• by Phil Leggiere , November 3, 2006

Consumer-oriented social networking sites like MySpace, YouTube and dozens of lesser knowns have quickly come to occupy center stage as the "next big thing" in online advertising. At the same time, they remain unproven as serious ad venues. Though they've succeeded in amassing tens of millions of eyeballs and incredible diversity of content, especially video, they've yet to establish the third indispensable component of online ad success, tools for tracking, measuring and ultimately targeting their users by interest and activity. In the conversation below, Randy Selman, CEO of rich media and digital asset management firm Onstream Media, discusses the infrastructure challenges facing user-generated content and video networks looking to target and monetize all their content.

Behavioral Insider: What do you see as the biggest gaps, misconceptions or limitations in current thinking by Web publishers and marketers about audience/advertising targeting?

Randy Selman: One assumption behind behavioral targeting is that advertisers can identify more likely prospects for their message based on knowing what users do and where they go online, that is, how they interact with content. This is perfectly logical but only feasible to the extent that content is made intelligible, easily searchable and navigable. Most Web publishers, and this is particularly true of the social network sites everyone is so excited about, really haven't done this in a way that makes for targeting in a sophisticated manner. Their goals are more ambitious than their ability to leverage their own content.

BI: What kinds of supportive infrastructures have you been working on that can bridge that gap?

Selman: Audience and advertising targeting techniques are continually evolving in much the same manner as other online technologies. To fully leverage these tools and capabilities in an ever-changing environment requires that organizations first recognize the full value of their content. In other words, to put it very succinctly, if you want to successfully traffic ads to content, then you need to be able to first locate and distribute the content successfully. From there, the more information you have about the content in the form of metadata, the more information you can leverage to provide a targeted ad experience. Our platform also enables us to build custom Web applications and integrate with third-party ad solutions as needed to provide the desired end-to-end online experience.

BI: What are some of the unique aspects of digital video asset management as it potentially relates to behavioral targeting?

Selman: To enable sophisticated targeting you need a rich set of video search, video playback, and ad serving features that may be customized as needed to meet specific customer requirements. For example, we've applied these capabilities to integrate with the DoubleClick DART Motif solution to provide one client, Rodale Publishing, with a custom video player that incorporates synchronized streaming video ads, advertising curtains and banner advertisements. Rodale receives detailed weekly reports of video site usage, which they may leverage as needed to make decisions regarding video publishing and ad campaign management.

Some other clients, including Discovery, are using the platform to transform high resolution streams into a myriad of streaming formats and bandwidth speeds. We also offer the means to provide automated ingest and digitization of long and short format programs, along with the ability to catalog each discrete scene as an independent media asset and comprehensively tag each video scene and its contents with relevant metadata.

Additionally, our fast multi-format playout center is capable of streaming video from multiple media player formats, such as Flash Video and Windows Media. This allows clients to deliver their assets on multiple platforms such as PCs, Macs, mobile smart phones, and iPods, or to package and sell video-on-demand. Our automated indexing and metadata tagging capabilities are used heavily by AOL to provide quick turnaround of timely content, such as news and sports content. AOL also leverages our professional editorial services to catalog content with an extensive set of descriptive metadata, which is then used to provide a compelling online video experience.

BI: How do you approach tracking and targeting consumer video usage?

Selman: What's happening is--whether we like or not--the marketplace is moving to rich media content. Everybody knows now they want to jump on the bandwagon but they don't know how. The first scenario is to provide meaning to the content they have.

Some of the tools for this include video indexing capabilities such as keyframe capture, closed caption extraction, speech-to-text, etc., [which can be combined] with professional editorial services to enhance the set of metadata associated with each encoded video. The set of video and clip metadata fields is completely customizable and is designed according to customer requirements. This set of rich metadata may be fully leveraged to integrate with enterprise advertising solutions designed to target synchronized advertising campaigns to specific video assets. It provides rich reporting capabilities that provide detailed information regarding video usage.

BI: Are we at the point where this data can be used for targeting applications?

Selman: The next step is to combine all this information with the native reporting capabilities provided by advertising solutions to provide a more complete multi-dimensional picture of video site usage and behavior by visitors. That's where we see leading edge publishers, and with them, advertisers, going. Once that's accomplished, what are now just unorganized assets can become richly targetable inventory.