Fill in the blank. The best practice for a/b testing two different calls to action for a __________ad is to test banners that are __________in every way, with the exception of the call to action.

The best practice for A/B testing two different calls to action for a banner ad is to test banners that are identical in every way, with the exception of the call to action. This ensures that any observed difference in performance is solely attributable to the call to action, rather than other design elements. If the banners were different in multiple ways, it would be impossible to isolate the impact of the call to action. Text ads typically have less visual variation, making the "banner, identical" option the most appropriate for isolating the CTA variable in a visual ad format.