experiment control groups versus experiments
I want to test four different ad texts for a new campaign. I would like to use the experiment mode since it automatically measures statistical differences.
The problem is that I have no 'control' group. I have four different new ads (experiments) that I want to compare to one another. Is it possible? The experimental settings require me to set a ratio of experimental to control changes.
I guess I could just set it up as a regular (not experimental) campaign and force the ads to rotate evenly and force them to use the same manual bid?
Thanks for the advice.
Re: experiment control groups versus experiments
You are correct!!;
This is very confusing with testing different ad-copies (Since, your control group is actually your current ad with the current ad-copy)
When testing ad-copy I recommend switching to an even rotation (of ads ), and test which ads are performing better. This is the easiest way, that after a short time shows results that are statistically significant . (You don't need a "statistical analysis" by the system "to accept" an ad-text with a CTR twice or 3 times better than the original text...)
Did you find any helpful responses or answers to your query? If yes, please mark it as the ‘Best Answer’