How T-Mobile Rehumanizes the Digital Experience

And the way we did that was, we did things like ask customers what they wanted. And U.S. customers of wireless carrier, what is the thing that you hate the most? Your contract because you don’t wanna be married to your wireless carrier, you barely want that relationship with them. So why would you want to spend two years with them and have to pay $500 to get out of that relationship? So the first thing we did was get rid of contracts, then we decided what else are we gonna do for you? We’re gonna give you unlimited data. We also introduced the concept of simple global and mobile without borders. For those of you that travel a lot, potentially internationally, how many times have you been overseas, come back, and been like, “Oh shit, I’ve got a $1,000 phone bill?” So we decided to get rid of that, and actually, as of this week, we now have international roaming coverage and text coverage for free in 200 countries. So you guys, clearly this is a little bit pitch that I’m going through here, so sorry for you AT&T and Verizon customers in the room. And then shout out to my T-Mobile Tuesday’s brethren who are in the room here.

So a couple years ago, we introduced our version of a loyalty program, which is basically giving free shit to our customers. Every Tuesday, I recommend downloading the app, whether you are or not a T-Mobile customer, you can still get access to some things and then do things, like we give away Netflix for free. So if you’re a T-Mobile customer, you get Netflix for free. So, great. This is a wonderful history lesson, I’m sure you’re thrilled. But the question is not about our history, it’s about our future. So we’ve got some numbers up here that talk a little bit about our growth, we’ve doubled our customer base, we have the lowest churn rate, which is basically how we measure off retention. So people aren’t leaving, everybody’s joining us, we’re doing great in revenue stock prices, what tripled in the last three years. Okay, fantastic. But what’s next? So in order to get to the next stage, we don’t wanna be a wireless company, we wanna be a digital company, we wanna be a technology company.

And obviously, those of you who are watching the news or, sort of, seen some of that come to light, can’t talk a lot about it today, but in order to do that, we need to rethink the conversations that we’re having and how we manage those conversations internally. So we wanna be able to have a conversation where we change the dynamic and we say to a customer, “We are what you want, we have what you need, we fit with your life, and most importantly, we love what you love.” So I’m gonna walk you through some examples of some tests that we’re doing and then talk a little bit about what that means from an organizational perspective. So, one way we talked about we fit with your life. So for you travelers, I mentioned, you can go to 200 countries for free, be able to text internationally. The conversation that you’re having with a business traveler, and the conversation that you’re having with a frequent traveler, budget traveler are completely different. Those are two fundamentally different humans with different behavior.

So this is a test where we shared a completely same value prop, same product, free international coverage, but to a, sort of, work audience, business audience versus a, sort of, generic travel audience. So 50% lift and engagement on the left here for that business audience. Another example, again, if you’re a Verizon customer, you might have seen this experience if you’re actually shopping on the T-Mobile site. We changed this experience in our cart to target the brand perception of Verizon customers by showing them what ex-Verizon customers had said about us. So you’re more likely to believe in the value proposition if you could hear it from one of your former peers, effectively, who’s telling you, “No, T-Mobile service is great.” I joined, and we saw an 80% lift for the Verizon audience in this instance. One of my personal favorites, how many baseball fans do we have in the room here? Okay. How many of you baseball fans are T-Mobile customers? Robbie, did you take the MLB TV offer this year?

Robbie: Yeah, I watched that on my T-Mobile plan years ago.

Lisa: I love it when this works. So we wanna be able to have a conversation with the customer that says, “We love what you love.” We have all of these wonderful benefits, one of which is we provide free MLB TV for our customers as part of our sponsorship package with MLB. The challenge we have is that often we wanna talk about what’s most important to us. When MLB TV launch was about a two week offer, Samsung, the GS9 was also launching, that device is the most important thing we have to talk about. But for a baseball audience, it is the most irrelevant thing and they really only wanna hear about baseball. So by simply ring-fencing, and I’ll talk a little bit about the technology behind this later, ring-fencing the audience that loves baseball and promoting this in our communications, we had a 300% lift in customer engagement and a 9% lift in actual shopping behavior for prospects. So you wouldn’t actually think that MLB TV and like that as an offer would get someone to convert, right?

