Display A/B testing - Ghost ads
We are planning to set up a display campaign targeting a remarketing audience of website visitors.
I would like to create a control group on this display ad. I saw this article mentioning Ghost ads as being the preferred approach: https://www.thinkwithgoogle.com/intl/en-gb/articles/a-revolution-in-measuring-ad-effectiveness-knowi...
How could I implement this measurement in GDN?
Re: Display A/B testing - Ghost ads
I'm not sure I fully understood the statistical technique suggested in the article; Iwould advise to refrain from any statistical technique that could bias the results of the test.
A classic A/B testing is based on randomly selected users in the test group and in the control group. This is how an A/B testing should be constructed. Any attempt to modify the users on either group could bias the statistical analysis.
Also note that a re-marketing list is already statistically biased, and I would be cautious. Those are users already visted your site the number of users (in the test group) might not be big enough to get a reliable result.
Did you find any helpful responses or answers to your query? If yes, please mark it as the ‘Best Answer’