Google Merchant Center Feed Rules: Automate Product Data Transformations

Published: 2026-06-09 | Category: Feed Management

Feed rules in Google Merchant Center let you transform product data after it enters your feed, without touching your source system. You can fix missing attributes, map values, and override fields at scale using conditional logic. This saves hours of manual feed editing and reduces disapprovals from attribute mismatches.

What Feed Rules Can Do

Feed rules apply transformations to any attribute in your feed. You can set static values, extract data from other fields, use regular expressions to reformat strings, and apply rules conditionally based on other attribute values. Rules run every time your feed is fetched or refreshed, so fixes stay in place automatically.

Common uses include: adding a missing condition attribute (set to "new" for all products), extracting a brand name from a long title, reformatting sizes to match Google's accepted values, and mapping internal category names to Google product category IDs.

Setting Up a Basic Feed Rule

Go to Products > Feeds > select your feed > Feed rules tab. Click the attribute you want to modify. You have three transformation options: set a static value, use another attribute's value (with optional regex processing), or combine multiple sources. For conditional rules, add an "IF" condition first, then specify the transformation to apply when that condition is true.

Example: if your feed uses "NEW" for the condition attribute but Google expects lowercase "new", create a rule for the condition attribute, add an IF condition where condition equals "NEW", then set the value to "new". All products matching that condition will have the field corrected.

Extracting Data with Regular Expressions

The regex extract option is powerful for pulling structured data out of free-text fields. If your product titles include a brand in a predictable pattern, you can extract it into the brand attribute. For example, a rule on the brand attribute can extract text matching a pattern like "^([A-Za-z]+)" from the title field, capturing the first word as the brand name.

Test your regex before saving. Google provides a preview mode that shows how the transformation will affect your actual products. Check at least 20 products before activating a new rule to catch edge cases.

Priority and Rule Order

When multiple rules apply to the same attribute, they run in order from top to bottom. The last rule to match wins. Drag rules to reorder them. Put general fallback rules at the top and specific overrides at the bottom so that the most precise rule has final say.

Conditional Rules for Category-Specific Fixes

Different product categories often need different transformations. Use the Google product category condition to scope rules to specific verticals. Electronics may need different size format handling than apparel. Creating category-scoped rules prevents a fix for one category from breaking another.

Monitoring Rule Impact

After saving rules, check the Diagnostics tab within 24 hours. Compare disapproval counts for the affected attributes before and after. If an attribute that was generating 200 disapprovals per day drops to single digits, the rule is working. If disapprovals stay flat or increase, the rule may be introducing a new issue.

Common Feed Rule Mistakes

Overusing static value rules is the most common mistake. Setting the condition attribute to "new" for every product causes problems for merchants who sell used or refurbished items. Always add an IF condition that scopes the rule to products where the attribute is actually missing or wrong.

Avoid chaining too many rules on a single attribute. More than three or four rules on one attribute becomes hard to debug. If a product still has an incorrect value, stepping through the rule chain mentally to find where it broke is time-consuming. Simpler rules are easier to maintain.

Feed Rules vs. Supplemental Feeds

Feed rules work on the primary feed and run server-side after upload. Supplemental feeds are separate files that override specific attributes for specific products identified by ID. Use feed rules for blanket transformations that apply to many products based on logic. Use supplemental feeds when you have a specific list of product IDs that need specific values that cannot be derived from existing data.

Both can be active simultaneously. Supplemental feeds take precedence over primary feed values, and feed rules run after supplemental data is merged in. Plan your data flow to avoid conflicts between the two.

Next Steps

Audit your current disapprovals in the Diagnostics tab and identify attributes with the highest error counts. Start with the attribute generating the most disapprovals and build one feed rule to address the most common error pattern. Monitor for 48 hours before adding more rules. A methodical approach prevents you from creating new issues while fixing existing ones.