Digging Deeper Into Social Banners: Interview with Seth Goldstein, CEO of Social Media

Earlier this week, Social Media Networks, one of the largest ad networks for Facebook and other social network platforms, announced the launch of Social Banners, a new advertising product based on users’ “FriendRank” – a new proprietary method of examining user behavior and calculating the likelihood that a user’s behavior will be influenced by a given friend’s.

While Social Media has many privacy and data portability waters to navigate as it collects greater amounts of behavioral data on Facebook users, in the end, “Social Banners” must produce measurable improvements in performance in order to gain significant traction. I sat down with Social Media CEO Seth Goldstein to learn more about why he thinks the Social Banners value proposition is real. (Disclosure: Social Media advertises on this blog.)

Seth, what kinds of advertisers do you think will benefit most from Social Banners and why?

Brands will benefit the most.  Social Banners provide brands with a way to spark social interactions involving their products and offerings from within the banners themselves (no application install necessary).

In the past, many brands have been adamant about building Pages and Applications on Facebook, and then buying media to promote those properties.  While this certainly isn’t a bad thing, it is riskier, and the success of the campaign largely hinges on the quality of the application.

However, the primary reason that brands want users to engage in their applications is not just to have an application or to have a page.  It’s because they want users to interact, recommend, and engage with their product socially, and applications and pages were the only things that facilitated this.  With Social Banners, we enable brands to shortcut the application and page installation process and have that social interaction take place within the banner, with just one click.

Additionally, for brands who *do* want an application, we feel that social banners are a more effective means of driving users to those applications, as we are able to mirror the functionality of the application inside of the banner, which provides a more seamless transition from banner to application, resulting in less user drop off upon install.

What kind of performance increases are you seeing so far, and do you expect to see?

Social banners tend to perform at approximately 2-4x the rate of traditional banners, depending on the social content within them.  However, we also have a proprietary technology called FriendRank that adds a new dimension to ad targeting by selecting the ads not only by application context and user behavior, but also by determining the relevant “social distance” of the viewer and the individual inside of the ad.

For example, if two of your friends recommended a movie to you, one of the two recommendations is going to have a higher psychological impact.   FriendRank determines which of these two people should be presented to you in order to maximize the effectiveness of the ad for the advertiser, and the relevancy of the ad for the user.  People have, on average, 150 friends.  It’s not enough to just stick some of them in ads.  You have to know who the right friends are, and that’s what FriendRank determines.

What case studies and examples are most illustrative of the potential value created by Social Banners?

We helped to create an application for BMW for their new 1 series car release.  The application, Joyrides, allowed individuals to select a car, a few of their friends, and a destination you would all drive to.  This information was shared with your friends through news feeds, status updates, and notifications.  In order to promote it, we created an ad which presented the viewer with pictures of several of their most trusted friends (as determined by FriendRank) which plainly asked, “Which of these friends has shotgun on your BMW Joyride?”  The content of the ad allowed users to transition seamlessly into the application, leaving little doubt as to what the application did.  And, there was statistically significant data showing that users were several times more likely to click on the ad if it contained someone we determined to be highly trusted, as opposed to one of the viewers random friends.