The AudienceCube’s accuracy rate ranges from 80% to 100%, depending on the number of available data points, claims Colwell. Not identifying individuals on their mobile devices is key for brands that “want to be able to have some level and degree of reliability that their message is being seen by the right person…but in a way that does not involve any bad press,” he added.
Among the clients experimenting with the company’s targeting technology are quick service restaurants, financial services, retailers, and consumer packaged goods companies. Adelphic recently raised $10 million in series A funding.
Another company that offers targeting capabilities that accounts for privacy preferences was founded by former AdMob scientist Kamakshi Sivaramakrishnan. Drawbridge (Q&A Nov. 2012), which raised $14 million in a recent round of funding, uses an algorithm to determine whether a user on a mobile device and a PC is the same person, based on information such as location, IP addresses, and browser type, and other anonymized data.
The San Mateo, CA-based company offers two products, "Drawbridge for Mobile Marketing" and "Drawbridge for Cross‑Screen Marketing." The company has matched more than 500 million users across devices to date, according to Sivaramakrishnan. Depending on user data and the devices being tracked, Drawbridge claims an average accuracy rate for its anonymized tracking system of between 60% and 90%.
When it comes to user privacy, Drawbridge takes a “conservative approach” to tying identities across devices, Sivaramakrishnan insisted.“If a user has opted out of being behaviorally targeted on one device, such as their desktop PC, we're going to take a very conservative stance and assume that the user has opted out of being behaviorally targeted on every device,” she said.
If that happens, Drawbridge relies on performance-based data, such as measuring click-thru-rates on ads. “I would essentially just use click stream behavior and if you're not actively clicking on ads, I have very little data,” Sivaramakrishnan said.
Sivaramakrishnan declined to say how many clients have had users opt out of receiving targeted messages, but noted that it rarely happens.
The demand for delivering the right message to consumers on the devices they’re using has prompted tech firms to offer up numerous products. Other ad targeting players include: BlueCava (Q&A 2011), which lets companies identify internet connected devices including computers, mobile phones, set-top boxes and gaming consoles); TapAd, which enables advertisers to deliver the same ad across different platforms); and the now defunct Ringleader Digital (Q&A 2011), a web-advertising firm which used HTML5′s database-storage capability as a cookie substitute and was hit with a privacy lawsuit in 2010.
As for whether the Do-Not-Track legislation could eventually include additional ad targeting technologies, it would be difficult for legislators to restrict the use of some identifying data, according to Matt Papertsian, research director at the research firm SiriusDecisions.
“Domain registry information [for example] is public data,” Papertsian noted. “Legislators would be locking public data, and I think they would have to think about the potential harm behind that.” At the same time, the possibility that companies could be forced to further reign in their ad targeting approaches continues to be a concern, say marketers and vendors.
Part of the problem is that some marketers still intend to use “spray-and-pray marketing,” techniques, according to Papertsian. “That’s one of the reasons why so many people have bought into privacy. They’re just sick of being over communicated to,” he said. “The irony is that the very thing that the government’s taking away is marketers’ ability to better target [their messages].”