To find gaps in a Shopify product catalog, start where Shopify and Google already report them: the Agentic sales channel status, the Shopify Catalog requirements, store policies, field coverage on the Channel mapping page, product categories, variant grouping and the Merchant Center data rules. Read individual descriptions only after those checks, because better copy cannot replace eligibility on a channel that uses Shopify Catalog. It can still help web search, crawling and other discovery routes.
I'm Felix-Sebastian Ament, one of the two managing directors of Commerce Labs GmbH, the company that builds and operates Sphen.ai, a Shopify app that looks for exactly these gaps. Treat my order of work as a vendor's opinion. Every check below works with Shopify's and Google's own tools, without my product.
Overview: what a catalog gap is
I call a gap any missing, wrong or contradictory value that stops a channel from including a product, classifying it correctly or describing it accurately. I sort gaps into three levels:
- Exclusion: the product does not reach Shopify Catalog, the structured source that AI channels such as ChatGPT and Microsoft Copilot read.
- Misclassification: the product is included, but its category, variants or identifiers are wrong.
- Thin data: the product is classified correctly, but its description lacks the facts buyers ask about.
I search in that order to avoid treating copy as a remedy for a Catalog exclusion. The exclusion limits that data route; it does not erase the value of better copy for web search or other discovery routes.
Separate a market restriction from a product gap
Before counting product gaps, check the channel's market rules: ChatGPT requires US sales, and Copilot direct checkout is limited to US customers. Meta and Google have their own supported-market lists for direct checkout; these lists do not define every form of AI product discovery. Review the supplemental terms and store policies too. A channel restriction is a separate finding from a missing product field.
Seven places to look
1. The Agentic sales channel.
Under Sales channels > Agentic, Shopify shows either the number of products you have in Shopify Catalog or an Ineligible warning. Clicking it shows the reason, for example a password-protected store or a plan that does not support Catalog access. The full list of messages is in Shopify's Catalog requirements.
For an eligible store, my first calculation is simple: active products in the store minus products in Catalog. This is only a starting comparison, not a count of missing products.
First check that both figures use the same unit and publication scope. Combined Listings or custom grouping can combine products into one listing, and active but unpublished products can also distort the comparison. The list comes from checks 2, 4 and 5.
Alongside the requirements in check 2, investigate excluded categories, Shopify's own category classification, sensitive content and detected B2B-only products. These are examples, not a complete cause list. Account standing instead determines eligibility for the whole store; an ineligible store shows the reason rather than a Catalog product count.
2. Per-product requirements.
Shopify requires a title, at least one product image and a price greater than 0. The product must be published to the online store, Hydrogen or Headless channels with an identifiable product URL. Hydrogen and Headless stores must use the route format Shopify specifies. On the Agentic plan without a published online store, products need the external product URL instead. Unlisted products, products hidden from search engines and sensitive or mature content are excluded. I check each of these conditions for every active product before anything else.
3. Store policies.
Every channel in Shopify's agentic storefronts requires completed Terms of service, Privacy policy and Return and refund policy. This is a channel requirement, not a per-product gap or a claim about the Catalog product count.
4. Field coverage on the Channel mapping page.
On the Channel mapping page for Shopify Catalog, each field shows a percentage: the share of your products that have a value for it. For Shopify Catalog, Shopify counts all of your products here, not only active ones. Fix [number] products opens the products without a value. In my opinion this is the fastest gap report Shopify offers today.
5. Product category.
Shopify asks you to assign every product to a category from its standard product taxonomy, and some existing products may be listed as uncategorized and need one. This matters twice. The category unlocks category metafields such as size, fabric or color. And Shopify excludes products it classifies as Uncategorized from Shopify Catalog. That classification uses your product information, so it can differ from the category in your admin.
