Table of Contents
- Products You Never Uploaded: How Automatic Discovery Works
- The Upside: A Safety Net for Small Stores
- The Downside: You Lose Control of Your Product Data
- How to Audit What Google Found
- Taking Control: Feed First, Crawl as Backup
- Platform Notes: Shopify, WooCommerce and Custom Sites
- Can You Turn Off Automatic Discovery?
- Quarterly Discovery Audit Checklist
- An Example of How It Goes Wrong
- Why This Matters More as AI Shopping Grows
- Frequently Asked Questions
Products You Never Uploaded: How Automatic Discovery Works
Many merchants open Merchant Center and find products they never submitted. Google's help center explains why: it can automatically find and add products from your online store by discovering product pages, extracting product details such as title, price and availability from the structured data on those pages, and submitting that information to your Merchant Center account for use in Shopping ads and other surfaces. Merchants sometimes see this labeled 'Found by Google' in the interface, and industry coverage often calls it a feedless setup.
To make it work, Google says you need to use structured data to add detailed product information to your site, allow Google to collect that information, and verify and claim your website by linking it to your Merchant Center account. Search Engine Land reported on the original rollout in 2024, noting that products can be added or removed automatically based on the crawl, and that the feature complements other data sources without duplication. A June 2026 French-language industry article describes the same three steps: discovering product pages, reading the data on them, and sending it to Merchant Center.
Process details come from Google's 'Add products automatically from your online store' help page and Search Engine Land (June 14, 2024). Whether automatic discovery is active in your account depends on your settings and website; check Merchant Center rather than assuming.
The Upside: A Safety Net for Small Stores
For a small store without the time or technical skills to manage a feed, automatic discovery is a genuine on-ramp. It gets products into Merchant Center quickly, powers free listings and can serve as a fallback if your regular feed breaks. Because it relies on schema markup, it also rewards stores that maintain good structured data, which helps with AI surfaces too (see structured data for AI shopping).
The Downside: You Lose Control of Your Product Data
Relying on the crawl exposes several compliance and performance risks:
- Sparse attributes. Markup rarely includes GTINs, brand, custom labels, product category, shipping overrides, or identifier exceptions, which feed-based setups handle explicitly. Missing attributes limit performance and can lead to warnings.
- Errors are copied faithfully. Whatever your page says in markup, including a stale price or a placeholder description, flows into Merchant Center.
- Unexpected products. Test pages, hidden products, internal items or discontinued pages that carry Product markup can be discovered and show up as offers.
- Policy exposure. Items you would never choose to advertise, such as restricted products, can enter your account and trigger disapprovals or account-level warnings.
- No explicit data source ownership. When troubleshooting, it is harder to know which source produced a value.
A 2026 blog post from Productrise argues merchants should not rely on feedless discovery for organic shopping visibility, for similar reasons: you want your most important data to be deliberate and complete. That is an opinion, but it matches what we see in suspension cases where unknown items were the trigger.
Google evaluates the items in your account against policy regardless of how they arrived. If a crawled page lists a restricted item or a misleading price, it counts against your account health.
How to Audit What Google Found
Do this once, then quarterly:
- In Merchant Center, open the Products area and filter by data source. Look for the source labeled as automatic or website-crawled alongside your feeds.
- Export the full list of offers from that source and compare it with your real catalog. Flag anything you do not recognize.
- For each unknown item, find the landing page. Common causes are old campaign pages, test products, bundles, gift cards and out-of-catalog accessories.
- Decide: fix the page, remove Product markup from it, block it from indexing if it should not exist, or add it properly to your feed.
- Check Diagnostics for issues attached to the crawled data source. These often highlight missing identifiers or mismatches.
See the Diagnostics guide for navigation, and approved products with no impressions if the discovered items are approved but silent.
Taking Control: Feed First, Crawl as Backup
The best-practice stance is: submit a real feed with complete attributes as your primary source, and treat the crawl as a backup that fills gaps. When both sources contain the same product, Google says automatic discovery complements other data sources without duplicating them, but conflicting values still have to be reconciled, so keep your site markup and feed aligned.
- Include GTIN, brand and MPN where they exist; see feed optimization.
- Use supplemental feeds for custom labels, promotions and shipping overrides.
- Make sure the shipping and return information in Merchant Center matches your policy pages. See return policy requirements.
- Maintain accurate Product schema on every product page, since it is the source for crawled data and for automatic item updates (covered in this guide).
Platform Notes: Shopify, WooCommerce and Custom Sites
Hosted platforms usually output reasonable Product markup by default, which makes them prime candidates for automatic discovery. Themes and apps can add duplicate or conflicting JSON-LD blocks, though, and that is where discovered data goes wrong. For platform-specific setup, our guides for Shopify and WooCommerce cover feed configuration, and each should be paired with a markup validation.
Can You Turn Off Automatic Discovery?
Google's help documentation presents automatic product addition as an option tied to claiming your site and allowing collection, and Merchant Center lets you manage data sources. If you do not want crawled items in your account, review your data source settings and the relevant collection permissions, and remove the source from your account if you prefer an entirely feed-based setup. Settings and wording change over time, so follow Google's current instructions in the help center rather than a screenshot from a blog post, including ours.
Quarterly Discovery Audit Checklist
Keep crawled data under control
The automatic or website data source is identified in Merchant Center
Every discovered offer corresponds to a real, sellable product
Pages that should not be advertised carry no Product markup or are noindexed
Feed and markup agree on price, availability and identifiers
Diagnostics reviewed for issues on every data source
Restricted-category items are not present in any source
An Example of How It Goes Wrong
Imagine a store that sells skincare. Years ago a developer created a hidden landing page for a sample-size promotion, marked up as a Product with a placeholder price of $0.01. The page was never linked in navigation but was still indexable. Google's crawl discovers it, extracts the markup and creates an offer. The next time a reviewer samples the account, a $0.01 item with a mismatched real price at checkout appears. That single item can trigger a price accuracy flag, and if the store has several like it, the pattern looks like deliberate manipulation.
None of this requires bad intent. It only requires a website with leftovers. Which is why a data source audit belongs in every quarterly review, and why checking your site with the same eyes a crawler uses is so valuable. Search your own site for pages with Product schema that are not in your catalog, and clean them up.
Quick ways to find stray Product pages
- Search Google for site:yourdomain.com with words like sample, test, gift, hidden or draft.
- Crawl your site with a free crawler and filter for pages containing Product schema.
- Compare the sitemap with your product catalog export.
Why This Matters More as AI Shopping Grows
AI-driven shopping surfaces depend on machine-readable product data. Merchants whose only source is the crawl are at the mercy of whatever their markup says, which is often thin. A deliberate feed with rich attributes, supplemented by accurate markup, gives AI surfaces more to work with. See the conversational attributes guide for the next layer of detail that only feed-based merchants can add.
Frequently Asked Questions
Why are there products in my Merchant Center I never uploaded? Google may have discovered them by crawling your site and reading structured data, if automatic product addition applies to your account.
Do I still need a feed if Google finds my products? For serious advertisers, yes. A feed gives complete control over identifiers, labels, shipping and promotions that markup rarely covers.
Can discovered products cause suspensions? They are evaluated against the same policies as any item, so a restricted or misleading discovered offer can create issues.
How do I fix wrong data in discovered products? Fix the markup on the landing page, since that is the source, and consider submitting a feed that provides correct values.
See What Google Sees on Your Store
Wrong markup becomes wrong product data. A free scan at gmcunbanned.com checks your store for the policy and landing page issues that most often trigger suspensions.
Run Free GMC Scan →