AdWords is now Google Ads. Our new name reflects the full range of advertising options we offer across Search, Display, YouTube, and more. Learn more

1.7K members online now
Improve your Google Ads performance and boost your ROI, CTR, and Quality Score
Guide Me
star_border
Reply

A/B testing

Visitor ✭ ✭ ✭
# 1
Visitor ✭ ✭ ✭

Hi Google experts!

I have a question regarding A/B testing. I need to test a desitnation URL for the keywords in the Google Adwords. I have tried out this option called Experiment. So I have taken 1 adgroup, duplicated keywords in the Editor. Half of them, I marked as Control, another half as Experiment (in those I have changed destination URLs). But Editor wouldnt allow me to post them as they are dupplicated and marked as red. Do you have any idea of how to deal with it? Have I chosen the right approach?

 

Thanks a lot!

1 Expert replyverified_user

Re: A/B testing

Participant ✭ ✭ ✭
# 2
Participant ✭ ✭ ✭
Hi Natalia

In order to do A/B testing for a destination URL, you don't need to create a duplicate ad group for that. Just ad a replica of the ad copy which you are using in the currently in the ad group with the new updated URL that you want to test. Also please update the campaign ad rotation setting to "Rotate Evenly" option so that both the ads receives nearly same number of impressions. It will be easy to judge which destination URL performed well if both ad copies serves nearly same number of impressions.

Hope this Helps!
Please keep us posted for further queries.

Abhishek

Re: A/B testing

Top Contributor
# 3
Top Contributor

Hello,

 

IMO, it's a kind of duplicacy only where you are trying to duplicate the same ad group with similar keywords.

 

As per my understanding, what you might try could be to add the keyword destination URL to this particular ad group and then select control and experiment. Find more details on how to set up the ACEs accurately in this help center reference: https://support.google.com/adwords/answer/2375383

 

Also, when you set up a campaign experiment, Control/experiment split setting has to be chosen in the beginning.

 

Let us know if that helps!

Pankaj

Re: A/B testing

[ Edited ]
Visitor ✭ ✭ ✭
# 4
Visitor ✭ ✭ ✭

Hi guys,
thank you for replies. To be honest I still can not figure out how it exactly works. I have never done those tests before.
Basically i do not need to check any ad copies and destination URLs of ads. I need to test destination URLs of keywords, since keywords URLS are "stronger" than Ads URLS. I have already read the information from google support on this issue.
So, to make it more simple;
1. I select and add group and create the experiment there with 50% control and 50% experiment,
2. I go to my keywords, mark them as "control and experiment", and then change a destination URLs. But then how Google Adwords keeps a track of my old URLS?

And in order to be accurate, I need to see if the same keywords work better with different destination URL. Could you please clarify it for me?

Re: A/B testing

Participant ✭ ✭ ✭
# 5
Participant ✭ ✭ ✭
Natalia, just make a new replica ad copy in the same ad group where you are using the old URL with a new destination URL. So now your ad group will have two ad copies, one with old URL and another with the new one which you want to test. And make changes to the campaign ad rotation setting to rotate evenly. That's it. I think its easier to this instead of making experiments for A/B testing.

My views!

Abhishek

Re: A/B testing

Visitor ✭ ✭ ✭
# 6
Visitor ✭ ✭ ✭
But I need to test the keywords urls for my case. How should I do the A/B testing then?

Re: A/B testing

Participant ✭ ✭ ✭
# 7
Participant ✭ ✭ ✭
If you have already setted up experiments then select experiments in
segment drop down menu and check that adding another keyword level
destination url option is available or not. If yes then add that, if not
then suggest that you wait for some more community members to reply with
more answers.

Abhishek
Send from handheld.

Re: A/B testing

Visitor ✭ ✭ ✭
# 8
Visitor ✭ ✭ ✭
the article here says in it is necessary to create a duplicate adgroup and mark the original one as Control, and the duplicated one with different URL as Experiment: http://searchengineland.com/how-to-use-ace-for-landing-page-testing-with-minimal-risk-83940
But what should I do if I have several adgroups in the Campaign? Google wont allow me to chose only 2 adroups for experiment..

Re: A/B testing

Participant ✭ ✭ ✭
# 9
Participant ✭ ✭ ✭
Natalia, read the first strategy which is exact same to what i mentioned in my earlier post . I didn't find anything in the article which states that creating a new ad group is must, its just an another way to do A/B testing. Creating another replica ad copy of the current running ad copy with new testing landing page is the easiest and best way to A/B testing for Landing Page.

Abhishek

Re: A/B testing

Explorer ✭ ✭ ☆
# 10
Explorer ✭ ✭ ☆
Hi Natalia,

Welcome to the community. Adding to the great answers below, I suggest you following.

If you wish to test your landing page, then here is some stuff that you can follow.

1) Please create two ad copies under ACE with same keywords and run them for some time to see what results it gives you. This will help you understand effect of your landing page, ad copy and keywords as well.

2) In another scenario, you may create another ad group with different ad copy, different keywords, but with same landing page. For budget part, you may divide your existing budget among them.

A/B testing will help you when you try variations, not repetitions. Hence, you really need to take an experimental approach to come to any conclusion.

I hope this helps.
Kapil M