Can You Identify Me Now? A Deep Dive On Verizon’s Data Practices

VerizonVerizon bills itself as a triple threat. It’s got mobile, it’s got television, it’s got broadband.

And those three channels form the foundation for a deterministic data cocktail that Precision Market Insights (PMI) – Verizon’s addressable advertising division – is tapping to power Precision ID, the carrier’s answer to the ever-elusive mobile cookie.

When Verizon Wireless first rolled out Precision Market Insights back in 2012, it was to meet market demand, the group’s VP, Colson Hillier, told AdExchanger.

“People have been asking us for a long time why we weren’t doing something more in advertising because of all of our consumer relationships and touchpoints, but that hasn’t been a core business for us until this point,” said Hillier, noting Verizon’s unique positioning as “one of the world’s largest advertisers.”

But now that Verizon has started to focus on the identity question, it’s come out with guns blazing – and they’re pointed directly at mobile.

“Mobile is now the centerpiece of our customers’ lives and it’s got a lot of promise as an ad platform, but today it’s also fraught with a lot of problems,” Hillier said. “Things that have been solved in other channels don’t work well in mobile.”

Cookies top the list. While Facebook is starting to tackle identity with the relaunch of its Atlas ad server, Verizon is taking on the cookie question with a five-pronged approach: deterministic identification, measurement, cross-device targeting, more mobile inventory and privacy-compliant actionable insight.

It’s a work in progress.

“We haven’t solved for all of these right now, but we’re building toward this proposition,” Hillier said, “The asset we bring to bear is the network intelligence that we manage in an anonymized way with clear opt-out.”

Tapping The Data

The size of Verizon’s subscriber base is nothing to sniff at. According to the company’s website, it’s up to 103.3 million wireless customers, 6.2 million Internet users and 5.3 million TV subscribers. It’s a wealth of potential data that Verizon can use to power its advertising business.

But Hillier is adamant about Verizon’s stance on personally identifiable information – PII is utterly verboten without an explicit opt-in.

Around the time Verizon launched the Precision Market Insights group, it also rolled out Verizon Selects, a part of PMI that enables subscribers to opt in to receive more targeted ads – both online and off – based on their actions inside the Verizon network, including web browsing and mobile app usage, as well as location, demographics, interest data and even credit card info and physical mailing address.

That information, Hillier said, is not shared outside the confines of Verizon, which includes both Verizon Communications and its mobile subsidiary, Verizon Wireless.

Back in July, Verizon enticed consumers to opt in to the Selects program with the launch of its Smart Rewards loyalty program, which provides access to offers, points and discounts on merchandise in exchange for enrolling in Selects. Before officially launching its loyalty initiative, customers were incentivized to join Selects with coupons and discounts through a series of pilot programs.

Although Verizon was unable to share the exact number of subscribers enrolled in Selects, a company representative did say that there are more than 2 million users in the rewards program thus far. (It’s relevant to note, however, that the number of Smart Rewards enrollees doesn’t serve as an exact benchmark for Select opt-ins since it’s possible for a user to subsequently opt out of Selects while remaining a member of the loyalty program.)

Whatever the case may be, it’s safe to say that the Smart Rewards program has driven a significant number of Select opt-ins. As Hillier noted: “It’s a quid pro quo exchange of value.”

While Verizon’s hashed anonymous identifier, Precision ID, isn’t wholly dependent on Selects – a user can allow Verizon to match certain bits of aggregated non-PII data against third-party data assets, which Hillier said is “nice” but doesn’t leverage “the true promise of mobile” – the two work best in tandem.

Users don’t opt into Precision ID, per se. Every subscriber on the Verizon network automatically has an ID. But what users do have control over is how that ID is used – in other words how effective it is for advertising purposes – depending on what they decide to do with their opt-in and opt-out settings.

When a user gives the nod to Selects in what Hillier called a “full opt-in scenario,” that person’s data stream, including app usage and web browsing history, is stitched together with a Precision ID – an alphanumeric string that changes every week – and used to create behavioral segments and campaign objectives. For example, Verizon can see that a customer is a frequent traveler who often spends time in Chicago. Based on location data and browsing history, it might make sense to serve that person an ad for Uber the next time he or she is in town.

“The ID helps with that targeting, but the data associated with the ID leverages what that person did based on their opted-in status,” Hillier said.