6. Variants.
If you set up custom variant grouping on the Channel mapping page, Shopify checks all products and can report [number] need review. That means two variants in a group share the same option combination. I also look for products where no variant has a price above 0, variants without a SKU, and products with at least two sellable variants whose names are all Title or Default Title. For another data check, Sphen's variant checks report four findings: no positively priced variant on a product, a missing variant SKU, generic names on all of at least two sellable variants, and unavailable variants on draft or archived products. For the generic-name finding, every sellable variant must be named Title or Default Title; one-variant products are exempt. Shopify's duplicate-option result is not one of the four. You fix all of these in Shopify.
7. Merchant Center consistency.
Google's product data specification requires the price to match the landing page and checkout. Availability must also match. When structured data is present, check its price and availability too, using Google's landing page requirements. Google's guide for the Google & YouTube app on Shopify also rules out placeholder brands such as N/A or Generic. Missing or incorrect GTINs can limit visibility. If a product has an assigned GTIN and you omit it, Google says it could be disapproved. This matters for AI shopping because Shopify makes products available to Google AI Mode and Gemini through Merchant Center. Your products can reach Merchant Center through the Google & YouTube sales channel, with automatic product and shipping sync turned on, or through a direct connection by a third-party feed provider or manual upload.
Before you trust the difference from the first check, remove the products Shopify will never pass on. Shopify excludes whole categories from Shopify Catalog, such as alcohol, tobacco and weapons, and also product add-ons such as extended warranties, installation and shipping insurance. Those products do not belong on your gap list. If you leave them in, the gap looks larger than anything you can fix.
If you also upload a product feed to OpenAI directly, add one more check. OpenAI's product feed schema asks you to omit unknown optional values instead of writing placeholders such as null, unknown or n/a. I count a placeholder as a gap unless the schema explicitly permits it, as it does for unknown availability.
How I sort the list afterwards is my own rule: first by level (exclusion before misclassification before thin data), then by revenue within each level. A bestseller without a category is more urgent than a hundred clearance items without alt text. If you have no revenue data at hand, start with the products buyers ask about most.
Which fields matter
Shopify's guidance on optimizing products for AI platforms lists details AI platforms and shopping sites consider, including: title, description, images, type, vendor, collections, tags, barcode and variants including option names.
When I rank gaps, I weight those fields differently. This is my working rule:
| Gap | Why I rank it high | Who closes it |
|---|---|---|
| No image, price 0, not published | Product is missing from Shopify Catalog | Store team, by hand |
| Missing or wrong category | Possible exclusion, no category metafields | Store team, picking a real taxonomy entry |
| Missing GTIN where one exists | Google: limited visibility or disapproval | Manufacturer or inventory data |
| Conflicting price or stock | Google requires matching values | Settle the source system, then sync |
| Generic option names | Variants cannot be told apart | Store team |
| Thin description | Fewer facts for answers | Draft possible, approval required |
One rule I do not bend: no AI should invent a GTIN, vendor, price or availability value, or a category outside Shopify's taxonomy. Choose a real taxonomy entry and check that it matches the product. Google tells you not to guess a GTIN, and an incorrect GTIN can limit visibility or lead to disapproval.
How I measure gaps and impact
A gap only becomes a number when it has a denominator. I express every gap as a share and write the base next to it: share of active products without a barcode is a metric you can repeat; lots of products lack barcodes is not. Keep one base per metric. Shopify's field coverage counts all of your products, so a share of active products will not match it. I record both and compare each only with itself.
Then I measure impact on three levels:
- Data: field coverage on the Channel mapping page and the number of products in Shopify Catalog, before and after the fix. Shopify notes that mapping changes are processed with a delay, so I record when a mapping change was made and compare results after it has been processed.
- Answers: I ask assistants real buyer questions over several runs and record whether the store is mentioned. Single answers vary. I read the direction.
- Revenue: orders from AI channels appear in the Shopify admin with channel or referrer attribution. This does not capture AI-influenced orders without retained attribution.
Sphen's readiness score is one way to keep the data level on a single number: a score out of 100, the count of products not yet AI-ready, and the top fixes ranked by impact.
