Just about every ad tech company wants to link desktop and mobile in order to retarget ads between the platforms. It’s assumed that it’ll happen eventually, but one company exiting stealth mode today claims the ability to do so now.
“We’re solving some of the biggest challenges in mobile advertising today,” said Kamakshi Sivaramakrishnan, CEO and founder of cross-device ad tech firm Drawbridge. Specifically, Drawbridge aims to fix “the lack of sophisticated targeting and lack of insights into audience behavior in mobile” through its proprietary “self-learning ad technology,” she said.
Ok, cool. But what makes Sivaramakrishnan’s startup so special if most big mobile companies are conceivably also looking to solve the same issue? For starters, Sivaramakrishnan. The Stanford Ph.D. previously worked as a senior research scientist at one of the largest mobile ad networks AdMob—and then joined Google’s display group when that company acquired AdMob in May 2010.
Sivaramakrishnan left Google in October of that year to start Drawbridge, where she's been quietly cooking up algorithms to associate desktop and mobile devices.
The quest to connect desktop and mobile ads took a hit in March when Apple began to deprecate the use of UDIDs (unique device identifiers) by mobile app developers and advertisers. UDIDs could be used to monitor users’ interactions with ads on a device and track conversions, but can also flare up privacy concerns. Drawbridge claims it can sidestep that issue by algorithmically correlating data to make essentially highly educated guesses as to whether a desktop user and mobile user are one and the same.
“We crunch vast amounts of data and build probabilistic models by which we make associations between a user on multiple devices. It’s a statistical technique, so we are not using any deterministic pieces of information to make these correlations,” said Sivaramakrishnan. In other words, without actually identifying a specific person's iPhone, for example, Drawbridge believes it can crunch enough data to figure out that a person surfing the Web on their mobile device is the same person identified surfing the Web on their laptop later on.
How exactly? Drawbridge says it tracks and collects anonymous desktop Web, mobile Web and mobile app user interactions, which can be blended to create an anonymous user handle. Its technology then tracks those requests over time in order to infer ties among requests. Got it? Sivaramakrishnan wouldn’t discuss what signals Drawbridge’s algorithms track to inform their modeling, but Mark Connon, evp of corporate and business development at mobile ad exchange Nexage, said “there are upwards of 100 different attributes mobile devices can put off [that don’t include a user’s personal information] that may or may not be used.”
“It’s not like I’m looking up certain pieces of information like a phone number or email handle or Twitter handle or anything of that nature,” said Sivaramakrishnan. “It is a learning mechanism. We need multiple observations before we can reliably say that this is most likely the same user on multiple devices.”
Once Drawbridge’s inference modeling hits a certain confidence threshold that a desktop user and a mobile user are the same, it pairs that user between those devices. With that knowledge in hand, Drawbridge, which works directly with publishers and ad exchanges, is then able to execute buys on behalf of advertisers targeting that presumed user.
Sivaramakrishnan said Drawbridge’s technology would allow advertisers to retarget consumers between devices, so that if, say, an automotive brand saw a user visited its desktop site, that brand could then recognize that same user on his or her mobile phone and run an in-app mobile ad promoting one of its models or vice versa.
For now, Drawbridge is focused on desktop and mobile devices, but Sivaramakrishnan the company’s technology would be able to scale up to include connected TV devices when they become more ubiquitous.
Drawbridge is in beta with a number of customers testing the platform and has also raised $6.5 million in Series A funding from top Silicon Valley VCs Kleiner Perkins Caufield & Byers and Sequoia Capital, which will be used to grow the team beyond its nearly 20 employees. Sivaramakrishnan said she’s looking to hire across the board: product, sales, marketing and engineering.