Your product data feed is a file (generally an XML or TXT file) that is sent from your website to the Google Merchant Center, containing all of your product information (it can also be sent to the Merchant Center via API connection).
Think of your data feed as a massive Excel file, where each row contains the information for a single product, and the columns that stretch across contain all the attributes related to that product.
Let’s say you sell office supplies. Listed in your feed are two calculators: the Texas Instruments BA II Plus Financial Calculator and the HP Prime Graphing Calculator. The former is primarily used by college-level business students, and the latter is used by high school and college-aged students studying geometry, trigonometry, and calculus.
There’s a lot of information related to these products that you need to include in your feed if you’d like for Google to serve your ad to the right prospective customers. Without sufficient information, Google might only serve your ad for the search calculator, which would likely be a disaster, as these users are likely not actively looking to purchase a $140 graphing calculator.
There are dozens of potential attributes that you can add to describe your product, some of which are required. Product title, price, brand, description, and a link to a landing page are required attributes. Size is not required, as this is an attribute that might help describe an item of clothing but does not help describe a Texas Instruments calculator.
See here for the list of all attributes, including the list of required attributes.
To start, every item in your feed must have an Item ID. Traditional brick and mortar retailers will know this as a SKU number. It’s a number that is somewhat arbitrary and is used exclusively by your own business to track inventory and other operational tasks within the backend of your website. This is different from UPC or GTIN, which we will cover shortly.
The Item ID for your Texas Instruments Financial Calculator might be 10001, and the Item ID for your HP Graphic Calculator might be 10002. Google Merchant Center will use these numbers to refer to your products (as opposed to title or another attribute).
GTIN or Global Trade Identification Number is a separate number that needs to be added to any products that are registered with the GS1 database.132 Most people recognize GTIN numbers as the 12-digit UPC barcodes that appear on product packaging, or the ISBN numbers that are assigned to books. Other GTIN examples include EAN numbers (the UPC of Europe), and JAN numbers (the UPC of Japan).
The GTIN of a product is consistent throughout the world, regardless of which merchant is selling the product. For example, the GTIN (UPC) for our Texas Instruments Financial Calculator is 033317206353, and the GTIN (UPC) for our HP Prime Graphing Calculator is 887758319833. A search for either of these numbers will bring you results for these specific products I am referring to.
The key difference between Item ID and GTIN is that ID can vary from merchant to merchant (you can even change your Item ID if need be), whereas GTIN remains consistent. You might list the Texas Instruments calculator as Item ID: 10001, but your competitor might list it as Item ID: 355 (hence why it is an arbitrary number). GTIN, however, is the globally recognized identification number for this product. The GTIN: 033317206353 for this Texas Instruments calculator would exist in both yours and your competitors’ feeds.
The advantage of including GTIN in your product feed is that Google will know exactly what product you are advertising. Google will likely already have data on what types of searches result in sales for that specific product. You are a step ahead of the game.
We often say that there are two parties that you must appeal to if you want to serve the perfect ad, be it on Google Shopping or elsewhere. You must first effectively communicate with both the machine (Google) and your customers. Therefore, even after you add the GTIN, other information is still vital in helping Google better understand your products, as well as describing your product to potential customers.
Few products are recognized by the general public by their GTIN number, so customer-facing attributes, including Title, Description, Image, and Price have equal importance.
While you might think that Price is something that can’t be optimized, consider a retailer who sells wine glasses. Does the retailer list their product as $3.99 per glass, or $7.50 for a set of two, or $18.99 for a set of four? Perhaps the retailer offers a discount on bulk orders, where an order of 100 or more glasses brings the price per unit down to just $1.89. This low price might be appealing to click on, but it is misleading and frustrating if most of your customers are only looking for a set of two. This is a conundrum that plagues any retailer who sells customizable products, or products that ship in various quantities. To be honest, there is no rule of thumb as to what works best. The best advice I can give is to start out with mirroring what your competition does. If the consensus is that your industry advertises the lowest possible price per unit, then that is likely what your customers are used to. However, if most customers are looking for a specific quantity or have some other standard expectation, then you should begin there. This is always something you can test over time.
Title is the most important customer-facing attribute that you list in your feed. The information listed in your product titles has a great impact on the searches that your ads will match to, and the order of the information provided can have a massive influence on your Click-Through-Rate and overall performance.
Let’s play “One of these things is not like the other” with the following Shopping results that appear for a search for Sony LED TV:
Is there a single person in the universe that would choose to click on the Crutchfield ad over a Best Buy ad? Set aside the star-ratings vs. free shipping aspect of the ads and just use the information derived from the product title.
The structure of Best Buy’s titles is specifically designed with the user in mind. They likely conducted customer research to determine the information that is most important for potential TV customers and structured their titles according to priority.
I had to purchase a new TV recently, and I would agree that brand, size, type are the top three bits of information that I look for—along with price, of course.
The Crutchfield title begins with brand and then immediately goes to the model number, XBR75X800G, a confusing string of numbers and letters that mean nothing to most users.
