T-Mobile - AI, Live Translation, and Key CMD Takeaways

Full Transcript

Don Kellogg 0m10s

Hello, and welcome to the two hundred and eighty third episode of week with Roger, a conversation between analysts about all things telecom, media, and technology by Recon Analytics. I'm Don Kellogg, and with me as always is Roger Entner. Roger, how are doing? Where are you?

Roger Entner 0m24s

I'm good. I'm here in Washington, DC in the Rayburn Building about to talk with lawmakers and staffers here about AI, the adoption of AI, and how different parts of the country are already falling behind, and the role of telecommunications in that. But I'll talk about that later. Maybe I'll record it or somebody records and we can share it with people. But I was also this week in New York at the T Mobile Capital Markets Day.

Don Kellogg 0m51s

Yeah. So how was that?

Roger Entner 0m53s

As always, they put down a good show. Very confident. Really, really good numbers, especially when you look into some of the details. I would say, don't look at what people say because there's a lot of hot air often around. Look at what they do.

Roger Entner 1m8s

One of the things that I thought was an underappreciated detail is they said, we are shifting our store mix from stores that are dealer owned to corporate owned. That is usually a sign that you're very bullish about the quality of your gross ads and that it's going up. Because stores are a fixed cost, and so if you have more customers coming and you know that, you're moving it over to a fixed cost rather than a variable cost. I thought that was very, very interesting. What was also interesting was how they're straddling the gap with FWA and fibre.

Roger Entner 1m47s

Because listeners of the show know that the new drug is convergence, and, you know, the combined footprint that they're going for is like 50,000,000 households. That's the size of Verizon, basically. And it's a much smaller share with with fiber, which is also really, really tough for them.

Don Kellogg 2m6s

Well, they don't wanna touch copper. Right?

Roger Entner 2m7s

They don't wanna touch copper.

Don Kellogg 2m9s

There's only so many pure play fiber companies out there.

Roger Entner 2m12s

Yeah. And they wanna be the first one because the second fiber guy has a really, really hard time. And so if you only wanna dance with the prettiest girl in the ball, you know, there might be a line, and there's only one. Right? And we saw that they bought US Internet, a very big name for a very small carrier in suburban Minneapolis, right?

Roger Entner 2m33s

But that's the slim pickings it becomes. So that was really cool. You know, the announcement that they are launching live translation on the network, I think is very telling, very transformative. If you can do AI translation on the phone, This is about the heaviest load, the most latency sensitive.

Don Kellogg 2m56s

Well, you're doing a translation in the network, though. It's not on the phone.

Roger Entner 3m0s

And in the network. Exactly. In the network, and you're switching live, like it was explained to me, like, it takes for example, if you're calling somebody in Portugal, it takes the hint from the area code, and then it start talking in Portuguese if you wanted to. And if the other person responds back in Spanish, it switches on Spanish on the fly. The other thing that they talked about in meetings afterwards, not publicly, was that the AI voice will feel and sound very similar, if not identical, to your voice.

Roger Entner 3m32s

You know, because they had this commercial where the granddaughter was talking to the grandmother in English, and and the grandmother was in speaking Spanish. First of all, shame on the mother. Sorry. But I was like, if this is like some kind of robotic voice, there's so much disassociation going on, you know, they might never know each other. I thought that was your voice, grandma, when they actually finally meet.

Roger Entner 3m54s

But no, it's gonna be like extremely similar, if not the same, of your actual voice. And that's really, really awesome. But here's the thing, if you can do live translation on a network, you can do anything. Right? And so here, T Mobile has the opportunity not to be just a quote unquote a dumb pipe.

Roger Entner 4m16s

Telecom is very is nicely profitable as a dumb pipe, but there have been so many opportunities where they could have taken part of the value chain and were completely bypassed. I remember in 2000 when was that? '7. I did effectiveness measurement for online advertising for Verizon. Nothing came out of it.

Roger Entner 4m38s

I talked with AT&T over the years about how to have a part of location based services. Carriers got zero, right? And there are so many other opportunities where the carriers have a legitimate right to be part of the value exchange. And shame of the industry if it now happens, I think we're way beyond three times, right? But shame on them if AI is the next thing where they're not part of the value exchange.

Don Kellogg 5m7s

One of the things that is important about AI is the latency really matters. Right? So the closer you can get to the customer, the deeper you can be in the network. Insane. Yeah.

Don Kellogg 5m17s

Right? The better the latency is, and that's really, really important for things like maybe not chatbots, but if you're doing autonomous driving or something like that Well latency matters. Right?

Roger Entner 5m27s

Well, you're not gonna do autonomous driving on a you said it yourself. The AI is on the car.

