Online outdoor apparel search results page highlighting teal women's hiking tops with product photos and prices.

Agentic Commerce Search

Power context-aware search and conversational experiences that elevate the shopper experience and drive more revenue. Agentic Commerce Search is the only ecommerce search engine that builds a custom small language model for every customer, giving you a data advantage.

These leading brands trust Salesforce to power their agentic search

Boot Barn Logo
Living Spaces Logo
Loft Logo
Pacsun Logo
Rally House Logo
Tillys Logo
West Marine Logo

Here’s what sets Agentic Commerce Search apart

85 %
rule reduction
Merchandisers eliminate 1,000s of search rules, reducing manual overhead.*
+ 33 %
clicks per user
When shoppers find what they’re looking for, they engage more.*
+ 17 %
add-to-cart per user
More engagement means larger basket sizes (and happier customers).*

Transform search into revenue

Turn your search bar into an AI-native search engine that understands shopper intent. Customers drive an average of 12% more revenue per user.**

**Cimulate customer success metrics.

Boot Barn e-commerce header displaying menu categories, free shipping on boots, and an extra 25% off sale banner featuring America's 250th milestone collection.

Continuously optimize your product rankings based on shopper behavior, catalog changes, and business objectives. Use this context to evolve experiences autonomously.

Grow basket size by keeping customers engaged with a deep understanding of matching and related products.

Search results interface on Northern Trail website displaying search term query for Eagles alongside recommended sports merchandise including a clear tote bag.

Deliver rich, conversational shopping experiences within your Agentforce Shopper Agent. Let context guide your shoppers to the perfect product.

Boot Barn e-commerce header displaying menu categories, free shipping on boots, and an extra 25% off sale banner featuring America's 250th milestone collection.
Search results interface on Northern Trail website displaying search term query for Eagles alongside recommended sports merchandise including a clear tote bag.

A custom small language model built for you

You have the catalog, but you don't have billions of clicks to teach an AI what “summer wedding guest” actually looks like in your inventory. Agentic Commerce Search knows and surfaces what shoppers want.

Visual graphic highlighting the difference made when integrating Agentic Commerce Search into user data systems.

Using frontier LLMs, we generate synthetic searches, clicks, add-to-carts, and purchases to create more behavioral signals than your web traffic allows. With these synthetic journeys, we understand intent to purchase for all types of shoppers on your site across your entire product catalog.

We tune the synthetic data with your real behavior data. With a typical ratio of 10:1 synthetic data to real behavior.

Diagram illustrating how small language models utilize synthetic data, product catalogs, customer profiles, and shopping journeys to power AI shopper agents and agentic search.

Once deployed, Agentic Commerce Search produces remarkably accurate results and drive increased engagement and conversion. It improves autonomously with every interaction, reducing merchandising rules by 85% and increasing clicks per user by 33%.***

***Cimulate customer success metrics.

Diagram showing how custom small language models convert catalog and shopper inputs into faster, cost-effective e-commerce recommendations and shopping cart actions.
Visual graphic highlighting the difference made when integrating Agentic Commerce Search into user data systems.
Diagram illustrating how small language models utilize synthetic data, product catalogs, customer profiles, and shopping journeys to power AI shopper agents and agentic search.
Diagram showing how custom small language models convert catalog and shopper inputs into faster, cost-effective e-commerce recommendations and shopping cart actions.

Intelligent search with simple setup

Seamlessly integrate Agentic Commerce Search into any commerce stack with our headless APIs or get started out-of-the-box in B2C Commerce.

Architecture diagram showing the search request flow from an e-commerce storefront search box to the agentic search API and backend catalog infrastructure.

Toggle on with a click and start building your small custom language model to power Agentic Commerce Search on your site.

Merchant admin dashboard displaying a provisioning status page for setting up an AI-driven e-commerce search engine with a progress tracker.

All we need to build your AI model is your product catalog feed and your shopper data. No complex set up. No rule configuration. Sit back and let the AI do the work.

Feature comparison slide detailing the benefits of Agentic Commerce Search over traditional search setup methods.

Optimized for Commerce. Built for your brand. Powered by synthetic data. Your shoppers get a natural language experience that mirrors how customers shop today. You get higher conversion and revenue.

E-commerce search interface showing product listings and category filters for country festival attire including ruffled mini dresses, midi dresses, and snip-toe boots.
Architecture diagram showing the search request flow from an e-commerce storefront search box to the agentic search API and backend catalog infrastructure.
Merchant admin dashboard displaying a provisioning status page for setting up an AI-driven e-commerce search engine with a progress tracker.
Feature comparison slide detailing the benefits of Agentic Commerce Search over traditional search setup methods.
E-commerce search interface showing product listings and category filters for country festival attire including ruffled mini dresses, midi dresses, and snip-toe boots.
Headshot of Shirley Gao

With Agentic Commerce Search, we deliver highly relevant, seamless discovery experiences that set a new standard for modern retail.”

Shirley Gao
Chief Information Officer, Pacsun

Agentic Commerce Search FAQs

Agentic Commerce Search is an AI-native search and discovery engine purpose-built for the complex demands of modern retail. Instead of merely matching keywords, it natively interprets natural queries and brings context-aware shopping experiences directly to your site. Guided Discovery delivers these rich, conversational shopping experiences directly within your Agentforce Shopper Agent.

Agentic Commerce Search builds a custom small language model trained to each customer’s data. This AI model powers the search, browse, and recommendations experience on your site. It is also integrated into Agentforce Shopper Agent, powering the search and conversational experience that

Agentic Commerce Search builds a commerce-optimized small language model (SLM) built specifically for your brand. This is an SLM trained on your specific catalog and shoppers that understands intent and delivers a natural language experience for how people shop today. Unlike a large language model (LLM) that is 7B+ parameters, this is a 1-7B parameter model that is highly optimized with your data and built for speed. This maintains an optimal site experience and fast speed on your site, delivering an optimal shopping experience.

Yes, ecommerce customers on any platform can get started with Agentic Commerce Search. Agentic Commerce Search is natively integrated into Salesforce B2C Commerce customers, enabling customers to get started with a toggle. Customers on other ecommerce platforms — such as Shopify, CommerceTools, SAP Hybris, Adobe Commerce, or their own homegrown platform — can integrate Agentic Commerce Search to their site with our headless APIs.

Getting started with Agentic Commerce Search is easy. You just need to provide your product catalog feed in any format and your shopper activity data from Salesforce Commerce Cloud or your ecommerce platform. There is no need to modify your catalog or create a highly attributed catalog nor is there a minimum traffic threshold or data prerequisite. From there, our AI automatically infers missing attributes and begins the synthetic simulation process to build your custom small language model that understands intent.

Customers can use Agentic Commerce Search no matter what platform they run their ecommerce business on. For Salesforce Commerce Cloud customers, Agentic Commerce Search is directly integrated into B2C Commerce. Customers can toggle it on with the click of a button and leverage their existing product catalog in B2C Commerce, giving them faster time to value without the need for hands-on engineering work. Customers powering their ecommerce site on other platforms — such as Shopify, Adobe Commerce, SAP, Commercetools, or their own home grown platform — can seamlessly integrate Agentic Commerce Search into any front end using our headless APIs.