Keeping in mind the Catch & Bid days, you could argue that Crutchfield might have an advantage when a user’s search contains the product model number. In general, if a user searches specifically for a model number, their intent-to-buy is likely greater than if they broadly searched for sony led tv.
However, even if that holds true, far more people are searching for brand or size than they are searching for specific model numbers. *
Title is just one aspect of a Google Shopping ad, but Best Buy is crushing it in this department. It’s also worth mentioning that Best Buy dominates the Google Shopping search results for any TV-related searches. To pull a screenshot for this example, I had to continually refresh the page to be served even one ad that wasn’t from Best Buy. That is unlikely to be a coincidence. It’s more likely indicative that their Google Shopping ads are very successful.
What’s more, the orange Seller-Ratings (stars) featured in the Best Buy ads are likely to have a greater impact on CTR, compared to the Free Shipping copy that is featured in the Crutchfield ad— this is based on our experience with Seller-Ratings, compared to clients who are not able to take advantage of this feature.
Seller-Ratings are not added to your Shopping ads via data feeds or your Google Ads account. Rather, they are automatically assigned to ads if Google is able to collect reviews from one of the following sources:
● Google Customer Reviews: A free program that collects post-purchase reviews on behalf of merchants
● Aggregated performance metrics from Google-led shopping research
● Ratings from Google Consumer Surveys: A market research platform we use to collect data for certain domains and businesses
● Shopping reviews for your store domain, which include reviews from various third-party sources 133
If Google doesn’t have information for your store, or if your store doesn’t meet the minimum seller rating thresholds, a seller ratings page may not load for your homepage. You can see if your store is eligible for, and serving, Seller-Ratings by navigating to the following URL, but replacing “www.example.com” with your homepage URL:
www.google.com/shopping/ratings/account/lookup?q=www.example.com
Many of the popular ecommerce content management systems (CMS) such as Shopify and BigCommerce offer a free Google Shopping app that allows you to set up a feed through API. If you’re unfamiliar with APIs, think of it as an open connection between two platforms. Without API, you are relying on an export of your feed data, which is sent to the Merchant Center on a recurring basis and cannot be changed once it is sent. Content API feeds create a two-lane highway where information can be exchanged instantly.
Shopify users leveraging the Google Shopping API App benefit from real-time updates to product data. If a product goes out of stock, or the price is updated on the site, it is automatically changed in the product feed. Advertisers not leveraging API connection would need to manually re-fetch the data into the Merchant Center or wait for the next automatic upload—typically the next day.
Content API feeds are a nice luxury but are not necessary for every advertiser. If you have fewer than 1,000 products and don’t see many changes to product inventory or pricing on a daily basis, a traditional XML or TXT feed is still a fine option.
Custom Labels are optional feed attributes that advertisers can use to add additional information into their product feeds. Advertisers can add up to five custom label fields, denoted by the syntax of custom_label_0, custom_label_1, etc., which can be used for segmentation and reporting purposes in your Google Shopping campaigns. They are organizational attributes.
In our calculator example, we might include a custom label for calculator type, where we assign a value of Financial to the Texas Instruments Financial Calculator and Graphing to our HP Graphing calculator. Let’s say we have more favorable profit margins on our graphing calculators compared to our financial calculators. In this case, we might use this custom label to separate our Shopping efforts into two campaigns, where our graphing campaign has more budget and a lower (more aggressive) Target ROAS goal.
Popular uses of custom labels include separating out products by product type, target audience, profit margin, overall price-point, best-sellers, loss-leaders, and products that you wish to exclude altogether. The information in your custom label fields can also be used to dynamically populate other, customer-facing aspects of your shopping feed, using a feed management tool like ProductsUP or a feed management partner agency like Sales & Orders.
Feed management software provides a solution to this issue without having to change anything on the client’s website. Software like ProductsUP allows you to intercept your product data before it gets sent to the Merchant Center, allowing you to create rules that will override individual components of your feed with the goal of improving your feed or creating a better experience for your customers.
For example, we have a client where we do not want to use the actual product title that is listed on the website as our Title in the shopping feed (their titles happen to feature the product model number, which, as we covered in our TV example above, is not an ideal title to submit to the Merchant Center). We use a set of rules that pull information from various custom label attributes to override the original Title.
Let’s say we want to rebuild our calculator titles from scratch, using other attributes from within the feed. After analyzing previous customer data and pairing that with search term trends, we find that the best order of information would include:
1. Brand
2. Financial or Graphing (represented by information we add in custom_label_0)
3. Color (black or white)
4. Product Type (calculator)
5. Model Number (represented by information we add in custom_label_1).
We would first need to add additional information into our custom label fields, so we’d work with our developer to set up schema on the website or some sort of automatic rule that includes this information in the original feed file for both custom_label_0 and for custom_label_1.
Then we would create a rule in ProductsUP that would follow the following logic:*
Title = [brand] “ “ [custom_label_0] “ “ [product_type] “—” [custom_label_1]
Once this rule is created, the final Titles that will be sent to the Merchant Center will look like the following:
● Texas Instruments Financial Calculator—BALIPlus
● HP Graphing Calculator—G8X92AA
