New UPC/Gtin broke Google Shoppping
Hi,
We had a campaign running for a client for few months without any major issue.
It was a smart campaign with target ROAS.
The ROAS was the targeted one.The client’s product hand’nt a UPC/Gtin thus far. He decided to integrate one this week as he needed to expend his sells with some marketplaces.
In the hours after he integrated the UPC/Gtin, the feed was changed for
and infos were provided.
identifier_unique
=yes
24h later, impressions decreased by 50%, 48h later by 85%UPC/Gtin are brand new, so when you lookup at them on some database, you will find them as valid, but unknown.
They have been removed since, and the feed has also been changed accordingly.
We’re trying to relaunch it with a smart campaign. Doesn’t work.
We’re trying with a Manual campaign, no more luck thus far (CPC have been set up higher to boost the impressions)We estimate that we are recieving 15% of the impressions we should receive.
Why did all this happen? Is there any way to revert this?
We are eventually thinking about duplicating the products with new url, Id etc… to present them as new products.Any feedback is welcomed.
Thanks.
The short answer is:
GTIN
to Google Shopping? Your massive drop in Google Shopping impressions is almost certainly due to Google’s algorithm treating the new, unvalidated UPC/GTINs as invalid or low-quality unique product identifiers.
When your products went from having no required identifier (where Google allows them to run, albeit with limited performance) to having new, valid-format but officially “unknown” GTINs, the system’s confidence in the product data likely cratered.
The system will favor products with established, known GTINs, especially if you set the
attribute to “yes.”identifier_exists
You need to immediately check the Merchant Center Diagnostics tab for disapproval or warning messages like “Invalid GTIN
” or “Incorrect product identifier.”
The way to fix this is to either permanently remove the GTINs and ensure the
attribute is set to “no” or “false,” or obtain official, fully validated GTINs and submit them, then wait 72 hours for Google to re-crawl and re-rank the products.identifier_exists
Duplicating the products is a desperate measure that might work but carries risks of policy violation and won’t solve the core data quality issue.
The long answer is:
Your experience is a classic, but often misunderstood, reaction from the Google Merchant Center (GMC) algorithm.
When your products had no GTINs, Google allowed them to run with the
field implicitly or explicitly set to “no” or “false,” which is acceptable for custom, vintage, or private-label goods.identifier_exists
However, once you introduced a new, valid-format GTIN
and changed the identifier_exists
attribute to “yes,” you essentially told Google: “This product now has a unique, globally recognized identifier.”
Because these new GTINs were unknown in Google’s or GS1’s comparison databases – as they were so new – the system flagged them as low-quality or potentially invalid.
For Google’s algorithm, an unknown GTIN
is often worse than no GTIN
at all, especially for a Smart Shopping campaign that relies heavily on product data quality for ranking.
This triggers a massive quality score and ranking drop in the auction, which is why your impressions plummeted.
Even increasing CPC on the Manual campaign won’t help much because the product listings themselves are being suppressed due to the data quality penalty.
Removing the GTINs and setting the feed back to
was the correct immediate action, but the impact is not instant.identifier_exists
="no"
Google’s systems can take 24-72 hours, or sometimes longer, to fully re-crawl the feed, process the changes, clear the data quality flag, and then re-enter the products into the ad auction at their former performance level.
You must be patient for this recovery period.
To prevent this kind of data quality issue from impacting your bidding and optimization in the future, especially with automated campaigns like Smart Shopping or Performance Max, you should use server-side tracking to send conversion data directly to Google Ads.
While this won’t fix the GTIN
issue itself, it ensures that your core campaign optimization engine is fed the highest quality conversion data, which improves the long-term performance stability of your Target ROAS strategy, even if the GTIN
issue causes a short-term ranking penalty.
An excellent and cost-effective solution for this is to implement the Google Ads API conversion tracking via a server-side setup.
You can set up a Google Tag Manager (GTM) Server Container and host it on a low-cost, scalable server like Stape or Google Cloud Platform (GCP).
This allows you to route conversion data (like a purchase event) directly from your server to Google Ads.
This server-side solution bypasses unreliable client-side issues like browser tracking prevention and ad-blockers, providing Google’s algorithms with clean, accurate conversion signals.
This superior data quality is critical for the Target ROAS strategy to function efficiently and help your campaigns recover and stabilize faster from feed quality dips.
If you must use new GTINs in the future, ensure they are fully registered and allow extra time for Google’s systems to recognize and validate them before making the feed change permanent.