Product Feed Optimization for Google Shopping and Free Listings

What product feed optimization actually involves

Most stores send product data through a platform integration (the Google & YouTube app on Shopify or Google for WooCommerce), a file, a Google Sheets template or the Merchant API. Whichever route you use, the primary data source is the only one that can add or remove products, so the first question in any feed project is where each attribute really comes from. Edits made directly in Merchant Center can be overwritten by the next API or platform sync.

Titles and descriptions

Google’s title specification allows 1 to 150 characters and notes that shoppers usually notice the first 70 characters or fewer, so the most important details belong at the start. Titles need to describe the product on the landing page, distinguish variants, and avoid all caps for emphasis, symbols, promotional text, prices, shipping or delivery information and company names.

Descriptions can be up to 5,000 characters, with the most important details in the first 160 to 500. Google’s description rules ask you to describe only the product (not accessories, other products, business history or policies), avoid comparisons and links, and keep category paths in google_product_category or product_type instead. If titles or descriptions are AI-generated, Google asks for them in structured_title or structured_description with the digital source type trained_algorithmic_media.

Product identifiers

Unique product identifiers are assigned by the manufacturer and are identical across retailers. Google recommends submitting GTIN, brand and MPN where they exist, and notes that products with an assigned GTIN submitted without one may have limited visibility. Identifiers should never be guessed, made up or copied from similar products; incorrect values can cause disapprovals. See Google’s unique product identifiers guide for the full rules.

Situation	What to submit
Branded product with an assigned GTIN	brand + gtin (and mpn if available)
GTIN assigned but you can’t find it	Leave gtin blank; don’t substitute an internal SKU
No identifiers assigned (vintage, one-of-a-kind)	Don’t include gtin, brand or mpn; set identifier_exists to false where needed
You make the product yourself, no official brand	Store name as brand, plus an MPN of your choice
Books	ISBN-13 (starts 978 or 979); ISBN-10 is deprecated

Variants and item_group_id

Each variant is a separate product with its own id and a shared item_group_id, so three sizes in three colors means nine products. According to Google’s item group ID specification, the attribute is required for variants in free listings and for Shopping ads targeting the US, UK, Germany, France, Japan and Brazil. Each variant needs a distinct landing URL (a path or query parameters such as ?color=green&size=small), and the title, color, price, availability and image on that URL must match the variant’s data.

• Use the parent SKU as the item_group_id and don’t submit the parent as a separate product.
• Keep item_group_id stable over time, and don’t rely on letter case to make IDs unique.
• Don’t group products that aren’t variants, such as a suit jacket and trousers sold separately.
• Still submit color, size, material, pattern, age_group and gender where they apply.

Images

Google’s image requirements disapprove images with promotional overlays, watermarks, added logos, placeholder graphics, borders or one photo reused across distinct products. JPEG and WebP are recommended, and the file extension must match the real format. Google has also announced a minimum of 500 x 500 pixels for all products beginning January 31, 2027, and recommends around 1500 x 1500 or above. Generative AI images must keep their IPTC DigitalSourceType metadata.

Price and availability accuracy

Google crawls landing pages and compares them with your data. A price that differs from the page, or that is injected with JavaScript after the page loads, can trigger a price mismatch, and repeated mismatches can lead to account suspension. Availability must match too; Google’s availability mismatch guide points to timing between site and data updates as a common cause.

Automatic item updates can correct sporadic price and availability problems from landing page data, but Google describes them as a safety net for a small share of products, not a replacement for regular updates. For temporary removals, use pause (up to 14 days) or excluded_destination rather than marking in-stock items out of stock, and use sale_price with sale_price_effective_date instead of overwriting price during a sale.

Supplemental data sources and attribute rules

A supplemental data source adds or updates attributes on products in a primary source, matched by exact ID, but can’t add or remove products. Attribute rules (previously called feed rules) transform values with operations such as Set to, Extract, Take latest, Prepend, Append, Find & Replace, Calculate and Split & Choose. Both require the “Advanced data source management” add-on, and rules should be saved as a draft and tested before they are applied.

We use these tools when the store platform can’t output the right value directly, for example adding missing material or gender attributes from a spreadsheet, or standardizing brand names. Where the platform can be fixed, we prefer fixing it there, because one source of truth is easier to maintain.

Feed labels and multiple countries

One data source can target several countries by editing its countries or using the shipping attribute per product, as described in Google’s multi-country guide. Additional languages need a new product source with translated data and landing pages. Feed labels group products for Shopping, Demand Gen or Performance Max campaigns; they can be up to 20 characters (uppercase letters, numbers, hyphens and underscores) and can’t be changed after the data source is created, so they are worth planning before setup. For more detail, read our guide to multiple countries and currencies.

How a feed project runs

1. Export the product data and the issue list from the Needs attention tab.
2. Map where each attribute comes from: store fields, app settings, file columns or rules.
3. Compare a sample of products by hand against their landing pages and structured data.
4. Fix the highest-impact problems at the source first, then layer supplemental data or rules where needed.
5. Re-check item status after the next upload and keep a change log so later syncs can be audited.

If products are already disapproved, our product disapprovals service focuses on diagnosing and clearing those issues first.

Sources

• Google Merchant Center Help — Title [title]
• Google Merchant Center Help — Description [description]
• Google Merchant Center Help — About unique product identifiers
• Google Merchant Center Help — Item group ID [item_group_id]
• Google Merchant Center Help — Image link [image_link]
• Google Merchant Center Help — How to fix: Mismatched product price
• Google Merchant Center Help — How to fix: Mismatched product availability
• Google Merchant Center Help — Automatic item updates
• Google Merchant Center Help — Create a product data source
• Google Merchant Center Help — Set up your attribute rules
• Google Merchant Center Help — Feed labels
• Google Merchant Center Help — Show products in multiple target countries