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

Ads
2.6K members online now
2.6K members online now
Get started with Google Ads - learn the basics to get set up for success
Guide Me
star_border
Reply

Split testing on display network

Visitor ✭ ✭ ✭
# 1
Visitor ✭ ✭ ✭

I'm getting ready to create a campaign on the content/display network using image ads.

 

If I want to split test the image ads, does that mean that I would create an Ad Group for each size per theme? So, for example, I would create an ad group just for the 300x250 size image ad (for a specific keyword theme) and another ad group for the 728x90?

 

Just wanted to make sure before I set it up ...

 

Thank you,

Ron

 

3 Expert replyverified_user
1 ACCEPTED SOLUTION

Accepted Solutions
Marked as Best Answer.
Solution
Accepted by Zee (Community Manager)
September 2015

Re: Split testing on display network

Top Contributor
# 8
Top Contributor

Hi Ron,

 

I've done it both ways. If you place all sizes within one ad group, setting "rotate evenly" doesn't really do that because not all sizes will be eligible on all pages. But I find it works out fairly well that way. When I put 2 ads the same size into one ad group along with other sizes, the 2 of the same size do rotate evenly. I have set a fairly high default bid for the group to be competitive with the larger image formats. But that doesn't seem to make the CPC for the smaller images higher than if place in a separate group. And I don't have to manage multiple sets of the same keywords.

 

To test ad copy or variations, place them in the same ad group together. For more complex testing, like bidding strategy, use experiments.

 

Best of Luck!

 

Pete

 

petebardo -- Deadhead doing AdWords

View solution in original post

Re: Split testing on display network

Participant ✭ ✭ ☆
# 2
Participant ✭ ✭ ☆

Hey Ron,

 

You can upload all the sizes of the image ads to the same ad-group -(regardless of the images sizes) -

Then, after you have some statistics, you can just compare the performances of each ad (you can manually compare the each set of images from the same size)...

 

Good luck,

 

Asaf

Re: Split testing on display network

Visitor ✭ ✭ ✭
# 3
Visitor ✭ ✭ ✭
Hi Asaf.
If I upload all the different size ads into one ad group, do I set the campaign to rotate the ads evenly or optimize?

Re: Split testing on display network

Participant ✭ ☆ ☆
# 4
Participant ✭ ☆ ☆

Hi Quickron,

 

If you have installed the conversion tracking code in your Thanks page or final page which comes after filling the contact us form or purchase, you can use the optimize for conversion setting or if you are not using conversion tracking you can use rotate show ads more evenly.

Dinesh
My Blog

Re: Split testing on display network

Follower ✭ ✭ ✭
# 5
Follower ✭ ✭ ✭
Hi Ron,

When uploading a new ad group I usually start with "Rotate: Show ads more evenly" so that all ads get a fair chance.

After they get enough impressions, I can see which have a decent CTR/conversion rate and then change the setting and optimize for conversion or clicks depending on the objective.

If you start optimizing from the beginning, the system may put most of the weight on one of the images and hardly give the others any impressions, preventing you from really testing the two ads properly.

Re: Split testing on display network

Top Contributor
# 6
Top Contributor
Hello, Ron.

It depends on what you want to test. If you want to test all sizes against each other, you can, as Asaf said, put them all in one ad group.

If you want to test variations of the same size (300x250 in blue versus the same size in pink) you need ads split by size.

However, when comparing different sizes, please note that some placements have ad units which display one size but not another one, so the competition is not quite even. Also, the necessary bids are different at placement level (if you're competing against somebody with a high bid for one size and no bid for a different size, one ad size gets the chance to show while others not). Therefore I think you should look not for the global performance of the ads, but for their performance on the same placements. And that could be time-consuming.

Or you can take a step back, let them all run for a significant number of impressions and (hopefully) clicks, and decide which one works and which one doesn't, and leaving the "why certain ads work better" consideration aside.

2 cents, from someone going deeper into display ads every day.
Calin Sandici, AdWords Top Contributor | Find me on: Google+ | Twitter | LinkedIn | myBlog
Was my response helpful? If yes, please mark it as the ‘Best Answer.’ Learn how here.

Re: Split testing on display network

Visitor ✭ ✭ ✭
# 7
Visitor ✭ ✭ ✭
Thank you all for your input.

@AdWiser ... yes, I was talking about testing variations of one size ad. But, it does sound like I should first run all the ad sizes in one ad group and then see the results over time.

I can then take the top performing sizes, move them into their own groups and split test variations.

Thanks again!
Ron
Marked as Best Answer.
Solution
Accepted by Zee (Community Manager)
September 2015

Re: Split testing on display network

Top Contributor
# 8
Top Contributor

Hi Ron,

 

I've done it both ways. If you place all sizes within one ad group, setting "rotate evenly" doesn't really do that because not all sizes will be eligible on all pages. But I find it works out fairly well that way. When I put 2 ads the same size into one ad group along with other sizes, the 2 of the same size do rotate evenly. I have set a fairly high default bid for the group to be competitive with the larger image formats. But that doesn't seem to make the CPC for the smaller images higher than if place in a separate group. And I don't have to manage multiple sets of the same keywords.

 

To test ad copy or variations, place them in the same ad group together. For more complex testing, like bidding strategy, use experiments.

 

Best of Luck!

 

Pete

 

petebardo -- Deadhead doing AdWords

Re: Split testing on display network

Top Contributor
# 9
Top Contributor

Hi Friends,

 

Recalled this thread moments ago when I read this: 

 

http://adwords.blogspot.com/2012/04/new-changes-to-ad-rotation.html

 

Quite relevant to this discussion.

 

-Tom

Tommy Sands, AdWords Top Contributor | Community Profile | Twitter | Philly Marketing Labs
Did you find any helpful responses or answers to your query? If yes, please mark it as the ‘Best Answer.’