When managing a large catalog of products an account with sufficient data is necessary to leverage Google’s smart biddingoptions. However, when you don’t have this luxury you have to utilize other automatons such as scripts, rules and bulk imports. In this case study I will explain how I improved this digital marketing campaign using automation.
Google Shopping Revenue Increased 980% compared to the previous period
Return on Ad Spend increased 98%
Lack of consistent new revenue from Google Ads
No data feed or shopping campaigns
Segmented each product into separate ad and product groups using Bulk Uploads
Created campaigns based on device, product popularity and seasonality
Developed scripts to programmatically do bid adjustments based on product performance
Used rules to send email alerts
Analyzed Google Merchant Center to improve feed quality
Switched to Target ROAS when enough conversion data available
To have a successful campaign I knew that I had to get more granular with my item segmentation. To do this efficiently I had to leverage bulk uploads. In a standard Google Shopping Campaigns you get all products put in one ad group with no device parting.
I used the above Excel format to separate every single ad group to have a single product in it. This will allowed me to know exactly what search queries are triggering which ads. Also, I gathered more data better performance product could receive higher bids or Target ROAS,
Google Ads Editor
To do this efficiently I used the Google Ads Editor to bulk create all ad groups with the itemCode – Item Name naming convention . This allowed me to save time and made the bulk upload easier.
To properly allocate budget I needed to segment my campaigns based on device, seasonality and popularity. My naming convention was Search Query Type – Device Type – Product Segmentation. Search query type was based done if it was generic, branded or specific item. I would bid less on generic campaigns, and more on item and branded terms. Device type would be between desktop, mobile and tablets.
From historical data desktop outperformed mobile and tablets so I would be willing to pay more per click . When it came to product segmentation I had campaigns that kept all products(4000+), best sellers(500+) and holiday specific products(Halloween, Christmas, etc). I would separate these with custom labels in Google Merchant Center.
To improve efficiency I created a script to increase bids on the product group and ad group level when they get conversions. I had to create a custom solution for this problem using documentation in Google Ads Scripts. First, it creates a variable productGroups and sets this equal to all product groups in the account via the AdsApp object. Next, it grabs all of the product groups that have had greater than 0 conversions yesterday and puts them into an array. Then a while loop iterates the array and each MaxCpc is increased. Finally, the script goes through the same process but for each shopping ad group.
For rules I set up alerts for search impression share, shopping sales, and low click though rates(CTR). When search impression share drops below 20% I get an email. This helps me to make the decision whether I should increase bids on that specific product(s) or if something quality score is wrong. The Shopping Sales rule keeps me updated on the previous days performance. I also check the quantity of the product purchase to determine a more aggressive bid increase as well. Finally, Low CTR rule tells me if there may be something copy or image wrong with product(s). If an ad is showing up for irrelevant terms I can go in and set that as a negative keyword. Or if an image happens to have a watermark I can update that as well.
Google Merchant Center Feed
In Google Merchant Center I check item diagnostics, and improve feed quality based on data. I did this by creating a supplementary feed to make alternative titles and descriptions based on the search terms report. For example if I saw my product is being search by gender, color and then brand. I would reorder that product in a supplementary feed with rules or manually. This can help to improve click through rate (CTR) and ultimately. quality score.
Once my campaigns started get adequate conversions of 20 conversions in the last 45 days I switch to Target ROAS. I set my target initially to what was individually recommended for each campaign. Then, I let the algorithm learn and improve over time. I increased the Target ROAS slightly over time to improve profits as well.
When setting up Google Shopping segmentation and data analysis is your friend. You should not start with smart shopping off the bat because the data will be a black box. Start with a well segmented campaign and set up rules and scripts to follow decisions you would make manually. Once you have the proper data switch to an automated bidding strategy to leverage what the Google algorithms do best. This is so you can focus on high level strategy and what marketers do best, communicating value.