Google Shopping 101: Building Your First Data Feed
In my last post, Google Shopping 101: Why Data Feed Quality Matters, I discussed the importance of taking the time to build a high-quality and complete data feed. In this post, I’m going to share the process I go through with my clients when they build their first data feeds.
Step 1: Identify your business’ product category
Google buckets products into general categories including:
- Variant products
- Apparel products (non-variant)
- Custom goods
- Media, video game software
- All other products
Variant products, categorized as ‘Apparel & Accessories’ in Google’s product taxonomy, are any group of identical products that only differ by the data feed attributes ‘color’, ‘material’, ‘pattern’ or ‘size’.
For example, a purse that comes in three different colors would be considered a variant product.
Non-variant products, also categorized as ‘Apparel & Accessories’ in Google’s product taxonomy, are only available in one size, color, material and pattern.
For more information on apparel products, check out this article on submitting apparel products to the Google Merchant Center.
Step 2: Determine and complete required attributes
After identifying what types of products your business sells, use Google’s summary of attribute requirements table, to determine what fields you are required to have in your data feed.
You may sell a wide range of products that fall under several of the categories listed in step tone If this is the case, you will need to make sure you have the attributes required for each category.
For example, if your company sells gardening gloves (apparel products) and hedge trimmers (all other products), you will need to make sure that the gardening glove line in the data feed includes attributes such as size and color.
The hedge trimmer line in the data feed will not need to include ‘size’ and ‘color’ attributes as those are not required fields for products categorized as “all other products”.
After this step, you should have your first, basic iteration of the product data feed. This should not be the last version of your data feed as there are a few additional steps that ensure the creation of a high-quality data feed.
Step 3: Identify business priorities and goals
While at first this might not seem tied to the development of a product data feed, identifying your business priorities and goals is crucial to its development and the overall success of your Google Shopping campaign.
Identify what products are more important to your business. What are you willing to pay to promote these high-value items? Also determine what products are less important and what you would be willing to pay for them.
Ideally, you should have some type of tiered grouping for all of your products based on their value or importance to the bottom line.
Step 4: Get granular and relevant
During the first round of data feed creation, you may have chosen to keep things fairly simple and high-level. Now, you need to complete one of the most important steps: making your data feed as detailed and granular as possible.
Two of the most important required attributes for the data feed are the ‘Google product category’ and the ‘product type’ fields. When completing the Google product category field, you are required to use Google’s product taxonomy. Make sure to get as granular in your categorization of each product as possible. The more information, the better (see Matt Wynne’s post, What is Google doing with Shopping Data?, where he outlines the effects of having more data in Google Shopping).
Use the ‘product type’ attribute, to create your own taxonomy. Our clients have found it most helpful to use their website as a guide.
Using your website as a guide is beneficial because your website is likely more granular than Google’s product taxonomy. Being granular in your categorization of products is crucial for bidding in Google Shopping. Google offers benchmarck CPC, benchmarck CTR and impression share at the most granular level you segment to. Knowing these competitive metrics can help you better manage bids for each specific product set, which will result in great performance.
In the previous step, I recommended creating tiers for the level of importance or value for each one of your products. With this information, create a ‘custom label’ attribute to label each product by your custom tiered system. Using the ‘custom label’ attribute to Identify products by their importance to your business will help performance of your Google Shopping campaigns because you can, and should, bid differently based on each product’s importance to the bottom line.
Step 5: Upload and troubleshoot your data feed
After you have gotten as detailed as possible in the next few iterations of your product data feed, upload the data feed to your Google Merchant Center account. A few hours after uploading, click “data quality” on the left-hand side to see what issues your data feed has. These issues can prevent certain products from showing:
Or, they can be mere optimization recommendations:
Export the errors into Excel and start fixing. If possible, fix these issues on your website to improve your data there as well.
If you choose to go live with your Google Shopping campaign after you upload your data feed for the first time, be sure to allow 3-5 business days for Google to review your data feed and website in detail.
Step 6: Keep improving
As seasonality sets in or products get added, be sure to go back and look at ways to improve your product data feed.
Your Google product data feed isn’t just used for Google Shopping – it can also be used for Facebook and web dynamic remarketing ads, so make sure your data feed is always relevant, detailed and of the highest quality.