News & Insights


Cleaning up your Google Shopping Data Feed

pwd staff OLIVER WOOD
Oliver Wood

|27th June 2017

Before creating shopping ads with Google AdWords, you first need to upload your product catalog into Google Merchant Center because this data is what feeds into your shopping ads and the quality of your data will impact your conversion rates.

Since the product catalog is the foundation of shopping ads, it often needs some cleanup before PPC experts can create ads, whether they are standard shopping ads, dynamic remarketing ads or TrueView ads. The role of the PPC expert is to translate and enrich data from the client’s catalog before presenting this information to the consumer.

Ensuring you have a clean feed before advertising will increase the likelihood of a good click through rate and engagement from shoppers, as well as making your products more likely to show up for relevant queries, and increase impressions to reach more shoppers.

Let’s start with an example of what an ad might look like when it is pulled directly from the product catalogue and the data feed is not optimised for Google Shopping Ads. The below ad for a performance road bike is from an existing product description. At a glance, it looks okay; the image is clear, there are no misspellings, and the description states what it is.

Below are the attributes that were pulled from the product catalog to create the feed and that information was used by Google for the ad. Certain attributes are required in the feed and are a must for good product data because it separates one product from another.

But cleanup is also about the consumer’s experience. Optimising this example for the Performance Road Bike starts with filling out all of the attributes, such as gender and size, so you reach the right audience. You do not want to show ads for women’s bikes to men who are searching for themselves. It may result in a high click-through-rate but unnecessary costs when the wrong type of consumer is clicking on your ad.

There are additional fields that can be cleaned up beyond just including gender and size. Instead of Performance Road Bikes, a more engaging description is Fuji Men’s Road Bike 46CM Black. Not only does the searcher know immediately that it is a men’s bike, it also includes the brand and color which increases the chances of the right prospective customer clicking on the ad.

Although men is a better descriptor and sounds more natural in this ad than male, Google’s requirement for the actual gender field is male. From Google’s perspective, male and men are two different values so cleaning up your Shopping data feed also include modifying it as needed to match Google’s standards.

Other differences between Google’s standards and a product catalogue may also include using the term refurbished instead of reconditioned and spelling Out of Stock instead of an abbreviation such as OOS. For many clients, their internal database will not exactly match Google standards so a quality check is needed for attributes around gender, age, color, condition, and availability.

PPC experts should also check for unique identifiers, such as a brand or item number because this attribute is matched against Google’s product catalog as a check to ensure the product really represents what it claims to represent and is also an opportunity to highlight what makes the product unique.

Think like the consumer

This advance work is not only for Google, it is also for the consumer. We all have short attention spans when it comes to the internet and want answers to our queries very quickly. Shoppers expects to see details for their product search, and are not satisfied with vague titles or descriptions. Cleaning up your feed before creating ads helps both PPC experts and consumers by matching the right product with relevant searches. This clean-up work also entails thinking from the perspective of the consumers. How would you describe the below image?

This description might include rain boots, spotted yellow rain boots, polka dot boots, or women’s rain boots, all of which could pull up the above image in search. What is important here is how the consumer searches the product, not the name assigned by the retailer.

In this example, you can compare different variations of the product in Google Trends for spotted boots versus polka dot boots to see how your target audience searches for this item and the volume for each keyword phrase. The phrase that is trending is the best descriptor for the ad. In other words, if polka dot boots is the most popular search terms, then that is the term to use in the feed.

Optimise the Titles

The client’s product catalogue may also have very general titles for each item which works fine internally but does not provide compelling information for the consumer. Review the values for the titles and incorporate the brand, product type, and attribute in the title whenever possible.

It is especially useful to have the brand name at the beginning of the title if it is a well known or luxury brand that is important to consumers. If you are uncertain which of your brands are important to searchers based on latest trends, use the AdWords keywords tool to determine the search volume and incorporate popular terms into the title.

Another helpful tool for Optimising the titles is the search query report in AdWords for ads that have been running a while on the Search Network. Some of those queries may include words you want to incorporate in the titles as well as descriptions.

The highest value keywords, such as the brand, need to be in the beginning of the title in case the title is truncated in the search results. And with some products, such as clothing and shoes, the data feed should also include sizes in the title of each product to drive more relevant clicks. Hopefully the product catalog does not have spammy titles that use all caps or stuff in keywords, but if so, ensure those are cleaned up as well.


Only use images that have a white background since the Google search results page is white. If images have a photoshopped logo or promotional text added to it, the ad will be rejected so ensure all images are just the product and not modified to squeeze in additional ad copy.

Since Google’s image search also includes shopping ads, it is crucial that product attributes include accurate color and category information. Google started displaying shopping ads on image search because of user behaviour so it is another opportunity for retailers to meet consumers where they are when they are engaged in searching for products.


  • Never assume a product catalog can be uploaded to a data feed “as is”.
  • Look for gaps in the raw data, such as gender.
  • Ensure all relevant attributes are filled out completely.
  • Use specific, rather than vague descriptions, such as with the bike example mentioned earlier.
  • Think like a consumer and use Google Trends to see popular terms.
  • Match Google’s standards for attributes.
  • Investing time in clean-up work now will result in a better ROI when the ads are launched.