tax, reviews, and search integrations add code paths that interact in ways nobody tested together.
- Changes without deploys. Promotions, content slots, and site preferences changed in Business Manager can alter the storefront with no code release.
- Server-side failures. Some errors happen on the server and never show in the browser, so they need logs tied to sessions.
- Platform code you cannot edit. Some issues sit in built-in platform code and have to be reported to Salesforce rather than changed directly.
How do you connect SFCC server logs to shopper sessions?
Connect server logs to sessions by passing the SFCC session ID into your session data and streaming logs from the SFCC side. Noibu's logs integration does this for stores on SiteGenesis or SFRA, so a checkout failure that only appears in server logs can be traced to the shoppers it affected.
How should SFCC teams roll out AI agents?
Rolling out operations agents on SFCC
- Start read-only. Run daily checks on errors, speed, and conversion across every site and locale.
- Log every change as a release. Include Business Manager changes, not just code versions.
- Route code through review. Agent-drafted changes go to your repository as pull requests, then through your normal deploy.
- Protect checkout and payment cartridges. Require developer review for anything in those code paths.
- Measure per site. Compare results for each site and locale, because one can improve while another regresses.
How does Noibu work on Salesforce Commerce Cloud?
Noibu supports Salesforce Commerce Cloud alongside Shopify, BigCommerce, Adobe Commerce, commercetools, and headless storefronts. Its agents cover errors, performance, conversion, A/B tests, ROAS, and accessibility, draft code changes as pull requests in your code host, and measure results after a person approves. Teams work with the agents in the Noibu console or from Claude through Noibu MCP.
Frequently asked questions
What are Agentforce Commerce agents?
They are Salesforce's commerce agents. Shopper Agent handles conversations with consumers from discovery through checkout, Buyer Agent serves B2B buyers in channels such as WhatsApp and SMS, and Merchant Agent helps merchandising teams manage catalogs and storefront rules in plain language.
Is Salesforce Commerce Cloud now called Agentforce Commerce?
Salesforce now markets its commerce products under the Agentforce Commerce name. Many teams still say Salesforce Commerce Cloud, SFCC, or B2C Commerce for the storefront platform.
Do Salesforce's commerce agents find site errors and slow pages?
They are built for shopping, buying, and merchandising work. Finding and resolving storefront errors, speed problems, and conversion leaks needs session, error, and performance data, which is the job of operations agents.
Does Noibu work with Salesforce Commerce Cloud?
Yes. Noibu supports Salesforce Commerce Cloud alongside Shopify, BigCommerce, Adobe Commerce, commercetools, and headless storefronts. For stores on SiteGenesis or SFRA, Noibu's logs integration also matches server logs to shopper sessions.
Can AI agents change code on an SFCC storefront?
They should draft changes rather than deploy them. Noibu's agents draft code changes as pull requests in your code host, and your team deploys them through its normal process after approval.
Related topics
- What is ecommerce log monitoring?
- Troubleshooting the Salesforce Commerce Cloud Apple Pay error
- AI agents for ecommerce: what they actually do
- What is agentic commerce?
See what Noibu's agents would surface on your SFCC store: run a free website audit →
