AdWords
2K members online now
2K members online now
Get started with AdWords - learn the basics to get set up for success
Guide Me
star_border
Reply

A/B Display Ad Creative Testing with a daily budget and targeting?

Visitor ✭ ✭ ✭
# 1
Visitor ✭ ✭ ✭

I've been trying to find out how to run different sets of creative on a display ad campaign. I currently have 3 variations on each size I'm using and 3 different labels that they all fall under. I can break them out into labels and view some metrics (clicks, impressions, etc) between each creative set, but I want to be able to view demographic performance as well.

 

I've read that experiments are the way to do A/B testing on different ads, but I need to be able to target geographically (we're a local financial institution). I've considered using different ad groups but I'm told that my ads will then bid against each other, rather than rotate out evenly which is my preference.

 

In a nutshell, I have 3 different sets of creative I want to test against each other and view their performance on a geographic, demographic, and interest level. How can I do this?

 

Thanks!

2 Expert replyverified_user

Re: A/B Display Ad Creative Testing with a daily budget and targeting?

Rising Star
# 2
Rising Star
Hey Daniel, how are things?

If I understood things correctly, I'd have one Ad Group for each Banner size, this way you could add the 3 variations for each size rotating evenly and have a more separate view on the results, and even select different placements if you wish to.

That's my view on this.

Hope this helps.

Leandro Martinez | Basta1Click

Re: A/B Display Ad Creative Testing with a daily budget and targeting?

Visitor ✭ ✭ ✭
# 3
Visitor ✭ ✭ ✭
Thanks for the reply. My concern with that is I'd like to see the performance of each set of creative as a whole, not broken down individually by ad size.

I might not have explained it well. I have 3 different sets of creative (different messages and design) with 8 different ad sizes in each set. I'd like to compare the performance of those 3 sets against each other.

I've been doing this by combining the ad sizes into sets using labels, but this only allows me to view overall performance of the set (clicks, views, etc) and I'd like to view, for example, which of the 3 performs better with women, or which works best with a certain age group.

Re: A/B Display Ad Creative Testing with a daily budget and targeting?

Rising Star
# 4
Rising Star

 Hey Daniel,

 

My view on cases like this with a considerable amount of data to test is that I personally don't run them all together. I'd prefer to set a date range, let's say a week, 15 days or even a month, and run them individually. So run only Creative 1 for let's say 15 days, stop, then, Creative 2 for 15 more days, and son. This for me takes longer, but it's the best way to truly achieve comparison.

 

If you try to run them all together, you would have to create one ad group per ad size rotating evenly (and compare creatives for each ad size), or throw them all together in one group (which with evenly rotation, there shouldn't be a problema bout competing, because google would select them to appear equal share of times). From what I understand, you want a easy way to lay your eyes and see the results? Some times that's not possible depending on the test you're running and would require you to gather everything to create the desirable view.

 


Daniel O wrote:
My concern with that is I'd like to see the performance of each set of creative as a whole, not broken down individually by ad size.

This is what I talked on the last lines up there. This is something that might create confusion at the time to create a/b tests. You should gather and download data to then compile everything together and create your desired view on the matter. In some cases we're able to create tests that doesn't require compilation and are just easy to hit the eyes and see what we want, but very often for a true and reliable test, you need to break things apart, and for the overall view, download and put together yourself.

 

It's like extracting information from a database. The database have all the information but it's not the way you want, it's just raw info, but you need to come up with filters, range, selectors and so on, download that to able to see what you want. Might be confusing, but I hope I could explain my view on this.

 

That's my opinion for this one. Hope it can help Smiley Happy

 


Leandro Martinez | Basta1Click

Re: A/B Display Ad Creative Testing with a daily budget and targeting?

Top Contributor
# 5
Top Contributor

Hi @Daniel O;

This is a nice question; and it is statistically hard to test since many variables involved.

The best way, would be to use "AdWords experiment". This requires, though, some background in  statistics.

A simplified version, without the complexity of statistics, but also without the full confidence of a real s statistical test, would be to test   the new creative in a banner size similar to the  current size,  and test  the difference in CTR. Make sure that you have enough impressions (at least thousands) so, the test will   have at least some statistical reliability.

 

Read more about AdWords experiments;

https://support.google.com/adwords/answer/2385204?hl=en

 

 

Moshe, AdWords Top Contributor , Twitter | Linkedin | Community Profile | Ad-Globe
Did you find any helpful responses or answers to your query? If yes, please mark it as the ‘Best Answer’

Re: A/B Display Ad Creative Testing with a daily budget and targeting?

Explorer ✭ ☆ ☆
# 6
Explorer ✭ ☆ ☆

I think the important thing is to give your experiment enough time, and the minimum is a month. Run them all 3 x 3 in even and sequential rotation, and should one perform so bad, it ought to fall off your radar before the month is out.