For more information on COVID-19 and how to prevent infection, visit
Log in to your account
Sign up

How To Create Google Ads and A/B Tests on AMP Pages

14 September 2018 | 0 comments | Posted by Che Kohler in nichemarket Advice

How to create A/B tests with AMP pages

So you've taken the plunge and committed to mobile and in doing so chose the path of the AMP protocal which in my opinion is a bit of risk, but hey the internet was built on a cowboy approach to everything. AMP's are designed with the idea for improving the user experience for mobile users visiting your site via organic traffic. 

You would first set up and host an AMP version of your site on your domain and Google would then index and self-host those pages on their CDN for faster response times on their mobile SERPs. 

In theory, it's a great idea, and for the end user it's a great experience, but for long-term growth and the site owner, you're getting a bad deal here. 

All that work for any a bump in organic? What about my other channels? Paid? Social Media? Email? While on most of the other you're still out of luck and will have to create manual ad campaigns and split tests, Google Ads has taken a step to make it easier.

Google Ads for AMP

Accelerated Mobile Pages (AMP) are pages that load near instantaneously for mobile users. To date, you could test the impact of your AMP pages using either Google Ads' campaign experiments or ad variations. 

This would mean you would have to have an AMP structure that makes it easy to find for example hosting your AMP pages on a subdomain replica of your site or a subfolder replica of your current site. 

For example, if I had a desktop URL for an article How AMP works and the URL was:
  • Main site -
The AMP versions could look something like this
  • AMP subdomain site
  • AMP subfolder site -
Which makes it easier for advertisers to do a quick find and replace or regex replacement to make sure mobile ads send users to the correct version. While achievable it can be quite painful to duplicate all your ads like that, hence the introduction of split ad testing.

Google Ads split tests for AMP

Google has introduced a new way for advertisers to compare the impact of using AMP pages vs. traditional web pages with Google Ads. Google already offered the ability to test the two different page types using Google Ads campaign experiments or ad variations. Now they've introduced a new method of testing for campaign experiments which Google calls “cookie-based splits.” 

This new testing methodology is more closely aligned with the cookie-based approach of Google Ads’ ad variations. Prashant Nair, Product Manager, Google Ads, recommends advertisers use this new methodology when testing AMP pages via campaign experiments. Google Ads’ campaign experiments default to using a testing type called “search-based splits,” which assigns users to either a test group or control group every time they enter a search. 

Using search-based splits, any given user may end up seeing both the test and control experiences throughout the day. Cookie-based splits work by assigning users to the test group or control group once and keeping them there. 

That means users will always see either the AMP pages (if they’re in the test group) or still see the traditional web pages (if they’re in the control group). Nair concludes the announcement by saying Google hopes this new testing type helps advertisers “better test the efficacy of driving your search ads traffic to AMP pages.”

Better oversight of your investment

As I've mentioned before opting to go AMP requires some technical implementation along with managing it going forward which requires a sizeable amount of resources like development cost, SEO costs and time. Which up until now has been, pretty much pie in the sky for early adopters of the tech. 

But now that we have Google Analytics, Google Tag Manager and Google Ads support sites can start to push their AMP initiatives and begin to collect data on the format. 

I'm personally very interested to see sites and Google start to release data and case studies around AMP since I'm not entirely convinced that it is the superior mobile option. 

So as we wait the fight will continue to see who reigns supreme as the best mobile format. For more on mobile options for your site check out our post - Progressive Web Apps Now Available for iOS

More on accelerated mobile pages

If you're interested in getting to know more about AMP or want to explore the option for your site, then I recommend checking out these articles.

Contact us

If you want to know more about paid search or mobile marketing for your site, don’t be shy we’re happy to assist. Simply contact us

Tags: mobile, Mobile Marketing, paid search, User Experience

Previous: {{ previousBlog.sTitle }}

Posted {{ previousBlog.dtDatePosting }}

Next: {{ nextBlog.sTitle }}

Posted {{ nextBlog.dtDatePosting }}

You might also like

Rights of a truck driver in the USA

What Are Truck Driver Rights in the USA

08 August 2022

Posted by Arumugam Madhaiyan in Get Court

A look at the various rights truck drivers all over the United States of America should be aware of and how they impact their work and their ability ...

Read more
Motorcycle accidents statistics

10 Motorcycle Accident Stats You Should Know

12 August 2022

Posted by Arumugam Madhaiyan in Get Court

A review of some of the key stats around motorcycle accidents and why you need to remain vigilant as someone who uses this mode of transportation

Read more

Leave us a comment


{{comment.iDayLastEdit}} day ago

{{comment.iDayLastEdit}} days ago


Sign up for our newsletter