How Ads Follow You from Phone to Desktop to Tablet

Advertisers are increasingly using technology that targets users across multiple devices, and it’s working.

Imagine you slack off at work and read up online about the latest Gibson 1959 Les Paul electric guitar replica. On the way home, you see an ad for the same model on your phone, reminding you this is “the most desirable Les Paul ever.” Then before bed on your tablet, you see another ad with new details about the guitar.

You may think the guitar gods have singled you out—it is your destiny to own this instrument!

For advertisers, the process is divine in its own right. Over the past year, companies have substantially and successfully stepped up repeat ad targeting to the same user across home and work computers, smart phones and tablets. With little fanfare, the strategy is fast becoming the new norm.

“You really have a convergence of three or four different things that are creating a tremendous amount of change,” says Philip Smolin, senior vice president for strategy at California-based digital advertising agency Turn. “There may be one wave that is small and it doesn’t move your boat very much, but when you have three or four medium size waves that all converge at the same time, then it becomes a massive wave, and that is happening right now.”

One of these recent waves has been greater sophistication of companies engaged in “probabilistic matching,” the study of millions of Web users to determine who is likely to be the same person across devices. For example,Drawbridge, which specializes in matching users across devices, says it has linked 1.2 billion users across 3.6 billion devices—up from 1.5 billion devices just a year ago.

Another trend making all this matching possible is the continuing transformation of Internet advertising into a marketplace of instant decisions, based on what companies know about the user. Firms you have never heard of, such as Drawbridge, Crosswise, and Tapad, learn about your devices and your interests by tracking billions of ad requests a day from Internet ad exchanges selling in real time. Potential buyers see the user’s device, IP address, browser, and other details, information that allows for a sort of fingerprinting.  “We are getting very smart about associating the anonymous identifiers across the various devices,” says Kamakshi Sivaramakrishnan, founder and CEO of Drawbridge.

For example, a cell phone and tablet accessing the same IP address at home would be one clue, as would searches for the same product. You might look for “Chevy Cruze” on your phone and then search for the same thing on your laptop. The same geographic location of the searches within a short time period, combined with other information, might suggest the same user.

In the last six months or so, these companies say they have sharply increased the accuracy of probabilistic matching. A Nielsen survey of Drawbridge data released in April found 97.3 percent accuracy in linking two or more devices; anearlier Nielsen survey of Tapad found 91.2 percent accuracy.

Another wave feeding the fast growth of cross-platform advertising is the stampede onto mobile devices. Just last month Google announced that users in the United States, Japan, and eight other countries now use mobile devices for more than half of their searches. U.S. mobile traffic soared 63 percent in 2014 alone, according to a report from Cisco.

Many consumers search on mobile devices but buy on larger-screen computers, giving advertisers ever more incentive to track across multiple screens. Ad agencies are also breaking down traditional walls between video, mobile, and display teams to forge a more integrated approach.

For example, Turn recently worked with an auto insurer’s campaign that started with a video ad on one platform and then moved to display ads on other devices. The results of such efforts are promising. Drawbridge says it ran a cross-platform campaign for women’s sandals in the middle of winter for a major fashion retailer and achieved three times greater response than traditional Internet advertising.

People who prefer not to be tracked can take some countermeasures, especially against what is called deterministic tracking. Signing into Google or Facebook as well as websites and apps using those logins confirms which devices you own.  So you can log off Facebook, Google, and other accounts, use different e-mail addresses to confuse marketers and use masking software such as Blur. “But the probabilistic stuff is really hard to stop because it is like all the detritus of one’s daily activities,” says Andrew Sudbury, chief technology officer and cofounder of privacy company Abine, which makes Blur.

By: Adam Tanner