Quantcast Hires Google Display’s Duggal To Expand ‘Lookalike Targeting’

Audience analytics firm Quantcast has brought in Google’s Jag Duggal to develop more services around targeting and measuring display campaigns. At Google (NSDQ: GOOG), Duggal worked on display strategy and in charge of “Product” for the company’s Brand Display offerings. He served on display head Neal Mohan’s senior team (though he did not report directly to Mohan), his departure comes as Google is working to add supply-side platform AdMeld to its mix of display services.

For Quantcast, adding Duggal is part of a larger strategy to expand its analytics capabilities at time when demand for ever-more sophisticated audience data is crucial. For example, this week ad agency holding company WPP Group launched its audience targeting unit Xaxis as marketers budgets are being funneled through demand side platforms for real-time bidding.

Duggal was responsible for driving the company’s brand display product suite and worked with Brad Bender, Google’s director of product development, and Mohan.

“In all cases, my focus will be on driving breakthroughs in performance in a real time media environment,” Duggal said in an e-mail. “As media becomes more addressable, superior data and algorithmic sophistication are required in order to take advantage of the opportunity that’s there for all sides of the marketplace. I’ll be focused on products that bring Quantcast’s strong position in both to publishers, marketers and agencies.”

He will begin his job at Quantcast working to build on the Quantcast “Real Time* Lookalike product. “The results over the last year have been really promising and they continue to improve at an impressive rate, which I’ve found to be rare,” he said.

Most of the major ad tech firms, like display and video services specialist Collective, demand side platform Turn and Audience Science, among many others already offer some version of “Lookalike” targeting, which is based in the obvious notion that certain types of consumers who share specific characteristics (say, being a 29-year-old mother or a 40-year-old cycling enthusiast) are apt to engage in similar online activities that can serve as a basis for targeted ads.

“In almost every circumstance, however, they lack both the data and the algorithmic sophistication to make Lookalike targeting really work,” Duggal said of the competition, though he did not refer to any rivals by name. “The holy-grail for all display advertising, particularly as we transition to the world of real-time audience targeting, is to reach performance that rivals what has been achieved in search.”

As display begins to operate more like what Google established with search, the best targeting algorithm wins. “Most technology starts with an assumption about what targeting type works best – social or contextual for example,” Duggal said. “Quantcast’s Lookalikes start with no pre-conceived notion of what works – turning the process on its head. We start with the event that an advertiser wants to drive (e.g. an online sale) and then our systems learn what targeting attributes work best and apply that knowledge second-by-second in real-time. The analysis is done by machines not people, at scale not in niches.”