What’s the most important factor in the success of a display ad? Size? Placement? Not surprisingly, it’s relevance.

That’s according to a study conducted by publisher Condé Nast and research firm McPheters & Company.

The study looked at the effectiveness of display ads that were content-relevant (eg. “food ads running on food sites, entertainment ads on entertainment sites, etc.“). The results: display ads running on a relevant site were 61% more likely to be recalled.

Sites related to social network, shopping, and food had the highest recall; search and portal websites had the lowest recall. The study also found that “There were large differences in recall by type of product advertised“. Again, this isn’t entirely surprising, although it is interesting that the simplest form of targeting (advertising on a site that is relevant to your product) appears to be so much more effective than anything else. I would like to know if any of the ads that weren’t categorized as ‘relevant‘ by the study were served using behavioral targeting. If they were, such as study might hint at the relative ineffectiveness of behavioral targeting.

Of course, there was some bad news: 63% of the banner ads shown to users who were given the ability to browse the internet freely for 30 minutes were not seen. The dreaded banner ad blindness.

On a side note, according to Condé Nast and McPheters & Company, TV and magazine ads excelled in the study. In its report on the findings, MediaPost details that “Full-page, four-color magazine ads were
determined to have 83% of the value of a 30-second television
commercial, while a typical Internet banner ad has 16% of the value

Obviously there are huge price differentials between online display ads and television and print but this provides a good example for why television and print haven’t died off as some have suspected and might be a cue to savvy brands that have cut back on TV and print to reconsider. I’d imagine there are some good print deals to be had today.

Photo credit: jbcurio via Flickr.