Merchant Center Business Agents: Complete Guide

Merchant Center Business Agents are Google's branded AI chat experience that lets shoppers have a conversation with your brand directly on Google Search. Powered by Gemini, these agents pull from your Merchant Center product feed and website data to answer questions, recommend products, and guide purchase decisions. Google launched Business Agents in January 2026 with early adopters including Lowe's, Reebok, and Poshmark.

Instead of a static product listing, shoppers can now ask questions like "What paint works best for a bathroom?" or "Do these shoes run true to size?" and get answers that come from your catalog and brand knowledge. It's a fundamental shift in how product discovery works on Google.

What Business Agents Do

Business Agents appear as a conversational chat widget in Google Search results. When a shopper searches for products in your category, they may see an option to chat with your brand's agent. The experience is powered by Google's Gemini model and grounded in two data sources: your Merchant Center product feed and your website content.

Think of it as a virtual sales associate that knows your entire catalog. The agent can:

• Answer product questions — materials, sizing, compatibility, use cases
• Recommend products based on the shopper's stated needs
• Compare items within your catalog to help narrow down options
• Provide availability and pricing pulled directly from your feed
• Hand off to human support when the conversation requires it

The agent carries your brand's visual identity — your colors, logo, and tone — so shoppers know they're interacting with your brand, not a generic Google experience. As JumpFly noted, it's like hiring a sales associate that's available 24/7 and knows every SKU in your catalog.

Who Can Use It

Business Agents are currently limited in availability. To be eligible, you need:

• US-based operations — the feature is US-only as of March 2026
• A verified Merchant Center account in good standing
• At least 50 approved product offers in your feed
• A claimed brand profile in Merchant Center

Early adopters that Google highlighted include Lowe's, Michael's, Poshmark, and Reebok. As Search Engine Roundtable reported, the feature has been rolling out gradually since the initial announcement, with more retailers gaining access over time.

How to Set Up Business Agents

If your account meets the eligibility requirements, you can find the Business Agent setup in Merchant Center under Marketing > Business Agent > Customize. As PPC News Feed covered, the customization process is straightforward with four main areas to configure:

• Welcome message — the greeting shoppers see when they open the chat. Set expectations for what the agent can help with.
• Conversation starters — suggested questions that appear as clickable prompts. Use these to guide shoppers toward common queries your catalog handles well.
• Brand colors — the chat widget matches your brand's color scheme, creating a cohesive experience.
• Support handoff — configure when and how the agent escalates to your human support team. You can set triggers for specific question types or let the agent decide when it can't provide a confident answer.

Until you customize the agent, Google shows a standard Google-branded version. It still works — pulling from your feed and website data — but it won't carry your brand identity. Customizing the agent is free and takes about 15 minutes.

New Data Attributes for AI Discovery

Alongside Business Agents, Google announced dozens of new Merchant Center data attributes designed specifically for conversational commerce. As Search Engine Journal reported, these attributes go well beyond traditional keywords and product descriptions.

The new attributes include:

• Common product questions and answers — structured Q&A pairs that the agent can reference directly
• Compatible accessories and add-ons — which products work together, enabling cross-sell recommendations
• Substitute products — alternatives the agent can suggest when an item is out of stock or not the right fit
• Use case descriptions — specific scenarios where the product excels, helping the agent match products to shopper needs
• Care and maintenance instructions — post-purchase information that builds buyer confidence

These attributes feed not just Business Agents but also AI Mode, Gemini, and other conversational surfaces across Google. The direction is clear: product data is becoming the foundation for AI-driven shopping. Retailers who enrich their feeds with these attributes early will have an advantage as these surfaces scale.

What's Coming Next

Google has outlined a roadmap for Business Agents that extends well beyond the current chat-and-recommend functionality. As Chain Store Age covered, the planned capabilities include:

• Train the agent on your own data — upload brand guidelines, FAQ documents, and product knowledge bases to make the agent more accurate and on-brand
• Customer insights and analytics — see what questions shoppers are asking, where conversations drop off, and which products generate the most engagement
• Product recommendations and offers — the agent will be able to surface promotions, bundle deals, and personalized offers during conversations
• Agentic checkout — shoppers will be able to complete purchases directly within the chat, without leaving the conversation

The checkout capability is particularly significant. It means the entire purchase funnel — from discovery through Q&A to transaction — could happen within a single conversational thread on Google Search. As Business Tech Weekly noted, this represents a fundamental rethinking of how online shopping transactions work.

How This Connects to Your Shopping Strategy

Business Agents are part of Google's broader agentic commerce push alongside Direct Offers in AI Mode and the Universal Commerce Protocol (UCP). For advertisers already running Shopping campaigns, Business Agents add a conversational layer on top of your existing product data.

The critical point: your product feed data becomes the agent's knowledge base. Every product title, description, attribute, and image in your Merchant Center feed is what the agent draws on to answer shopper questions. Incomplete or poorly structured feed data means the agent gives incomplete or poorly structured answers.

This raises the bar on feed quality in specific ways:

• Product titles need to be descriptive, not just keyword-optimized. The agent uses titles to understand what a product is and how to describe it in conversation. A title stuffed with keywords but lacking clear product identification will confuse the agent. Use a title optimization tool to balance SEO with clarity.
• Descriptions matter more than ever. For standard Shopping ads, many advertisers treat descriptions as an afterthought. For Business Agents, descriptions are a primary data source. Write them for humans, not algorithms.
• Attribute completeness drives answer quality. The more attributes you fill in — materials, dimensions, compatibility, care instructions — the more questions the agent can answer confidently. Feed quality monitoring helps you identify gaps.
• Track what's working at the product level. As Business Agents generate traffic, you'll want to understand which products are converting and which are wasting spend. Product-level analytics give you that visibility.

If you're already investing in feed optimization, you're building the foundation that Business Agents need. The retailers who will benefit most from this feature are the ones with the richest, most complete product data.

Frequently Asked Questions