Google Shopping Segmentation For ROAS Success

Blog Categories:  PPC  

Google Shopping can be scary if you’ve spent most of your AdWords life in Search and Display campaigns. It’s easy to start a Shopping campaign with a huge product catalog at your disposal, not see results in the first month, and get scared off. If that’s the case, humor me and consider a few questions before you conclude Google Shopping just isn’t right for your online store.

Is ROAS (return on ad spend) the main issue? In other words, do you have substantial spend and conversion volume, but CPCs (cost-per-clicks) are just too expensive to rationalize continuing with the campaign?

I wanted to share an example of some basic segmentation that really paid off for a retail client. The Shopping campaign initially was paused due to poor ROAS. See initial stats over the first 10-month run:

Sales: 203

ROAS: 1.79 (goal is 3.00)

The Pitch: For Shopping, the volume was never an issue. If we could harness areas of the campaign that were working and exclude the others, there was really an opportunity here. We believe we’ve found the proper segmentation to exploit this opportunity. Additionally, our team is more experienced in Shopping segmentation at this point, so our management process for these campaigns has more concrete structure.

Strategy: Use granular segmentation to strategically increase revenue in the most effective pockets of the campaign, while decreasing costs in pockets of lower revenue.


  1. Segment products by brand, then by product type.
  2. Starting at product-type level, note Conv/Click (average value per click) and determine new “Goal CPC,” which would allow for ROAS of 3.
  3. Implement projected bid to best meet “Goal CPC,” and then monitor.
  4. If Conv/Click is too low (say under 0.30) to realistically gain any impression share, EXCLUDE the product type to eliminate any wasted spend.
  5. If there is minimal data at the product-type level, apply 1 to 4, but at the brand level.

Not only did this allow us to eliminate wasteful spending for ineffective products/brands, but also we increased and scaled in areas with an already high ROAS (likely less competitive brands).

Tip: Notice the “Click Share” and “Search Impr Share” columns, and the low numbers in areas of higher ROAS. This indicates we can raise bids in these areas to gain more traffic.

If a pocket already has an ROAS of 5+, we may wish to push up bids even further here, to increase volume and still stay within our target ROAS.

Ready to Make Every Click Count?