Jason Puckett

February 22, 2016

We Analyzed the A/B Testing Plans of 196 Digital Advertisers, Here's What We Learned.

There are some very common questions in the ad testing world: “How much ad testing should we do?” or “How do our ad test results stack up against other companies?”. Unless you’ve hired a vendor that can relay that information, it’s tough to know how you stack up against other growth marketing teams.

If you’re not familiar with the concepts of A/B Ad Testing or Multivariate Ad Testing, please stop and read some information on the topics. If you are familiar with the concepts and are, in fact, executing a testing plan of your own, here’s some information to help benchmark your strategies.

What Types of Tests Are We Talking About?

The data we are looking at below is from advertisers testing ad creative & targeting combinations on their search, display, remarketing and mobile ads. Meaning how different ad creative performs for specific audiences, ad groups, geographies, devices, etc.

AdBasis has a pretty rare perspective - we get the opportunity to see how today’s top advertisers are implementing their testing plans. We analyzed 196 of our customers, some do a LOT of testing, some a moderate amount… regardless of the activity tier they fall into, we have their data.

We looked at these 196 companies' average monthly testing numbers over the past 13 months. Here’s what we found:

What are Some of the Average Monthly Testing Numbers for These 196 Companies?

We have compiled some data to help you determine if your testing strategy holds up.

Average number of ad tests created per month - 7.43

The average company investing in ad testing is creating over 7 new creative tests per month.

Average Increase to CTR Per Experiment / Per Month - 54.8%

You may think this number is astronomical, and you are right. We actually removed the outliers from this equation that made the numbers even LESS believable (we removed anything that improved CTR by over 100%). Before we removed the outliers, the average increase in CTR was 104%. Which is absolutely crazy.

This tells us two things: (1) Some of our clients are ridiculously good at testing and needed to be removed from the sample and (2) there is such low hanging fruit in the creative testing realm, that on average our customers are improving CTRs in their AdWords accounts by an average of 54% MoM.

Average Increase to Conversion Rate Per Experiment / Per Month - 22%

This is a really solid statistic. It’s telling us that there is money to be made in creative testing and optimization. This is an impressive stat for one main reason, many of these companies have been testing in our platform for many months. Therefore, in order to achieve MoM growth rates like this, they continuously have to be improving over their baselines.

Give Creative Testing a Try

We have excluded the CPA & ROI data from this analysis because not all of our clients track ROI in AdWords and it would have skewed the sample. But please give ad testing a try. Uncovering the optimal ads for very specific situations DOES take work, but the work pays off.

If you’d like to see the full report, or if you’d like to know more about ad testing, please subscribe to our blog, or contact us at HELLO@AdBasis.com

Have a question for us? We'd love to hear from you!

comments powered by Disqus


Popular Articles

Understanding Conversion Layers and Executing Ad Experiments

August, 14th 2014 - Optimization

3 Steps for Building a Paid Acquisition A/B Testing Plan

October, 15th 2015 - Optimization

All Things Tested - 3Q Digital's Sean McEntee

August, 31st 2015 - "All Things Tested"