Start with Google Merchant Center, extend the same feed to Microsoft Merchant Center, then make sure your catalog is ready for ChatGPT. US shoppers now discover products in three very different places: Google, Microsoft’s Bing and Copilot, and ChatGPT. Each receives your catalog differently, but all three run on the same raw material.
This comparison covers where each one shows your products, what it needs from you and how to decide the order for your own store. If you would rather have the foundation built for you, see our Google Merchant Center agency page.
The short answer
Google Merchant Center comes first because it has the largest reach and the strictest data requirements. Meeting its standards prepares you for the rest, so the order is also the cheapest path: do the hard cleanup once, then reuse it. Then extend the same feed to Microsoft Merchant Center. Then make sure your catalog is ready for ChatGPT and other AI assistants. One clean catalog serves all three.
Side by side
| Google Merchant Center | Microsoft Merchant Center | ChatGPT Shopping | |
|---|---|---|---|
| Where products show | Google Search, Shopping tab, Images, YouTube, AI Mode and Gemini | Bing, Microsoft Shopping and Copilot | Product recommendations inside ChatGPT |
| How products get in | Product feed, API or platform integration | Product feed or import from Google Merchant Center | Through platform catalogs and merchant programs, which are still evolving |
| Free visibility | Yes, free listings | Yes, in Microsoft Shopping surfaces | Recommendations are not paid placements |
| Ads | Shopping and Performance Max | Microsoft Advertising Shopping campaigns | Not the main model today |
| Data quality bar | High: strict policies and review | Similar to Google | Depends on the structured data the assistant can access |
| Effort if you already have a Google feed | Already done | Low: import, then verify | Depends on your platform; mostly data cleanup |
The ads row is a description of the channels, not part of our service: Avafa works on feeds and catalog data and does not manage ad spend.
Google Merchant Center: the foundation
Google Merchant Center is where most stores already have a feed, and its requirements are the most detailed: identifiers, attributes, shipping, returns and policy compliance. A feed that passes Google’s review is usually ready for everything else.
When we look at a Google feed, three things tell us most. First, how many products carry a brand and a GTIN where one exists. Second, whether titles contain the product type and key attribute rather than only a model name. Third, whether shipping and return settings reflect what the store actually does. A feed can be approved and still be weak on all three.
Microsoft Merchant Center: the easy second step
Microsoft Merchant Center can import an existing Google Merchant Center setup, which makes it one of the fastest ways to reach more shoppers. It also feeds Copilot, Microsoft’s assistant, which launched in-chat checkout in January 2026. Our Microsoft Merchant Center and Copilot setup guide covers the steps.
An import is a starting point, not a finish line. After it runs, compare a sample of products in Microsoft against your site: price, availability, shipping and images. Then review any products Microsoft flags, because the two platforms apply similar but not identical rules. Details on the service side are on our Microsoft Merchant Center agency page.
ChatGPT: the newest and least settled
OpenAI launched Instant Checkout in ChatGPT in September 2025, but its shopping plans have changed since then. What matters for stores is being recommended when shoppers ask for products, which depends on accessible, complete product data. Unlike the other two, there is no single account to open and no feed screen to fill in, which is why the work here is mostly preparation. See how to get your products into ChatGPT Shopping for the routes that exist today.
How do I decide the order for my store?
Score your store against five criteria. Most stores end up with the same order, but the table shows where yours might differ.
| What to check | Why it matters | What good looks like | Common mistake |
|---|---|---|---|
| Google feed health | Every other channel inherits its problems | Few disapproved products, identifiers and attributes filled | Treating «approved» as «good» |
| Where your US buyers search | Decides which channel is worth the effort first | You know your traffic split by source and device | Assuming Bing traffic is negligible without checking |
| Attribute coverage | AI assistants match requests against attributes | Size, color, material and use filled for most of the catalog | Leaving attributes in the description text only |
| Platform support | Your ecommerce platform may handle part of the distribution | You know what your platform offers for Google, Microsoft and AI assistants | Building a custom feed that duplicates a built-in one |
| Capacity to maintain | Price and stock need to stay in sync | One owner and an update process for the feed | A one-time setup that nobody maintains |
What does the sequence look like in practice? (A hypothetical example)
Imagine a distributor with a few thousand SKUs, a Google feed that is approved but thin, and no Microsoft account. This is an illustration, not a real account.
- Phase one, Google: audit the feed, fix titles, identifiers and shipping on the best-selling categories, and set a regular update.
- Phase two, Microsoft: import the improved feed, verify a sample against the site and review flagged products.
- Phase three, ChatGPT readiness: confirm crawler access and structured data on product pages, check what the ecommerce platform offers for AI assistants, and make sure the catalog attributes are filled.
Notice that phases two and three are small because phase one was done properly. If the distributor had started with ChatGPT, it would have discovered the same data gaps with fewer tools to fix them.
What should I ask my team or agency before starting?
- Which system is the source of truth for price and stock, and how often does it reach the feed?
- Who owns the feed after setup?
- Will Microsoft be an import of the Google feed, or a separate feed?
- What is the plan for products that fail review?
- How will we measure each channel separately in analytics?
- What exactly is being promised for AI channels? Any answer that includes guaranteed placement is a red flag.
When does this approach not make sense?
If you sell mainly through one marketplace and have no store of your own, the work moves to that marketplace’s listings. If your products have no public prices, the shopping feeds will not fit. And if your catalog is very small, doing Google well and checking the others manually may be enough.
How they connect
All three run on the same raw material: your product catalog. Build it once, well, and distribute it. That is the idea behind one product feed for every AI channel. For how each assistant behaves with shoppers, read ChatGPT vs Gemini vs Copilot, and for the bigger picture, agentic commerce explained.
How Avafa works
Avafa is a consultancy based in Spain that works remotely with US ecommerce brands, manufacturers, distributors and retailers. We audit the catalog, build the foundation in Google Merchant Center, extend it to Microsoft Merchant Center, and prepare the data for AI assistants. We do not promise placement in AI answers and we do not manage ad spend. Our Google Merchant Center agency service builds that foundation, and the ChatGPT Shopping agency page covers the AI-channel side.
FAQs
Can I copy my Google feed to Microsoft Merchant Center?
Yes. Microsoft Merchant Center offers an import from Google Merchant Center, so you don’t need to rebuild your feed. Check a sample after the import, since the platforms’ rules are similar but not identical.
Do I pay to appear in ChatGPT?
Product recommendations in ChatGPT are not sold as ad placements. Visibility depends on your product data and how accessible it is.
Which one should I do first?
Google Merchant Center. Its standards are the strictest, so meeting them prepares you for the rest.
Do I need all three?
Not necessarily. Google is the base for almost every store, while Microsoft and ChatGPT depend on where your buyers are. Because the catalog is shared, adding a channel later costs much less than starting from scratch.
Will a perfect feed get me into ChatGPT?
No feed can guarantee that. A complete, accurate catalog makes it possible for the assistant to match your products, but the assistant decides what to show.
How often should the feed update?
As often as your price and stock change. If they change several times a day, a feed that updates once a day will regularly disagree with your site.
