Catalog Optimization for Dynamic Product Ads: Google Shopping, Performance Max and Meta Advantage+

1. 01

   On Google Shopping or Meta Advantage+ catalog ads there's no headline to write. The feed row is the ad: title, image, price, attributes.

2. 02

   Google ranks Shopping ads on bids plus relevance, and it reads relevance from your product data. A vague title loses auctions that a precise one wins for less.

3. 03

   Fixing catalog data is the cheapest bid increase you'll find: higher relevance and click-through rate on the same budget.

4. 04

   Custom labels turn the feed into a bidding tool. Margin and bestseller status stop being spreadsheet columns and become asset groups in Performance Max and product sets on Meta.

5. 05

   AI shopping agents read the same catalog your ads do. The complete, consistent data that lifts ROAS is also what gets a product picked in an AI answer.

Most Google Shopping ads optimization advice starts with titles. We start with eligibility: a perfect title on a disapproved product earns nothing.

1. 1

   Fix eligibility

   In Merchant Center, go to Products and filter on Needs attention. Clear disapprovals before warnings: missing or invalid gtin, placeholder brand values like "Generic", price or availability that doesn't match the landing page, promotional overlays on images.

2. 2

   Rewrite titles

   Product type, brand and the defining attribute go in the first 70 characters. Use the rest of the 150-character limit for variant details such as size, color or material.

3. 3

   Categorize precisely

   Set google_product_category to the deepest correct level and build a 3 to 4 level product_type from your own taxonomy. It becomes your bidding and reporting structure.

4. 4

   Upgrade images

   Highest resolution, no text or watermarks, extra angles in additional_image_link.

5. 5

   Complete attributes

   Fill color, size, gender, age_group, material, pattern and item_group_id for variants. Use the exact color names that appear on the landing page.

6. 6

   Add differentiators

   Submit product_highlight, product_detail and, where it fits, question_and_answer to answer pre-purchase questions.

7. 7

   Label for bidding

   Add custom_label_0 to custom_label_4 for margin, bestseller status, season or price band, so campaigns can bid on business value.

The Google product data specification sets the hard limits: title up to 150 characters, description up to 5,000, up to 10 additional images, 2 to 100 product highlights of up to 150 characters each, up to 100 product_detail entries and up to 30 question and answer pairs. If your titles are AI-generated, Google asks for them in structured_title. To check a feed against these limits and the other requirements before you submit, use our Google Merchant Center requirements checklist.

A common apparel edge case: every size of a style carries the same GTIN, because the ERP stores the barcode at style level. Each size and color is a separate trade item with its own code. Fix it in the source system; a feed rule breaks at the next import.

Does Google Shopping show all products?

No. A product can sit in Merchant Center and never earn an impression, usually for one of four reasons.

• Disapproved. Google states that "disapproved products won't appear on Shopping ads or free listings" (Google Merchant Center Help). Invalid GTINs and price mismatches are typical causes.
• Limited performance. Warnings for missing brand, mpn, color, size, gender or age_group keep the product live with fewer impressions. Unresolved warnings can escalate to account review.
• Not relevant to the query. If the title and attributes don't describe what the shopper typed, the product isn't eligible for that search, whatever you bid.
• Starved by the campaign. Products outside a listing group, or behind a few budget-hungry bestsellers, get almost no traffic.

Merchant Center never flags that last one. Sort product-level reporting by impressions, lowest first, to find it.

How to rank in Google Shopping (free listings and ads)

Google says Shopping tab listings are ranked on "relevance to a user's current search terms", while Shopping ads mix bids with relevance. Both reward the same things: a title in the shopper's words, precise categorization, complete attributes, a correct GTIN and an accurate price. Our Google Merchant Center guide covers every attribute, and the product feed optimization guide shows how to run these fixes across channels from one source.

How Meta catalog ads work

Meta describes Advantage+ catalog ads as a way to automatically promote relevant products from an entire catalog, using pixel and app events to build audiences. They either retarget past viewers or prospect broadly for new buyers. Advantage+ sales campaigns took 20% of retail Meta spend in Q1 2026, down from a 38% peak a year earlier, while overall Meta spend grew 13% (Karooya). eMarketer reported that Advantage+ shopping campaigns delivered almost 22% higher returns than the average Meta ad.

Setup in five steps

1. Create a catalog in Commerce Manager and connect a scheduled feed, a platform integration such as Shopify, or the Catalog API.
2. Install the Meta Pixel and Conversions API with ViewContent, AddToCart and Purchase events. Meta requires that content_ids "must match the IDs found in your product catalog".
3. Check the catalog match rate in Events Manager and fix mismatched IDs. A product the pixel can't connect to the catalog can't be retargeted.
4. Build product sets with filters on product_type, brand, price, custom_label_* or internal_label, reusing the margin and bestseller labels from Google. Keep sets large, because products outside a set aren't considered.
5. Launch a sales campaign with catalog ads, then test price or discount overlays and video assets.

Step 3 is where we find the most broken setups: a catalog keyed on variant IDs while a third-party tag sends SKUs. Everything looks live, and retargeting misses most people.

Facebook product feed requirements

Meta's catalog field reference lists the required fields.

Meta accepts Google-format feeds, so one master catalog can serve both channels. Two hand-maintained feeds drift apart within weeks, so run both from one source, either your PIM or one of the product feed management tools for Google and Meta ads. Recommended fields that improve delivery include google_product_category or fb_product_category, color, gender, age_group, material, pattern, item_group_id, gtin, product_type, up to 20 additional_image_link images and custom_label_0 to custom_label_4. Meta also offers internal_label, which it recommends over custom labels for filtering product sets because it can be updated without triggering a policy review.

Instagram Shopping in 2026: catalog setup and product tagging

Instagram Shopping runs on the same Meta catalog. To tag products, the professional account needs an approved Instagram Shop linked to a catalog, and only products with an approved review status show up on posts (Meta for Developers). Feed posts take up to 20 product tags, carousels up to 20 (5 per slide) and Reels up to 30.

A product rejected in catalog review vanishes from tags and ads alike.

Measure by item ID. Campaign ROAS hides what happened, because budget shifts between products as their data improves. Compare four weeks before a change with four after, and log which fields changed when.

The link to AI shopping agents

AI shopping agents read the fields you fix for ads. Google states that product_highlight helps products get discovered across AI-driven surfaces such as AI Mode (Google Merchant Center Help), and Google AI Mode shopping draws on Merchant Center data. From the shopper's side:

Each reason maps to a field the old "Linen Shirt Sand" row lacked.

A catalog optimized for paid ads is most of the way to Agentic Commerce Optimization, and the PDP work in our product page optimization guide helps landing page experience and AI recommendations at once.