AdWords
3.5K members online now
3.5K members online now
Improve your AdWords performance and boost your ROI, CTR, and Quality Score
Guide Me
star_border
Reply

Running experiments on new Banner set

Follower ✭ ✭ ✭
# 1
Follower ✭ ✭ ✭

Hello All,

 

I am trying to use experiments by myself where I d like to keep me "optimize per conversion rotation" but still upload and test new banners (let s say for a 10%).

 

How do I do this? I see that all my ads have now this little logo next to them, why? and what does it mean to experiment only or control only or both?

 

thanks for helping!

1 Expert replyverified_user

Re: Running experiments on new Banner set

Top Contributor
# 2
Top Contributor

Hello floriana,

 

When you set up the experiment settings in the beginning while going into the campaign settings, it asks you how much you want to control and how much you want to experiment. It comes under Control/experiment split which can vary from

90% control/10% experiment to 10% control/90% experiment.

 

what does it mean to experiment only or control only or both?

 

By default it will choose control and experiment when you save the settings. Now it depends whether you want to control only, experiment only or control and experiment both.

 

Once you run the experiment for any particular time period, then you can decide whether you want to launch those changes fully across your campaign or not. For more details, you can refer to this.

 

Hope this helps!

Pankaj 

Re: Running experiments on new Banner set

Follower ✭ ✭ ✭
# 3
Follower ✭ ✭ ✭

Thanks Pankaj but I still don t understand one think.

 

I decided to set 10% "experiment" and the rest "control" (which I hope it means that everything will be as before - ad rotation optimized per conversions) but ALL my ads are now on "control and experiment". Is this ok or should I manually change all the "old banners" to "control" so to tell the system to not change anything from before?

 

 

Marked as Best Answer.
Solution
Accepted by Karl (Explorer ✭ ✭ ☆)
September 2015

Re: Running experiments on new Banner set

Top Contributor
# 4
Top Contributor

Hello again floriana,

 

As per the Google notes, it reads (please read carefully text I marked in blue):

 

"You can make up to 1,000 experimental changes to your campaign's keywords and bids. You can also define entire ad groups to be a part of an experiment and set an experiment bid for the ad group.

 

Make sure that the settings for your keyword or ad are the same as the settings for the ad group they're in – otherwise, they won't show. For example, if an experiment-only keyword or ad is in a control-only ad group, it will receive no impressions. To receive impressions, the experiment-only keyword or ad will need to be in an ad group set to either experiment-only or experiment and control."

 

As per my understanding from above notes, ad group level experimental settings are the key to focus, otherwise keyword or the ad inside that ad group which you want to focus will not be involved in the experiment.

 

Hope that makes sense!

Pankaj

 

Re: Running experiments on new Banner set

Follower ✭ ✭ ✭
# 5
Follower ✭ ✭ ✭

Thank you Pankaj. What I mainly meant is that I didn´t know that running an experiment affect the whole adgroup.

So, even if I want to try out "only" one banner, I have to set the rest of them on "control".

 

I was just worried at first cos I didn t know that an action on those from my side was requested.

 

now it s clearer. How do I see my results though? And how much time/impressions should I allow, in order to judge if a banner should be kept or not?

 

 

thanks!

Re: Running experiments on new Banner set

Top Contributor
# 6
Top Contributor

Hello again floriana,

 

In order to see how the experiments are moving on, click on the segment Experiment and Google will show you the details of that particular campaign. To see how your individual ads are performing, go to ad level inside the campaign and then again click on segment and choose Experiment. It will tell you how ad have been doing since the time you lanuched your experiment.

 

As per Google you can understand the experimental report as:

 

"If your experimental data is statistically significant, meaning that it's likely that any differences in performance aren't due to chance, we'll display an Arrow up up arrow or Arrow down down arrow next to that data depending on whether your performance has increased or decreased. As many as three arrows can appear in the same direction, and the more arrows in the same direction, the more statistically significant the results are. One arrow means that we're 95 percent certain that the change is not due to chance, two arrows means we're 99 percent certain, and three arrows means we're 99.9 percent certain. Two gray arrows Arrows in opposite directions mean the results are not statistically significant.

The more statistically significant the results, the more likely that the results you see in your experiment will continue if you apply these experimental changes. The more traffic you have on a keyword, ad, or ad group, the faster you're likely to get statistically significant results."

 

I hope this should help further!

Pankaj