If audience data is truly the currency of digital advertising and media-buying, then a new “predictive” audience data-targeting platform being rolled out today, could finally do for the mobile advertising marketplace what the litany of advertising technology platforms have done for the online display ad marketplace: become an organizing principle enabling advertisers and agencies to plan, buy and measure the effectiveness of vast amounts of unsold inventory.
The reason that no material player has yet emerged in the mobile audience data marketplace, say the founders of the new company, Adelphic Mobile, is the complexity of data signals that need to be aggregated, processed and analyzed to effectively target mobile users.
In the online display marketplace, the only data comes from a relatively limited array of sources — content and ad servers, ISPs and the user’s browser — says Adelphic Co-Founder Changfeng Wang. In the mobile ad marketplace, the number of data sources necessary to target a user is magnified by the complexity of the mobile industry’s infrastructure, including varying handsets, devices, carriers, browsers, and applications as well as the physical location of the user.
It’s the reason, says Co-Founder Jennifer Lum, that only a fraction — maybe as little as a third of all mobile ad impressions currently being served — are even sold.
Wang and Lum know something about the mobile advertising marketplace. They are mobile industry vets who were key players at companies such as Nokia, Enpocket, m-Qube, Quattro and ultimately Apple iAd, where they met and decided to form a venture that would help organize the disparate mobile data infrastructure.
Their venture is also backed by some pretty savvy investors, including $2 million in seed funding from successful venture capital firm Matrix Partners, which is helping to manage its growth. Matrix General Partner Antonio Rodriguez has even joined Adelphic’s board.
Wang says the real power of Adelphic is its powerful predictive modeling system, which he dubbed a Predictive Data Platform. An expert in machine learning, Wang developed a system that utilizes sophisticated models that can identify the prospective value of a user by sorting vast arrays of data, including the context of the mobile content they are browsing, as well as the offline physical location they are browsing from.
The trick, he says, was developing a system that could scale it across the billions of mobile user impressions.
After completing beta trials with mobile-savvy shops such as Digitas and Mobext, he says Adelphic is ready for its public general release.
Currently, he says Adelphic is processing 12 billion mobile user impressions daily, and he estimates it will add another “5 billion to 8 billion impressions” each month until it achieves a critical mass of the industry — about 100 billion mobile user impressions daily.
“We think we have the critical mass to serve campaigns today, and we’re looking at being at really full production about a half a year from now,” says Lum, adding: “The key term here is ‘at scale,’ which is the ability to provide huge volumes and to do that effectively.”
To read this article in its entirety, visit MediaPost.