So it takes a lot to switch carriers, but you’re seeing a significant lift when you actually talk about something that they love. So the question is, how do we do this? And to be fair, this is an evolution, this is a journey that we are on ourselves. For us, this started with the idea of, could you change your team’s behaviors? Not so much just the technology, but the actual culture of your company. And this manifests for us in a team and we use this analogy of a party host. So a party host generally, and we’ve got several great event hosts here today, are some of the best marketers in the world. They know who you are, they know why you’re there, they have invited you for exactly the reason that’s gonna bring you to that party, and then when you show up, they’re curating that experience for you. They are obsessing about what is happening with you in that moment, whether it’s…you know, you’re gonna be there, and you’re gonna be compelled by going to the buffet and having that smoked salmon because you’re new to Seattle. Or, “I’m gonna take you over to the craft cocktail bar because, you know, I know that’s really what you want today.”

I saw a red drink a little earlier today. So, you know, and by early, I mean before noon. Or if it’s…you know, there’s a bouncy house out back and you’re gonna claim it’s your kids that wanna go on it but we know it’s you. So these ideas, these are the great marketers obsessing about that customer experience every day, based on meaningful insights of using data to make decisions and working on continuous improvement through test and learn. A second element that we’re working on is rethinking how you structure your communications. So again, I mentioned that when we launched baseball, the most important thing we had to talk about was the launch of a Samsung device. So that’s, sort of, your general campaign. And there’s always gonna be those things, there’s always gonna be the primary offer, the major launch that you’re thinking about. But we’re, sort of, turning it on its head and saying, “If that customer has taken an action, that is the only thing that you should talk to them about.” So if there’s a trigger out there that they’ve abandoned a cart, it’s their birthday, right? You know something about them that is the most important and only thing you should talk to them about.

The interest models, sorry, the customer interest score are more about someone has shown an interest. So they’ve been on your website, clicking through, looking for Netflix, you know that they’re a Samsung user because you’ve, sort of, seen them show that interest or they’ve clicked through on one of your media displays, for example. The models are an interesting one and this is an evolving space for us, where you’re actually using machine learning and AI to do the decisioning for you. So it’s no longer, sort of, “I’m going to decide what is most important,” or “We’re gonna have, you know, 50 people in a room debating which thing is most important,” which I’m sure everyone experiences around here. But, “I’m gonna leave that up to the decisioning models and I’m gonna use a little bit of my information on what I know and what I prefer, combined with what that model is gonna tell me.” So you’re rethinking, sort of, your structure based on the intent that your trying to drive. The other way, this is a little bit of an eye chart on purpose but for your data scientists and tech nerds in the room, this is a shot of Adobe Experience, oh sorry, Adobe Audience Manager.

And the actual…This is what was sitting behind the MLB campaign. So this is actually a screenshot of something that one of the folks on our team, Clayton Smith, produced, and it was a combined audience using display ad impressions. So what was actually happening in media so that if someone saw our media ads on mlb.com, whether they click through or not and they showed back up on the website, they were part of that audience. If they had redeemed in T-Mobile Tuesday’s the prior year so, again, we know they love baseball because they got it last year, again, it was the only thing that they saw. If we saw searches and look like models and we had a few other sources of information, that became who those baseball lovers were to us, and what they were gonna see in that moment that led to those amazing results. So this is just one example of different ways we’re using data to drive decisions. So I mentioned that through my team of experienced managers are obsessing about data, then they’re working with, sort of, their data scientist brother and say, “What can I do differently today using some other source of data?

Another key element of this is democratizing this. So I think, at least for us, there’s always gonna be that guy, you know, you got that one guy in your building, right? The one data scientist, the one target guy, which leads to an amazing backlog because everything you’re trying to do goes through one guy. Our guys are called Derwin and Chase. So there’s two of them, technically. And part of that, you know, we’re putting barriers in our own way that says, “You have to have custom code to be able to do a test,” or these, you know, superpowers are so extreme that no one can possibly learn them. We know as marketers that the technology has evolved, well, that’s not true anymore. So if we can, one, choose the technology and have the frontend UIs to just make changes, and then we democratize those skills by allowing and teaching our marketers to use them, shocking, right? That we have automatically increased the velocity of our organizations. So we as T-Mobile are moving incredibly quickly, the pace of change in our industry is extreme. So the question is, how do we match that? So if you guys wanted to 50 X your content tomorrow, could you do that? We couldn’t either, right?

So this is a methodology by which we’re choosing to, like, instill velocity in the business by democratizing all of those superpowers. Whether it be analytics, testing, target, or actual content management, trying to get to the point where you have a vertically integrated marketer who’s not just coming up with ideas, but just doing the thing, which is surprisingly difficult in a large enterprise, but this is the direction we’re moving. So key element of our strategy here. So the interesting part of this is, again, we have this team, we now have these superpowers, we’re on this journey but it’s by no means easy, right? You’re talking about rescaling your populations of workers, there’s a huge amount of, sort of, emotional anxiety that goes along with that, right? But in the space of change that we’re dealing with and in terms of what’s coming next, we’re at a moment and a pivot point where our next step is potentially doubling our customer base, potentially expanding our product lines from, you know, 1 to 10 in the next year. So there’s an immediacy to this need, to be able to run content at speed, to be able to make decisions based on data, and to be able to test and learn.

