Google Shopping optimization: Manage your best and worst selling products [new template]

Introducing shopping outliers, a new Supermetrics template from OIKIO Digital Performance Agency.

Advertising in Google’s Shopping network provides retailers with an excellent opportunity to promote online and local inventory in order to boost traffic, visibility, and online sales. As Google’s automation keeps improving, it challenges performance marketers to find new ways to both understand and collaborate with the algorithms making the bidding decisions. 

As any shopping specialist can tell, one of the most powerful ways to improve Google Shopping Ads’ performance is to make improvements to the product data feed required to run the ads. 

However, for many advertisers, this may mean adding product-related data to enhance visibility in search results (eg. product category, color, brand, EAN code, etc). While these are all useful pieces of information, they are simply descriptive data points, not performance-based data points. 

Now that Smart Shopping — a nearly fully automated campaign subtype from Google — is performing annoyingly well compared to “standard Shopping”, the common shopping advertiser may find themselves with only little to work with, and may thus find their Shopping performance to plateau after a while.

After all, how do you manage something that is not designed to be managed?

At OIKIO Digital Performance Agency, we came up with a partial solution to this and want to share this with others as well. We can better manage Smart Shopping (or any Shopping advertising in general) by managing the actual data the system gets to use via product data feed. 

This new approach has helped increase our clients’ Google Shopping ads’ performance by 15-35 % from before and has been validated across our clientele. Seems like it works. 

Introducing the shopping outliers template

The shopping outliers template creates a new supplemental feed for Google Merchant Center that is used to power Google Shopping advertising. This supplemental feed adds a new custom label for all products associated with the feed by combining two different performance data points together.

The feed format is universal and is built around required metrics. This means that it’s applicable to all shopping advertisers. The template pulls data from Google Ads (product cost data) and mixes that with data from Google Analytics (product purchase data). In order to function properly, the template requires that the user has their ecommerce tracking implemented properly via Google Analytics and that the product ID numbers match in Google Ads and Google Analytics.

The idea here is that even if Google’s automated bidding systems aim to increase ecommerce revenue as much as possible, they may still cling on to some products that may garner attention (eg. clicks), but are not products that people end up buying. This template helps reduce this behavior by allowing better management of these poor-performing products.

Based on the data from the past 21 days, the template automatically groups all products found in the product feed into the following categories and assigns them with a custom label:

1. Top seller: more than average number of product purchases
2. Ok: close to the average number of product purchases
3. Low: fewer than average product purchases
4. Cooldown: fewer than average product purchases, but product cost is more than the average product cost
5. Product not found: the product has spent money but doesn’t have product purchases data

If you’re using the template, you can customize these thresholds to your liking to better match your needs.

The template also tells you how many of your shopping feed’s product items would be labeled with ‘cooldown’ and what the historical spend ratio has been for those products. 

The benefits of using this template

Many shopping specialists are already familiar with labeling their best-selling products and managing them in a different campaign. This template adds the poorly performing products into their own custom label, which you can then either exclude from your shopping campaigns or set up a new campaign for them with a smaller budget and thus increase overall shopping performance. 

With this template, you can now show preferential treatment to both the best performing products and the least performing products (thus the template name “outliers”).

The exclusion is not permanent, however, as that could hinder long-term performance. Instead, the “poorly performing products” are categorized based on whether the product’s cost is above the average product cost and there have been little to no purchases for that product. As the template gets the data from the past 21 days, eventually the poorly performing products’ cost will be lower than the average cost, which will replace the custom label, effectively giving these products a new opportunity. We call this process the cooldown period.

This template is another example of complimentary automation solutions currently crafted at OIKIO, which means that it’s designed to complement Google’s own AI solutions rather than contradict them.

Playing the devil’s advocate here for a second: What’s the most likely downside of using this template?

Since the template only considers direct relationships with product cost and purchases, it does leave out indirect relationships completely (e.g. user clicked on a product and ended up buying a completely different product or bought it offline). At OIKIO, we haven’t seen this to be an issue with our clients, but we figured it should be mentioned.

Also, the template pulls the product purchase data from all traffic sources. This means the data also includes purchases that were not attributed to google / cpc. This was a conscious decision to try to implement cross-channel purchase behavior, but also to weed out the true “non-performers” for better performance management. 

How to get started with the template?

Follow these steps to get started with the shopping outliers template:

1. Check that you meet the template requirements
   1. Google Ads: You are currently doing shopping advertising and you have access to both Google Ads (for data fetching) and Google Merchant Center (for uploading the supplemental feed).
   2. Google Analytics: ecommerce tracking enabled and the product ID numbers match with your shopping feed values.
2. Having access to the template
   1. Copy the template and go to Add-ons → Supermetrics → Duplicate this file for another account.
3. Customizing the criteria
   1. Instead of using the predefined criteria on performance groups click the tick boxes in ‘Setup and overview’ and see the impact your criteria had.
4. Selecting available custom label
   1. In ‘Merchant Center Supplemental Feed’ sheet choose an available custom label.
5. Before adding the supplemental feed to Merchant Center
   1. Keep the ‘Merchant Center Supplemental Feed’ as the first sheet in this file (left side of the ‘Info’ sheet).
6. Adding the supplemental feed to Merchant Center
   1. Add the Google Sheet as supplemental feed and create a feed rule that takes the values in the supplemental feed to the primary feed.
7. Using the performance groups
   1. When all the steps above have been done, you have now the option to exclude cooldown labeled products and boost the high performance products in Google Ads.

How to customize the template?

Based on your business and your needs, you can customize the template in various ways.

By default, the template uses the predefined threshold of what is considered as a top seller purchase performance for an individual product.

If the purchase threshold needs to be determined by other criteria, you can click the checkbox and determine the performance thresholds. 

In terms of product cost level, you can either use the default average product cost or a custom criterion.

If the product cost threshold needs to be set higher or lower than the default provided, you can click on the checkbox and set your own.

If the item IDs in the Merchant Center consist of any letters, it’s crucial that the IDs in the template have the same format with Merchant Center i.e. upper and lower case considered. If the letters are in lowercase, no action is required, but in case that they are uppercase, by clicking the box (see below) also the ‘Merchant Center Supplemental Feed’ will treat the IDs as uppercase and you are good to go. 

Next, you’ll want to take a look at the overview of the number and share of the  cooldown labeled items, the cost of the cooldown labeled items, as well as the total cost of the non-cooldown labeled items. If you’re not happy with the shares, go ahead and tune your criteria. In order to have an impact, the share of cooldown labeled items (cost) should be more than 15% rather than less.

After finalizing the product and cost threshold set up, you can find the final product feed from the ‘Merchant Center Supplemental Feed’ tab. Now, this feed can be implemented in the Merchant Center as a Supplemental Feed. 

And remember, please keep the sheet ‘Merchant Center Supplemental Feed’ as the first sheet in this file (on the left of this ‘Info’ sheet). Merchant Center only considers the first sheet in files with multiple sheets.

Over to you

That’s it! You’re now ready to start using the shopping outliers template to drive better ROAS from your Google Shopping ads.