How-To: Control Device Bidding Tests to Drive More Conversions at Minimal Risk
Published: November 2, 2016
Author: Kevin Vlcek
As we all know, mobile has become the juggernaut of the SEM industry and now holds the lion’s share of web traffic over its desktop and tablet counterparts. So, it’s time to start thinking about how device adjustments are truly affecting overall campaign performance. I’ve discovered a way to try and push mobile volume in a controlled testing environment where opportunities lie.
Get more conversion volume in a test-controlled environment on mobile in markets where conversions are low.
Identify areas where CPA is low and has room for position movement (dependent on your goals) by device in a particular campaign – or, in my case, a particular market. Once you’ve identified this campaign(s) based on your analysis, select the campaign and create a new draft:
Once the draft is created, you need to name the experiment based on the type of test you’d like to run, e.g. “Chicago Mobile Bid Increase: +20%”. Next, you go into the campaign settings and increase the mobile bid on the campaign by whatever amount you’d like to test (in this case, 20%). Now this is where is where the controlled environment part comes into play. After the changes have been made, select “Apply” and “Run an Experiment,” and decide how much traffic you’re willing to send to the draft campaign with the higher mobile bid.
In my case, it was a 50/50 experiment and if the results drove more conversions within our target CPA, we would apply the bid modifier to the original campaign. Once your experiment parameters are decided upon, you click “Create,” and the experiment will begin. Again, the major benefit is only sending a portion of this traffic to the test campaign in case it delivers low performance.
As mentioned before, the focus of these tests was to try and get more conversion volume in low-converting markets. Results obviously varied depending on the market but were very positive overall!
In the above example, the draft campaign with the higher mobile bid modifier brought in over double the conversions at a marginally higher cost, lowering CPA by $5 and increasing CVR from 25% to 42%. Since the results were very favorable, we applied this modifier to the original campaign. Though not all results were as favorable, the impact to the account was much more controlled using this method. Now that Google has enabled modifiers across all devices, I highly recommend playing around with Draft & Experiments on Desktop, Mobile and Tablet.