Crimefighting Tech Put to Marketing Use

Imagine using your camera phone to snap a photo of a brand logo in a magazine, send that image to a short code, and instantly receive a mobile game in reply.

Sounds far-fetched, but that precise scenario is now playing out in Germany, courtesy of Coca-Cola and its regional CokeFridge marketing initiative.

In a campaign that’s now underway, German readers of teen magazines YAM!, Starflash and Maedchen are invited to snap images of the CokeFridge logo as it appears in print ads, and send them to Coke.

In instant reply, they receive a mobile soccer game called Home Turf. If a recipient’s phone can’t play the Java-based game, he or she receives mobile wallpaper instead.

The campaign uses i-Scout object recognition technology from Neven Vision to identify Coke logos that are sent to the company via multimedia message service (MMS). Neven Vision CEO Hartmut Neven said the response rate for the CokeFridge campaign has so far exceeded the average for text messaging campaigns in the same market, if you take into account the modest penetration of MMS versus the ubiquitous SMS.

This is the first ad campaign that uses i-Scout technology. The platform has so far been used for face recognition purposes by several law enforcement and security clients, including the Los Angeles Police Department and the U.S. military. The company has only recently decided to license it to advertisers, setting up sales operations in the U.K., Tokyo, Southeast Asia and the U.S.

Neven said several major advertisers will soon launch similar campaigns in different countries. He said a U.S. campaign would break before the end of the year, but declined to identify clients.

Neven said the Coke campaign so far has a near-perfect success rate when it comes to recognizing incoming logos — an easy feat compared with face recognition. However, he said it remains to be seen how well it will perform for more complex campaigns and applications involving a great many images.

“In the advertising space, we do not know yet if we can… scale to ten million images,” he said.

Related reading