Mar 18th 2015
A team of researchers observed and categorized a sample of 38,151 distinct banner advertisements from 1,076 advertisers across 2,184 distinct marketing initiatives. These initiatives spanned 16 verticals served between November 2013 and June 2014, totaling 23,397,541,849 impressions served in the United States.
When determining performance for a creative category or set of attributes, they looked at impression weighted click-through and conversion rate performance. they compared performance of one
attribute (for example, “blue background”) to either the performance for all ads in that vertical or ads with different attributes within the same category (for example, “how do red backgrounds perform against blue backgrounds for retail campaigns?”). The resulting lifts in performance were recorded and compared to produce the insights.
Results from each vertical were then used in a regression tree analysis to determine the overall impact of these measurable attributes on ad performance. The resulting R2 values identified the
likelihood that a specific set of factors influenced ad performance for a given vertical.
Key insights include:
Download the Guide Here: Guide_To_Creative_Optimization