Our client had been running Adwords text and display ads for years and had a strong understanding of the Adwords mechanics. They also had specific, achievable goals for ROI within their account, but were struggling to make shopping profitable. They hired WTM Digital to manage their shopping campaigns, with a specific goal of 10% Cost-to-Revenue in shopping. In the 4 months before our contract, average Cost-to-Revenue for shopping was 13.9%, and the client wanted a sub 10% figure before the Christmas shopping season so spend could be increased just before Christmas.

Our objective was to reduce our Cost-to-Revenue below client’s goals without reducing sales volume. This client needed to see similar revenue figures with increased ROI to justify shopping campaigns as well as the additional fees associated with hiring an agency.


Our first step was to evaluate all the shopping campaigns for opportunities to close this ROI gap. The first issue we settled on was segmenting all the shopping campaigns by device type, as desktop traffic was already nearly hitting client goals, but mobile traffic was hindering the account as a whole. The client was aware of this, and was modifying bids by device in an attempt to spend mobile dollars better.

The problem with device bid modifiers for shopping is that they can’t overcome the differences between conversion rates across devices, and they make product group and product level bids difficult to analyze. Without a way to see what average CPCs are across all 3 device types, it is often too difficult to understand what is the ideal bid point, and when a product is no longer profitable.

Average conversion rates per device before our work are below:

We segmented the shopping campaigns to focus on ROI and conversion rate. A well-segmented campaign reduces the number of variables, giving clearer data and allows us to improve impression share, better optimize daily budgets, and manage CPCs and bids.


The results were near immediate profitability, just by treating each device accordingly. Account average Cost-to-Revenue dropped from 13.7% to 5.2% in just under 2 months. Separating by device not only freed up the desktop campaigns to spend more per day at a more profitable level, but also exposed data that allowed our team to make more changes to the mobile campaigns until they were also profitable. By segmenting across devices we didn’t just increase profitability by removing mobile, but actually improved it to the point where even mobile, our worst market segment, was exceeding client goals.


Simply excluding mobile would have increased ROI, but the lost traffic would mean a decrease in overall revenue. With our strategy in place, this client spent 38% less than prior to our work, but generated 21% more revenue.

Case Study Highlights

Industry: Computer Parts / Accessories


  • Reach a 10% or lower “cost-to-revenue”
  • Maintain overall revenue


  • Desktop and Mobile traffic individually hit our Cost-to-Revenue goal
  • Overall revenue is up +20%, despite a decrease in spend
  • CPCs decreased, as did clicks, but we hit a better audience when devices were treated separately