So it’s nothing based on opinion and gut feel but actual fact and performance. And that the customers are telling you what they want and we can have a conversation with them that’s appropriate to what they want. So before I close here, that’s another shout out to some of our Tuesdays friends here. We have been working with [inaudible 00:15:59], I know landed another…talked a little bit about, sort of, using the love scores in your communications. So we’re doing something a little bit similar, where, again, for those of you who are familiar with the love dialogue in the T-Mobile app, and the T-Mobile Tuesdays app, we use the survey mechanisms to collect NPS on our customers. You can, sort of, know who your promoters and detractors are. And so we, sort of, asked the question, “What if we could use that information and actually target on that basis?” So could we just say, “You’ve told us that you love us, we love you too. We love you so much that you’ve become the Beta users on our next product launch. We love you so much that you get unique access to, say, a hypothetical launch of a TV product later this year, which we bought last year. We love you so much that we are gonna have that unique conversation with you every day based on what you tell us you want. So I guess I just have one question for you guys, ‘Are you with us?'”

Woman: Okay. So we do have some questions. The first question for you, Lisa, is, you mentioned ring-fencing customers, what does that mean?

Lisa: Sure. So effectively, we build audiences. So when I started to say ring-fencing, we are segmenting customers on certain basis of attributes, and building them an audience manager and then presenting them differentiated content. So when I say ring-fencing, it’s effectively segmentation, depending on your preference of language. I know marketers have different ways of saying it. So, yeah.

Woman: Industry jargon. So thank you. Second question, how do you measure customer engagement?

Lisa: So that’s a fairly broad question. So I think in each of these instances, in the examples, it was a little bit different. I think, depending on the channel, it tends to be a little bit different. So you have customer engagement as it’s manifested in particular channels, right? Whether it be through click-through rate, sort of, engagement, and multi-page visits. Again, we’re, sort of, gonna get deep into depending on your channel preference in this room, there are different metrics that you would use for engagement and, sort of, measures of success. I think when we think about overall financial success, engagement is about retention. So you’ve stayed with T-Mobile, that churn numbers or representation of that, you continue to be a customer with T-Mobile, and then if you sort of backup from that, it’s engagement with our products. So, do you use the products regularly? And then you can, sort of, say, engagement in a channel and our marketing asset, that, kind of, backs up from there.

Woman: Awesome. Any other last-minute questions, no? In the back there?

Man: With the democratizing data, did you get pushback from legal…from other multiple sources? How did you wrangle the problem of making sure the data that go on right here with Netflix?

Lisa: So I wouldn’t say legal, it’s usually our blocker on that. I would say in terms of democratizing data and the blockers that you have, I think it’s more about getting your actual data teams aligned. So your data scientists and your analysts. So, first of all, you have to have a clean foundation, right? The data infrastructure and the data architecture has to be set up in a way that can be interpreted by your UIs, right? Your user interfaces and your frontend systems. So at an enterprise of our scale, we’ve got whole technology teams and data technology teams sitting at that lower level that sort of govern the data lakes, right? Then you’ve got, sort of, your interpretation layers and then you’re, kind of, reporting layers on top.

So when you’re democratizing data, you, sort of, have to have agreement on governance and then you can give access to, sort of, your frontend UI layers to your marketers and analysts, or marketers and basic analysts, without going into manipulation of actual source data. So it’s like…it’s a lot of a deep information there. But, yeah, I don’t think the intent is really ever to have your marketers in the engineering technology systems, right? You would have governance around those. And obviously, we have protections and things like customer information and things of that nature. Usually, when you’re looking at engagement data, you can anonymize it so you’re not saying, you know, “I know your phone number, and you have done this thing, and I know your name and you’re…” all this kind of stuff. You’re filtering it through, sort of, your frontend systems.

Man 2: Hold on. What is the process? Is there a huge portion or does the organization have small portions?

Lisa: For the actual data and for Azure layers? That’s a good question.

Man 2: Keep talking.

Lisa: Yeah, I actually don’t know. I mean, I can, sort of, give you a range but that’s not my department. I can’t speak to their head count.