OpenX’s new OpenAudience aims to be "the first people-based marketing for programmatic ads"

Suppose advertisers could target and retarget consumers across the web with the kind of consistency and profiling they can do on Facebook.

That’s the lofty goal for a new initiative from ad exchange OpenX. It recently launched private testing of OpenAudience, which Chief Product Officer Todd Parsons told ClickZ was the “first true people-based marketing for programmatic ads.”

OpenX’s goal with OpenAudience

People-based marketing (PBM) is term used by a variety of identity resolution providers and others. While definitions differ, it usually means the ability to target the same individual consistently across multiple environments, including offline, and to create a profile that includes demographics, activities and other identifying aspects that go beyond simple website browsing or app use.

The most effectively realized version of PBM, Parsons pointed out, is inside “walled gardens” like Facebook, Google, Pinterest, or Instagram, where logged-in users definitively present an identity that is consistent and targetable across the entire platform.

OpenX’s goal is to do nothing less, but for the open web.

An uber-matching table

To get there, it is building a user identity graph that currently consists of about 240 million profiles of users in the U.S., which Parsons said represents virtually every digital consumer in this country.

Additionally, OpenX recently announced it was the first ad exchange to completely move its infrastructure to the cloud, which the company said was an essential first step toward offering this PBM initiative.

And, while PBM has been a hallmark of identity providers like LiveRamp and Tapad, OpenX does not see them as competitors. In fact, it is employing those providers’ identity graphs as just another data layer for its effort.

Essentially, Parsons confirmed, his company is creating an uber-matching table that takes data layers from many sources, deterministically matches them with persistent identifiers like email addresses, and tracks them by such means as synched cookies, mobile device IDs and connected TV identifiers to create trackable profiles.

Additional layers of data

While that might sound like what a provider like LiveRamp already does, Parsons points to two key differences: this will be an open API that OpenX hopes will be adopted across the industry, and participating publishers can actually receive additional layers of data about their customers and visitors.

So, for instance, if User A logs onto the website of Publisher X, and has previously given appropriate consent, Publisher X can call upon OpenAudience to add additional data about that user.

She might be a frequent visitor to, say, the website of a sports magazine and another one that promotes Star Trek, two refinements of her profile that could help Publisher X better target content and ads to her.

Other publisher sharing

User A’s profile is anonymized, Parsons said, and if she declined to provide appropriate consent, those additional layers would not be provided to the publisher. He added that LiveRamp has described OpenAudience as the first effort “to push identity into publishers.”

Publishers are key customers of OpenAudience, and Parsons said that could be expanded to other kinds of brands, like retailers and product makers. This approach differs from other publisher-sharing efforts like Sonobi’s, he said, because it is a step in developing a new open identity graph that is optimized for the programmatic delivery of ads.

As the OpenAudience effort is still in private testing, Parsons declined to release any stats on the number of participating publishers, the accuracy of the identity match table or the accuracy of the PBM-based ad targeting. He added that the testing is expected to conclude by Q3.