Verizon has partnerships with marketing data providers like Experian Marketing Services and Oracle’s BlueKai to enable anonymous matches between the Precision ID identifier and third-party data. Although there’s deterministic linkage back to the hashed ID, Verizon’s data partners are not able to collect or save the data profiles.

Big Red also integrates with demand-side platforms (DSPs), including RUN, Turn and video platform BrightRoll to make it easier for advertisers to access Verizon’s PMI offerings. Although the main relationship right now is between the buyers and the DSPs, over time, Hillier said Verizon hopes to provide some form of execution within its own technology stack.

“What we hand over to our DSP partners is nothing more than a unique hashed ID which the DSP uses to run its execution anytime it sees that ID,” Hillier said. “They never see why those IDs were loaded in.”

Although the third-party data partnerships are great for enhancing scale, Hillier said that brands and agencies have also been clamoring for a way to bridge the gap between their first-party data and Verizon’s targeting capability through Precision ID-based segments.

That’s why Verizon recently partnered with Acxiom’s data onboarding service LiveRamp to bring offline and point-of-sale data into the mix. And like with Facebook’s Atlas, that enriched consumer data isn’t leaving Verizon’s sight.

“We do not take consumers’ data and put it out into the wild. … The balance for me is to try and figure out how to deliver the utility that the ad ecosystem wants in a platform where we can respect the consumer’s privacy,” Hillier said. “We will never sacrifice our core business and our commitment to privacy because there’s an additional dollar to be made by pumping data out into the ecosystem.”

Back To The (Walled) Garden

Deterministic knowledge is a powerful tool – or weapon – and if being called a “walled garden” is the price Precision Market Insights has to pay for not leveraging customer data outside of its network, so be it, Hillier said.

“Ultimately, we don’t see ourselves as a data provider; we see ourselves as an ad platform that helps brands and consumers connect,” he said. “And yeah, maybe that is somewhat of a walled garden, but we think we can provide a lot of value in what we do.”

As for the idea of a universal ID that would allow advertisers to track what’s happening among the various gardens, Hillier said he doesn’t think the industry’s headed in that direction.

“Rather than a universal ID, I think there will probably be really rich algorithms that can tie multiple IDs together into a rationalized campaign,” said Hillier, who admitted that this wouldn’t really be an option, though, if all persistent IDs revolved every seven days like Precision ID does.

“In that case, even if you’re running a campaign with one of our strategic partners and you try to say, ‘The next time I see this Precision ID, remember that it was in a segment of 18- to 24-year-old males,’ you can’t because it doesn’t exist anymore,” Hillier said. “It’s not, if you will, a persistent cookie.”

Which is good for privacy, but challenging from an attribution standpoint.

Scale is also an issue – but that’s where a probabilistic player like Drawbridge or Tapad (a Facebook Atlas partner), could step in. Both Tapad and Drawbridge recently rolled out analytics offerings as part of their respective platforms.

“Facebook is not going to point all of its clients to Tapad, for example, and say, ‘Hey, they’ve got this great analytics solution, so you should go and leverage Tapad,’” Hillier said. “They become more of an ingredient to help round out all those impressions that Facebook already has on the first-party sites – but in a scale game like that, it’s probably not a horrible place to be.”

Having started to prove the worth of using Precision ID – in one example provided by Verizon, Kraft and its agency Starcom were able to add 80% incremental mobile uniques to its baseline audience during a Lunchables campaign at 23% lower cost per unique than other targeting providers – Hillier has his eye on increasing the program’s scale through lookalike modeling, building out Verizon’s TV and mobile inventory and developing in-house execution technology.

He’s also interested in enhancing Verizon’s cross-device targeting, as well as what can be done to “federate” or match Precision ID to other identifiers, like those provided by Apple, Facebook and Google, in a privacy-conscious way.

In addition to all that, Verizon is now in the process of rounding out what it calls the “Founding 15,” a group of blue-chip brands, including Kraft and 1-800-Flowers, that will test campaigns using Selects data and get early access to new products as they’re rolled out.

“We’re very focused on how we’re going to look at consumer interactions across television, mobile and PC on our networks,” Hillier said. “And then we’re going to look at how to get scale outside of our own first-party data – whatever makes the most sense in terms of taking first-party learnings and using them to change the overall customer journey, not just the ad impression.”

 

Add a comment

XHTML: You can use these tags: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>