May 10, 2012 · 2 minutes

With the announcement today of $6.5 million in Series A funding, Drawbridge has come out of stealth mode to reveal its solution to delivering highly targeted advertising to mobile devices. And – get excited, math-heads – it involves statistical triangulation on a massive scale.

Under the leadership of former AdMob lead scientist Kamakshi Sivaramakrishnan, Drawbridge has found a way to match user behavior between devices without the user having to be signed in to any particular account. The company’s algorithms learn by analyzing vast amounts of anonymous cross-device data. That means Drawbridge can predict with a high degree of confidence that, say, the anonymous dude who was just playing Angry Birds on his iPad is the same guy who just booked a flight on his desktop.

Sivaramakrishnan, who holds a PhD from Stanford in information theory and worked at Google for six months following its AdMob acquisition, says this machine learning creates a more relevant and personalized advertising experience for mobile users.

On the phone yesterday, I asked her if that meant ads are going to start following us into our pockets. Her response was most Googlian. Actually, she said with a laugh, it means "ads are going to be relevant in your pocket."

The math behind Drawbridge’s offering is not trivial. Jim Goetz of Sequoia Capital, who shared the round with Kleiner Perkins Caulfield & Byers, describes it as a “Turing-class problem.”

“We’re very confident there are very few startups, let alone large companies, that would have the skillset to pull this off,” Goetz told me.

The company’s goal, he said, is to democratize optimization technology “across the ecosystem.”

While desktop advertising is a $30 billion industry, the Internet technology sector is yet to crack mobile. Facebook admitted in its S-1 filing that monetizing mobile is its number one weakness, while Google can so far offer targeted mobile advertising only on its own properties.

Meanwhile, the rise of mobile has contributed to a 12 percent year-on-year drop in average cost-per-click price for Google Ads. And while the mobile advertising market is today only a fraction of desktop advertising (expected to reach $2.61 billion in the US this year, according to eMarketer), it is set for explosive growth, especially now that users spend more times in apps than they do on the desktop Web.

AdMob helped lead the way on mobile display advertising, but there have been few significant advances in targeting capabilities since Google acquired the company for $750 million in 2009. That means advertisers so far don’t enjoy the sophisticated targeting and performance metrics for mobile that they can get on the desktop. Drawbridge, which has been in stealth for a year, working with partners on the advertising side and the supply side, has the potential to bridge that gap.

Sivaramakrishnan, for one, doesn’t appear to think Google will be the one to solve the problem, even though last year the search giant hogged 51.7 percent of all mobile revenues. She went into the company in 2010 expecting it to have cracked the code on everything around mobile advertising. "It appeared to me that that was not quite the case,” she said.

She gave the job six months and then left to find a better way.