Do you have a Google Merchant Center strategy?
Research indicates that Google Gemini heavily weights any product-related data point to one of these two centers, with the data in the Manufacturing Center overriding the Merchant Center if there is a conflict. The graphic below shows that Google Merchant Center is a foundational layer augmented by Layer 2 sources, and it overrides everything else for Google queries—and possibly other LLM queries as well.
The three-layer commerce and product discovery tech stack
Where you compete is no longer a single layer. The stack has three interdependent layers, and weaknesses in any one of them collapses the whole.
The data feed powering Gemini and Copilot
Most manufacturers we talk to have not set these up, or set them up years ago and have not updated them for the age of AI. Few have a documented strategy for what data flows in, how it is structured, or how to measure whether it is working. And that gap is widening.
FEED 01
Google Manufacturer Center
BRAND DIRECT MANUFACTURER FEED
Authoritative product attributes published by the brand: titles, descriptions, GTINs, rich attributes, brand-canonical imagery.
Feeds Gemini and Google AI Mode’s understanding of what a product is. Often underused.
FEED 02
Google Merchant Center
RETAILER-PUBLISHED FEED
Product offers with price, availability, retailer-specific date. Powers Shopping ads and Google’s commerce surfaces.
If retail feeds disagree with your Manufacturer Center data, AI sees inconsistency and discounts authority.
FEED 03
Bing & Microsoft equivalents
MICROSOFT MERCHANT + SHOPPING
Bing parallel infrastructure feeding Copilot’s product knowledge in commerce contexts.
Copilot (embedded in Office, Edge, Windows) pulls from this. B@B teams touch it whether they know it or not.
FEED 04
Schema.org & structured data
YOUR OWN WEBSITES & PDPs
Schema.org Product, Offer, AggregateRating, and related types embedded in your product detail pages.
The connective tissue. Without schema, your site is text. With it, your site is data.
Merchant Center and Manufacturer Center are not the same
The two get confused constantly, and most teams treat them as one thing. They are not. Think of Merchant Center as your digital storefront, and Manufacturer Center as the official encyclopedia for those same products. You need both, and each does a different job.
The two usually work side by side through the Global Trade Item Number (GTIN), the unique barcode number for a product. When a retailer lists your product in Merchant Center, they supply the GTIN. Google matches it to your approved data in Manufacturer Center. As a result, the store shows current pricing, and Google shows your brand’s official details and imagery.
Here is why Manufacturer Center matters most. As mentioned, our research indicates that Google Gemini heavily weights product data from these two centers, and Manufacturer Center overrides Merchant Center when the two conflict. That makes Manufacturer Center the foundational layer. It’s the brand-authoritative record that controls what AI says about your products, with Layer 2 sources building on top of it. It is not the only signal Google reads, and no one can say exactly where the ranking starts.
Here’s what we can say with confidence. When your Manufacturer Center record is accurate, Google and Gemini have the best chance of getting your product right on Google queries and likely other AI queries, as well.
Setup, syndication, and end-to-end tracking
Ntara’s Google Merchant Center onboarding service gets companies properly configured across the Google Commerce Graph and builds a measurement baseline to track results. This type of engagement typically runs two to four weeks and includes four core workstreams.
Setup
This is for a net new setup or review of an existing Merchant Center or Manufacturer Center account. We diagnose the current state, identify gaps, and configure or reconfigure to align with best practices for AI surface syndication.
Data export
We configure a data export via flat file or REST API. The feed covers six required data fields. It also covers optional fields that improve the quality of AI recommendations, including a canonical image URL and rich attribute data.
AEO monitoring
We establish a baseline for monitoring AI engine optimization (AEO) using off-the-shelf tools or a client-provided baseline. This tracks LLM mention volume and recommendation accuracy on key platforms before and after the feed goes live.
Expansion
After Google Merchant Center and Manufacturer Center are in place, we expand to Bing Merchant Center and then explore emerging agentic checkout protocols (ACP, UCP) as the landscape matures.
Engagement structure
Three phases. Four weeks.
PHASE 1
Discovery
WEEKS 0-1
- Stakeholder kickoff
- Intake questionaire (all functions)
- Document & system access
- 6-10 stakeholder interviews
PHASE 2
Diagnostic
WEEKS 1-2
- Technical & data environment review
- Content & product data audit
- Scoring vs. Ntara readiness framework
- Gap analysis across all 4 dimensions
PHASE 3
Synthesis & readout
WEEKS 3-4
- Internal synthesis & report development
- Executive readout workshop (half-day)
- Scored readiness report delivery
- Prioritized roadmap delivery
What this looks like in practice
This onboarding service is designed for companies that want to take control of their Google Merchant Center accounts and improve the accuracy of AI recommendations in their category. This is a good fit if:
- You have a Merchant Center account but no documented strategy or measurement baseline
- AI referral traffic has changed and you do not know whether your feeds are contributing
- Your distributors are outranking you in AI-generated recommendations
- You completed an agentic commerce readiness audit and data feeds were flagged as a gap
Ready to get your product data in front of AI?
Let us scope the engagement. The first conversation starts with your current Merchant Center state.
