Monday Afternoon #HeroConf Keynote Speaker Brad Geddes, Mobile Ad Testing

Monday Afternoon #HeroConf Keynote Speaker Brad Geddes, Mobile Ad Testing

6th May 2014

There is a ton of data inside platforms like AdWords and BingAds, but the likelihood that we’re using all of it to our top advantage is nearly impossible. Brad Geddes is known for being a thought leader within the paid search industry and laid out quite a few ways that we can better optimize our ad copy for mobile and desktop.

Ad Testing Basics

Pre-Enhanced Campaigns, we didn’t need to worry about if our ad copy was targeted for devices. In the “good old days” we would simply split our budgets across three different campaigns that were enabled for one of three networks. Now, we have to pay more attention to where we put our ad copy and how it is performing against others that are also optimized for Mobile or All Devices.

With mobile traffic booming, we know we are going to need a better PPC strategy to be able to capture this new volume. However, there are only an estimated 3% of ads actually opted into the mobile preferred setting per ad. To be able to increase conversion rate and percentage by device, we need to get mobile advertising under control.

Because we work in an environment where we can control the majority of our surroundings (Keywords, ads, campaigns, locations, etc), we should be doing a better job of controlling our ad copy preferences. Here is the procedural rundown from Brad on how to test

  1. Decide what piece of adcopy you’re going to test.
    Will you test the headline? The Description lines? The landing page, or the display URL? Maybe you test them all. Whatever you do, make sure that you are being consistent. “Free Shipping” is an entirely different phrase than “Free Shipping!” so be careful. Use the tools you have on hand. Excel is your friend.
  2. Pick a data set to test and stick with it.
    For clients with very little data or massive enterprise accounts, use “multi-ad group testing” that allows for the same messaging to be displayed across all adgroups.
  3. Apply Labels to your copy depending upon what piece of the puzzle you are testing.
    Make a code and stick to it. This will allow you to remember which pieces of the ad copy you are testing and be able to sort them at a glance. If one set of copy is performing horribly across all adgroups, it will be very easy to turn it off or delete it completely
  4. Let your tests gather data.
    Use the proper ad rotation and minimum viable data testing metrics to be able to gauge your ad’s success.
  5. Put together a clear plan for testing your ad copy to different landing pages.
  6. Report on complete metrics using Conversions per Impression or Profit per Impression.

Ad Rotation & Minimum Viable Data Testing

True Story: If you let AdWords choose your copy based on CTR, aka the Optimize for Clicks setting, Google will make a hasty decision that is likely to throw off your entire test because they will make a decision off of less data. Google isn’t taking into consideration time of day or day of week when they choose the better ad.

Choosing Rotate for 90 days or Rotate Indefinitely will give you much more control to be able to choose the ad that is best performing after a predetermined amount of data. Rotate for 90 days at least allows the ads to assume a minimum viable amount of information before Google can make a decision on how to optimize for clicks.

Often, there is an issue when determining what a minimum amount of data actually is. More impressions will always provide more clarity to the tests.



If you’re looking for a new ad testing tool to help make this easier for you, Certified Knowledge is beta testing a tool which is still in beta called AdAlytics. It will help you start making more informed decisions on what ad copy is actually best performing.

Clearly Test Landing Pages

Ad Copy and keywords will perform differently given a different landing page, but we need to have a solid testing plan in place when testing. Outlining your structure in advance will help make sure that you stay on target and don’t make silly mistakes later in the process.

Brad suggests duplicating the ad and running two pieces of copy alongside each other going to different landing pages simultaneously.


Ultimately, AdWords doesn’t bring you conversions. It brings you traffic. And if you have a poorly converting landing page, then you aren’t going to get any return on investment.

Report on Complete Information

There are so many metrics that we can point to when trying to make optimizations to improve an account. Sometimes when reviewing CPC at the ad level, we can see a hint of the ad quality score. When several ads are receiving less than $1.00 CPC and others are receiving $4.00 CPC, we should be removing the higher cost ad copy because it could be increasing costs.

In addition to reviewing CPC, we try to rank copy by its conversion rate, conversion volume, or click through rate. But really, what we need is to be able to do is combine all of these metrics so it is easier for us to make decisions and to discuss with our boss (or client) exactly what is working and what isn’t.

Brad suggests that we instead look at Conversions per Impression (CPI), Profit per Impression (PPI), or Revenue per Impression (RPI) instead of the several metrics we typically look to when determining the superior copy. These metrics take into consideration your full branding plan, your click through rate and your conversion rate.

About the author

Jasmine Aye Jasmine works at Distilled in their Seattle office specializing in paid search. She has managed accounts in several verticals including finance, ecommerce, non-profit, and travel. She has a fierce determination to achieve excellent growth for her clients and is a passionate people person


  • This field is for validation purposes and should be left unchanged.