T-Mobile is sharpening its programmatic sales pitch and releasing new ad platform features this week.
One cosmetic adjustment: The ad revenue group’s name is changing from T-Mobile Marketing Solutions to T-Mobile Advertising Solutions.
“When you say ‘marketing solutions,’ half the people think you’re a CRM and half think you’re the T-Mobile marketing group that buys the media,” said Mike Peralta, GM of the newly rebranded ad solutions group.
But T-Mobile has also launched App Insights, an app-based analytics platform that’s been in beta for the past year. The insights product allows marketers to track, segment and target T-Mobile wireless subscribers based on the apps they have installed on their phone and their engagement patterns, including when they open an app, the WiFi networks they join and the domains they visit in a web browser.
T-Mobile aggregates this data within its App Insights product and uses the information for analytics and to create “personas,” which is its name for cohorts. Grouping people as business travelers, for example, might be based on whether they have ADP or Concur on their phone and regularly book travel through Expedia, said Jess Zhu, T-Mobile’s head of advertising products and development.
Advertisers can target people with ads using the T-Mobile DSP, but no outside buying platform can target using T-Mobile carrier data.
Other carriers have tried – and failed – to create a strong programmatic ad offering, but T-Mobile is taking a different and therefore successful tack, according to Albert Thompson, managing director of digital for agency Walter Isaacson, which was a beta partner for the App Insights solution. Thompson was also a former marketer at Bell Atlantic, which merged with Verizon, “so I already had a perspective on what a carrier could bring to advertisers,” he said.
Rather than trying to create a media empire with an ad business alongside (à la Verizon or AT&T), he said, T-Mobile is coming at what could be a multibillion-dollar ad revenue opportunity squarely from a data perspective.
Tiptoeing around mobile data
But therein lies one of its major challenges: As a carrier, T-Mobile must tread extremely carefully when using subscriber data.
With that in mind, although T-Mobile sees location data, it avoids using it because of the potential privacy perils. T-Mobile doesn’t incorporate location data into the App Insights platform or its own DSP, and if an advertiser wants to use location data, it has to bring its own and work with a separate vendor.
T-Mobile is right to be cautious. Unlike a vendor that makes all or most of its revenue from advertising, T-Mobile also has to protect its wireless business. Putting regulatory scrutiny aside, if T-Mobile started targeting subscribers using location data, it might turn off its subscribers and become a selling point for competitive carriers to woo people away.
But T-Mobile does collect domain-level web tracking data at the network level. Web browser apps are unique in that carriers can observe whether someone jumps from, say, CNN.com to ESPN.com to AdExchanger.com, whereas they can only see whether an app like Netflix or ESPN is open and not which shows people watch or the sports they follow.
Although T-Mobile uses web-tracking data to build its personas, the data isn’t available for marketers to use themselves within the App Insights product, Zhu said.
T-Mobile also draws the line at data collection on Apple’s iOS. No iOS data, including opted-in IDFAs, is available in App Insights and T-Mobile doesn’t target any Apple users via its DSP.
T-Mobile is missing out on a revenue opportunity, but it’s not worth the risk of getting on Apple’s bad side.
Also, even if T-Mobile did start activating iOS audiences and tracking them for analytics, the data would have a lower fidelity, Zhu said, because there are no longer ubiquitous IDs that track users across apps.
Even so, T-Mobile is “kicking the tires” on incorporating iOS into its App Insights product and overall ad business, Peralta said.
For now, however, the primary iOS use case is studying how Apple device owners compare with Android device owners who have the same app-install and engagement profiles, Zhu said. Apple customers generally have higher purchase power, she said, “but their behaviors and interests may not be actually significantly different from their Android peers.”
Currently, T-Mobile allows advertisers to create addressable audience segments on Android, but only for audiences that are retargeted based on known mobile advertising IDs.
Zhu said T-Mobile is exploring new methods of addressability, likely involving a clean room data solution. “That is on the roadmap for us,” she said, “because we get that not everyone has advertising IDs and, in fact, advertising IDs will probably go away in a year or two.”
Eventually, ad addressability will probably only be possible within walled garden environments, Zhu said, which is why T-Mobile is building more predictive segmenting capabilities that could be an alternative to one-to-one targeting.
Growth by conquest
T-Mobile may be slow-rolling its own ad revenue growth due to potential privacy conflicts – but in other areas, such as conquesting and creating finely tuned mobile audiences on Android, it’s taking the more agnostic (i.e., hands-off) approach that advertisers want.
“Conquesting is actually a primary use case,” Zhu said. “This is a tool that’s readily available to all competitors in the space.”
Wavemaker, another agency beta partner for T-Mobile’s App Insights product, has used the solution to help understand competitive differentiators for app-based businesses, said Delphine Fabre-Hernoux, Wavemaker’s chief data and analytics officer.
DoorDash, for instance, a Wavemaker client, could use App Insights data to observe which of its users also have apps for direct competitors installed. It can see which app is used more or less over time and create segments solely of competitive app users to power a conquesting campaign.
In that case, DoorDash isn’t just analyzing its own campaigns. It can also see whether Uber or Grubhub campaigns are successfully winning market share or pulling away DoorDash’s existing users.
McDonald’s was another beta partner for App Insights late last year. The restaurant chain was able to track app install campaigns based on retention across its own app, the Burger King app and the Popeye’s app to see whether it was managing to retain mobile customers or whether people only installed the app to secure a cheap meal before uninstalling.
But although T-Mobile itself takes a conservative stance on using location data and how it creates personas – and requires a minimum threshold of some thousands of users before it tracks an app’s audience – brands can use the data within App Insights to do their own thing.
One of the things Thompson said he finds most exciting about the App Insights product is the ability to create LGBTQ audience segments. That’s not one of the built-in T-Mobile personas (there are none based on gender or cultural identity), but a marketer could target phones with Grindr installed, for example, or use those audiences for analytics. (However, a T-Mobile spokesperson said LGBTQ dating apps are not available in the platform for user targeting and analytics).
This opens up new avenues for audience targeting, Thompson said. Targeting people by identity, especially LGBTQ or Spanish-speaking audiences, is difficult. There are media properties specifically for those segments, but advertisers want to scale campaigns and find those audiences wherever they are, not just on one or two properties.
With the T-Mobile data, Thompson said it becomes more feasible for, say, a fashion brand to promote a new line of unisex clothes, which are becoming more popular but still can be hard to find when most retailers are built around men’s and women’s sections. Brands Thompson has worked with have also had success using the platform for hiring campaigns, because they’re able to diversify their candidate pools with ads targeting LGBTQ audiences or native Spanish speakers.
So, even though T-Mobile is building guardrails around some data use cases, it may still court controversy because of its open platform nature – but that’s what advertisers are clamoring for.
“The reason why I really wanted to work with T-Mobile is that the world is changing,” Fabre-Hernoux said. “We are losing mobile data signals right and left – that is a topic that keeps me awake at night – and T-Mobile can plug signals back in.”
