Evaluating an ad test is all about aggregated data. In the last edition of this series, I demonstrated how this can lead to the wrong conclusions by going just below the top level. Today we’ll take a look at the lowest level.Continue reading
Ad tests rely on data. The first part of this series was about how using the data the wrong way leads to questionable decisions. This second part is about the data itself being questionable.
When it comes to A/B testing Google Ads, experienced marketers have agreed on one thing for more than a decade: You can’t leave it to Google. You have to take control. The best practices on how to do this are also widely accepted. But guess what – it’s all nonsense. It doesn’t work and never has. Continue reading