Decision table by store size and data maturity
How I would organize the search:
| Starting point | How I would search |
|---|---|
| Few products, one range | Walk the seven places by hand and keep one list |
| Mid-size catalog, data only in Shopify | Field coverage and categories first, then variants, then descriptions ranked by revenue |
| Large catalog with a PIM or ERP | Close gaps at the source, not in Shopify, and map Catalog fields to maintained metafields |
| Large catalog, inconsistent data | Use a tool for the counting, review drafts in batches, measure each change |
| Mostly selling to businesses | First check whether agentic storefronts apply at all |
Key takeaways
- Look for exclusions first, then misclassification, then thin descriptions.
- A difference between active-product and Catalog counts is a signal to investigate. Check publication scope and grouping before counting missing products; the per-product checks produce the list.
- Field coverage on the Channel mapping page shows, per field, the share of all your products that have a value.
- Products that Shopify classifies as Uncategorized are excluded from Shopify Catalog. That classification can differ from the category in your admin.
- Count forbidden placeholder values such as n/a as gaps. Keep values the schema explicitly permits, such as OpenAI's unknown availability.
Negative fit: when a systematic gap search is not worth it
- Very small catalogs. If you can open every product in one sitting, a process costs more than it returns.
- B2B-only selling. Shopify's agentic storefronts support only direct-to-consumer sales and exclude B2B-only products when Shopify can identify them.
- Excluded categories. Shopify does not allow categories such as alcohol, tobacco or weapons in Shopify Catalog. A gap search aimed at AI shopping will not change that.
- No agreed source of truth. Until someone decides whether the ERP, the PIM or Shopify holds the correct price, you will keep finding conflicts you cannot close for good.
Next step
Run the seven checks on your own store and write each gap down as a share, with its base named. If you want the counting done for you, Sphen's catalog audit reads your store without changing it and returns the readiness score, the number of products that are not AI-ready and your top fixes. Proposed fixes wait in Recommendations until you approve them. Verified source data is needed for GTIN and vendor. Check the category against the standard taxonomy in Shopify admin. Budget for that check separately from copy review.
Start with Shopify's free Agentic status and field coverage checks, then record the first gap and who will fix it.
Frequently asked questions (FAQ)
What is the fastest way to see which products are missing from Shopify Catalog?
First compare your number of active products with the number Shopify shows for Shopify Catalog under Sales channels > Agentic. The difference alone is not a count of missing products: first check grouping, publication scope and excluded products. To find which products are missing, check the per-product requirements, field coverage and product category.
What does field coverage on the Channel mapping page mean?
It shows, for each field, the share of your products that have a value. For Shopify Catalog, Shopify counts all of your products, not only active ones. Fix [number] products opens the products without one.
Why does an Uncategorized product matter?
Shopify excludes products it classifies as Uncategorized from Shopify Catalog. Its classification uses product information and can differ from the category set in the admin. The admin category also controls the available category metafields.
Should I fill missing GTINs with AI?
No. Google says to submit a GTIN only if you are sure it is correct, and an incorrect GTIN can limit visibility or lead to disapproval.
Sources
- Shopify Catalog — Shopify Help Center
- Requirements for being included in Shopify Catalog — Shopify Help Center
- Customizing your product data for Shopify Catalog — Shopify Help Center
- Optimizing your products for AI platforms — Shopify Help Center
- Shopify agentic storefronts — Shopify Help Center
- Shopify Catalog and product discovery for agentic storefronts — Shopify Help Center
- Selling on Google AI Mode and Gemini — Shopify Help Center
- Product category — Shopify Help Center
- Product data specification — Google Merchant Center Help
- Product feed reference — OpenAI Developers
- Variant checks — SPHEN.AI
- Readiness score — SPHEN.AI
- Catalog audit — SPHEN.AI
- Reviewing and applying recommendations — SPHEN.AI Help
- Shopify: Selling on ChatGPT — Shopify Help Center
- Shopify: Selling on Microsoft Copilot — Shopify Help Center
- Shopify: Selling on Meta — Shopify Help Center
- Google Merchant Center: Landing page requirements — Google
- Google Merchant Center: Product data quality for the Google & YouTube app — Google
- Google Merchant Center: GTIN — Google