Don Kellogg 5m32s

Yeah. Computer vision. Computer vision. Put it that way.

Roger Entner 5m35s

Exactly. And by the way, as a part of, you know, what I'm gonna talk here on this panel here is that the network really matters and latency matters. And you have to give T Mobile kudos that they have five gs advanced with L4S, which is like the low loss latency, low that makes it as good as you can currently get. And so that's a really an important step forward. As that permeates through the network, you know, we're running 6,000 respondents, 30,000 respondents on the AI experience, 6,000 a week, 30,000 a month.

Roger Entner 6m10s

Right? It's an insane number. We're going to look very closely if the T Mobile AI experience on a mobile device is going to be different than that of AT&T and Verizon, who don't have that technology yet. We will see the same thing then for FWA, where you have like an L4S, FWA versus not. Does that make a difference?

Roger Entner 6m32s

I'm gonna look in the data, right? Do I have a hunch? Yeah. But that's me having a hunch, and logic would dictate, yes, it should make a difference. We look in the data.

Roger Entner 6m41s

Takes probably a couple months.

Don Kellogg 6m43s

It'll be interesting to see. I mean, a lot of this stuff, like, you know, some of the Live Translate stuff has been available on Android for a while. You know, we see Apple continues to slowly take share from Android. So I'm still skeptical in the end use case in terms of whether or not this sells devices or changes behaviors, but we'll see.

Roger Entner 7m1s

And by the way, what we're seeing in the data, people don't often do AI. By the way, we just put final touches on a research note. Risich, Mehwada and I collaborated on one, and Risich did really the heavy lift on that, where you see such a strong difference between paid and not paid, that the people who are not paying for AI don't really know how to use it, are frustrated by it, and a lot of them are giving up again, because the damn thing doesn't do what it's promised to do. Whereas the people who are spending money on this, getting deep on it, are getting outsized rewards and benefits from it. Some of the data will show, like 75% in rural America say AI doesn't make an impact either way on me.

Roger Entner 7m49s

They don't even realize that they're getting left behind. The proverbial car doesn't get faster when everybody else, you know, is driving now 100 miles an hour and they tug over along with 30. They just don't know any better. What I'm seeing at least right now from that paper is it becomes a demand problem, not a supply problem. And Rishis and I are gonna look at what impact technology and BIT has for that.

Roger Entner 8m12s

Should be fascinating as hell. Stay tuned. Don't know yet where we'll post it besides our webpage, but it's really, really cool stuff. But coming back to T Mobile, they still have the best growth story in the industry. Really, really solid.

Don Kellogg 8m26s

Right. Although I was a little disappointed that they're changing some of their reporting, where they're not gonna be reporting individual lines anymore.

Roger Entner 8m32s

Well, that's a deliberate move. Absolutely.

Don Kellogg 8m35s

It's a deliberate move, but I'm not I'm not sure it's it's investor friendly. Right? To be fair, they had a bunch of free lines in there anyway.

Roger Entner 8m42s

So Well, the thing is so it makes it now much harder to compare. What also happens is churn will go down, because churn on an account level is lower than on a line level, as we know. And I would call it very Roger friendly, or very reconanalytics friendly, because we have the analytics to be able to tell lines from accounts. And so it just makes our data and makes the people the people who work with us, and now it's almost everybody, it doesn't take much away from them. It's the investors and the investment analysts with whom we don't.

Roger Entner 9m18s

We also work with most of the investment analysts and we give them our data. So it doesn't really have that much of an impact, at least on the sell side. And, you know, we might get some customers on the buy side, you know, we've already sold things to hedge funds. So personally, I was very happy because it makes our data richer. But yeah, I mentioned that to Srini.

Roger Entner 9m38s

Said like, that's a deliberate move, you know, to make yourself less comparable. And they're like, yeah, because who is the leader in churn? AT&T. And AT&T is the one that doesn't report accounts. Because once you have accounts, you can do really nifty things with math, as you and I have shown, and gain insights that are, I think, beyond what they wanted.

Roger Entner 10m0s

Because we told T Mobile, you give us accounts and you give us lines, we can back out your free lines, and here it is. Since they're taking away now the line part, I guess our number was pretty close.

Don Kellogg 10m13s

Yes. We could do criminology on this. Right? I think in the spirit of disclosure, it would be nice if they still reported this stuff, but, like, we'll laugh all the way to the bank on it. Right?

Don Kellogg 10m21s

So that's fair.

Roger Entner 10m23s

Right. And then they are like, hey. Why can't they tell good stories around us? But anyway, well, we'll talk about it next week. Okay?

Don Kellogg 10m31s

Alright. Enjoy your panel. We'll talk next week.

Roger Entner 10m33s

Thank you. Bye bye.