What’s the Best Way to Split Test Google Ad Landing Pages?
When thinking of ways to improve a Google Ad Campaigns performance, testing a new landing page always comes up.
There are times when it can really make a difference in performance and so you should know the best way to split-test different landing pages in your Google Ad Campaign.
There are a few different ways.
Sign up for free access to this post
Free members get access to complete blog posts like this one, plus exclusive Google Ads strategies and case studies.
When thinking of ways to improve a Google Ad Campaigns performance, testing a new landing page always comes up.
There are times when it can really make a difference in performance and so you should know the best way to split-test different landing pages in your Google Ad Campaign.
There are a few different ways.
You can use split-testing software like Mida or Unbounce that will split traffic between 2+ different landing page variations without updating the ads in your Google Ad account.
You can run a Google Ad Experiment.
You can run 2-3 ads together in an ad group with the only difference being the landing page they’re sent to. Be sure to set your ad rotation to “rotate indefinitely” so new ads can get a fair shake otherwise they may struggle to get impressions since by default Google Ads will show the better performing ad.
Of these 3, the last one is the easiest to set up but you’ll have to decide when the test has reached statistical significance and when to pause the underperforming ad. When using a Google Ad Experiment or A/B testing software, it’ll tell you essentially once a test has reached statistical significance and which variation won.
Google Ad Experiments are not really meant for split-testing landing pages, it’s moreso meant for split-testing different settings on your Google Ad Campaigns like bid strategy. You can only run one Google Ad Experiment at a time per campaign so it makes it difficult to a/b test more than 1 landing page variation per ad group.
Also once you end the experiment it ends it for all ad groups in that campaign so if you were testing multiple landing pages in different ad groups this wouldn’t be ideal because they’re unlikely to finish the A/B test at the same time.
Just know that unless you commit to spending enough money to get conclusive results from the split-test then there’s no point in running them and you’re better off just making improvements to your landing pages without testing them.
Not every change needs to be A/B tested.
A/B testing is kind of a buzzword and some companies tend to think more A/B testing is always better, this just isn’t true.
The more limited your resources, the less you should probably be A/B testing and when you do you should test big changes. If you have more resources then you can test smaller things like CTA button colors or whatever but most businesses that aren’t Google I’d argue shouldn’t be A/B testing since they have limited resources and aren't’ statisticians.