How To Split Test PPC Ads in Google Adwords

24 August

How To Split Test PPC Ads in Google Adwords

We've heard time and again the importance of having a high CTR (Click Through Rate) on Text Ads. At the same time there are many factors which can affect this number. One of these factors is the quality of your Ad Copy, and its appeal to the searcher. This can be summarized by

Poorly written copy = less clicks = lower CTR = higher average CPC (Cost Per Click)

Well written copy = more clicks = higher CTR = lower average CPC

This means that well written copy can take your campaign from under-performing to market-leading. Writing good ad copy doesn't have to be a guessing game though.

This is where Split Testing Ads becomes so important in Google Adwords. Here are some best practices that I've used to boost CTR's and get better pricing than the competition...

Step 1. Throw Spaghetti on the wall, and then see what sticks. Try creating multiple ads with different ad copy and headlines. Let them run for a while and see which ads get better response rates. This is a natural selection approach that works really well when it'd hard to determine if your market prefers one tone compared to another.


I recommend trying no more than five or six ads at one time as you will dilute the effectiveness of the test with more ads. Note: Make sure to have your campaign settings set to rotate ads evenly. This will ensure your ads get relatively the same amount of impressions per ad. This helps ensure you compare apples to apples.

Step 2. Look for common themes among the top performers. Do certain titles seems to resonate more? Or is it similar body copy that tends to do well? Determine which voice and tone seems to work the best.

Step 3. Test the top two ads and variations on the body copy and titles. If you have two variations of titles and two variations of body copy, you should have four variations of ads running against each other. See the example below:

Title A Copy A & Title B Copy B were the original ads. By testing out variations on each you should find 1 of these to have a higher CTR than the others...


By split testing these four ads you now have found the winning variation based on these four versions. You can also deduce the best elements from the ads based on these results for future testing. In this case it is Title A and Body Copy B that positively influenced the performance of the original two versions of ads, while Title B and Body Copy A were less effective at increasing CTR.

Important Considerations: Pay attention that your ads get at least 1,000 impressions to have a high enough sample size.10,000 is better. Why? Making decisions based off of only 100 impressions can leave big variations in sample size. Statistically you could have an error of up to 1% which can add up to big time money in fiercely competitive markets.

1,000 impressions will tell you the difference between a .5% CTR and a .9% CTR. For rapid tests that need to happen quickly this will at least put you in the right neighborhood. However Getting 10,000 impressions will give you the extra layer of definition telling you the difference between .50% and..59%.

There are many more considerations for raising CTR, such as keyword selection, campaign settings, etc. As a starting place though, you can never go wrong with testing your ads. This alone can produce some monster sized gains in CTR's. Good luck and happy testing!


TAGS: Digital Marketing, Copywriting