Use a real test, because ad reports can triple your results
Updated: Oct 11
Your ad report and your ad test can tell two different stories.
The study. Researchers from Northwestern University and Facebook used 15 large US ad experiments on Facebook. They compared the true test results with common methods that use data without a test.
The finding. The methods without a test were often wrong. In many cases they showed a much larger effect than the test. Sometimes the difference was three times or more.
Why it matters. Most teams judge ads with reports, not tests. People who see an ad are often people who will buy anyway. A report cannot remove this. So the report gives the ad too much credit.
The takeaway. Use the test tools on your ad platform, such as a lift study or a holdout group, for your largest campaigns.
Want to read more?
Subscribe to plaintalkbusiness.com to keep reading this exclusive post.
