Supplemental feeds: the free way
to fix your Shopping feed
Paid feed management tools like Feedonomics and DataFeedWatch aren't the only way to amend your product data. A supplemental feed overwrites what you already submit to Merchant Center - without touching the raw data itself, and without a monthly fee.
Here's what a supplemental feed is, how to build one, and the best practices to keep it working for you rather than against you.
A supplemental feed lets you overwrite your Shopping feed data. Without touching the raw feed, and without paying for a tool to do it.
What is a supplemental feed?
A supplemental feed is an additional data source your GMC specialist can add into Merchant Center that overwrites the raw product data you already submit. It's most commonly built as a Google Sheet.
The key requirement is a consistent key to verify the changes against. Your product IDs need to sit in the first column - this is what Google uses to match your supplemental data against your existing raw feed.
Every additional column after that - column B, column C, and so on - represents a product attribute. Whatever value you put in those columns overwrites the equivalent attribute in your raw feed once the supplemental feed is pushed through Merchant Center.
The reason to work this way rather than editing the raw feed directly is simple. It keeps the raw data from being overwritten at source, it's a far lighter lift than a developer task, and it's completely free compared to a paid feed management tool like Feedonomics or DataFeedWatch.
How columns behave - the part that catches people out
You can optimise any attribute in the feed this way. But the behaviour of each column depends entirely on whether it's included and whether every row in it is filled in.
- Add a column with values in every row - the raw data for that attribute is overwritten for every product.
- Add a column but leave some rows blank - those blank rows will overwrite the raw data with nothing, which is rarely what you want.
- Leave a column out entirely - the raw data for that attribute is left untouched.
Only add a column if you're happy overwriting the raw data for every row in it. If you're not ready to populate a full column, remove it from the sheet rather than leaving it blank - a blank cell still overwrites the raw value with nothing.
You can also use a supplemental feed to add attributes that don't exist in your raw feed at all. One worth considering as AI shopping surfaces grow is a conversational, FAQ-style attribute - again, kept separate from the raw feed rather than bolted onto it. For the full checklist of feed attributes and everything else Shopping-related, see the Google Shopping Ads & Feed Optimisation Hub.
Tips and best practices
Always the first column, always consistent
Your product ID is what Google grounds every change against. It needs to sit in the first column and match the ID format in your raw feed exactly, or the overwrite won't apply.
You can undo it as easily as you applied it
If something's not working, remove the supplemental feed and refresh - the raw data reverts on its own. From there you can amend the sheet and re-upload the supplemental feed with the changes you actually wanted.
Further reading
The 5 Shopping feed mistakes I find in 9 out of 10 accounts
The most common feed errors holding back Shopping performance - and what to fix first.
Custom labels: how to actually use them
How to build, maintain and align custom labels to the business so Shopping and PMax campaigns scale instead of plateau.
Performance Max: built to run itself - here's how to take control
The levers that actually move the needle once PMax is live, and what to strip out before it flatters in platform numbers.