AT&T, Inc. (T) Presents Barclays Global Technology, Media and Telecommunications Broker Conference Call - (Transcript)

|
About: AT&T Inc. (T)
by: SA Transcripts
Subscribers Only
Earning Call Audio

AT&T, Inc. (NYSE:T) Barclays Global Technology, Media and Telecommunications Broker Conference Call December 6, 2018 11:30 AM ET

Executives

Scott Mair - President, Operations

Analysts

Kannan Venkateshwar - Barclays Bank

Kannan Venkateshwar

All right. Good morning, everybody. Thanks for joining us today. My name is Kannan Venkateshwar. I cover the cable, media and telecoms space for Barclays. I'm very to happy to have with me here Scott Mair from AT&T, who's the President of Operations and is responsible for solutions design across the network, planning and engineering, integration of acquired assets and the entire CapEx budget.

Thanks, Scott, for being here. Before we start, I guess you have a safe harbor statement.

Scott Mair

Yes. Absolutely. As usual, we'll keep the lawyers happy here. So some of the comments that I'm going to be making will be forward-looking and have inherent risks and uncertainties associated with them, and so the results may differ over time. So if you need more information, please check out the -- our website or our SEC filings. Also, I will add we're in a spectrum auction silent period, so we won't be talking about spectrum today.

Question-and-Answer Session

Q - Kannan Venkateshwar

Describe -- I mean, when you think about the world right now, there's obviously a lot of changes from the network perspective. So could you talk a little bit about what your priorities have been, maybe, in the last couple of years and how that's morphed as we go deeper into FirstNet and 5G and so on and so forth? What -- how have your focus areas changed?

Scott Mair

Sure. So it's been a really busy year in 2018 for us. We have some large builds going on. I'd really put the priorities in the following list, not necessarily in priority order. But FirstNet is number one, right? We have some large commitments with FirstNet, and we can unpack all these as we go here. And so that FirstNet build has also enabled us to do additional builds. So we call our program our 5G Evolution program, how do we put more spectrum on sites and modernize our sites. So big build going on associated with that. And then that takes us into 5G, kind of a current topic right now. So a lot of time and energy on 5G. So that's certainly new. More time spent there in 2018. As we look at the other couple of things that I wake -- I stay awake at night thinking about, and that is our fiber build, right? We have our FCC commitment of 12.5 million homes passed that we're working on. And finally, we've been on this journey around virtualization. So how do we get our network to be more software-defined, more nimble and be able to introduce features and capabilities a lot quicker? So we're -- have a great program around that at AT&T. And so I mean, when you take those five, that's really where I spend my time. And you look at '17 and '18, couple of changes there, more FirstNet, more 5G. And '19 is going to be more of the same.

Kannan Venkateshwar

So looking at FirstNet, I mean, if you could start there, if you could just update us on the build-out so far and what the priorities there are in the coming years?

Scott Mair

Sure. So first, we're off to a great start, and it's -- we at AT&T are really, really proud to serve the first responder community, right, and that they do important work. When disasters happen, performance and connectivity really matters. And so we're building out the FirstNet network. Two deliverables this year. One is we have built a dedicated core network. So that's kind of the back end of the network. That deliverable was due at the end of March of this year, and we achieved that. And also then, putting the Band 14 spectrum to work. That was part of the FirstNet deal that we did with the FirstNet authority in U.S. government. That build is well underway. We had a target to achieve about 33% of our commitments for Band 14 deployment, and we hit that. In fact, we made that six months early, right? So we're really, really aggressive with our build right now.

We want to get the Band 14 spectrum out there to serve our first responders. And where there's Band -- where there's not Band 14, our first responders can also use our commercial network and get the same priority and preemption and services that they purchase from us on FirstNet. So that build is well underway. Really, really happy with it. In fact, I would say proof points matter. And during the last 2 or 3 hurricanes, couple of hurricanes, Florence and Michael, and then also the wildfires in northern and southern California here, our network performed really, really well. Very proud of it. In fact, we opened up our network for others as well. Because in times of crisis, connectivity matters, and we put all the other things aside. And so the FirstNet network and the response that we got from first responders has been outstanding, and so we're off and running. And next year, we'll be building out a lot more Band 14.

Kannan Venkateshwar

So when you think about the use case for FirstNet, is this primarily voice? Or are there other applications? And when you think about the milestones in terms of the rollout, how should we think about how these applications roll out in a couple of years?

Scott Mair

Yes. So as part of our FirstNet product, we have a whole suite of products that will roll out over time. Some are available now. Some will -- are being developed, and others will come over time. First responder community has unique and specific needs, and so our commitment has been to support them in any way they need. We're building an ecosystem of developers and people that can build solutions for FirstNet and make them available. And so when you look at it, it's unique to that customer base. So that's a very, very large customer base. So those product sets, we're selling today. In fact with FirstNet, so far, we've got 3,600 agencies who've already signed onto FirstNet and over 250,000 lines. Primarily, those are voice lines at this point. IoT will come because there's going to be a lot of IoT applications. But right now, it's primarily voice that we're seeing, and we're getting a lot of feedback. And one thing about the first responder community is the relationship we have both with the FirstNet authority as well as the first responder community itself is really, really robust, and it's really good because it's a partnership. And the feedback we get on things that they need really helps us make sure that we're spending time on the right product sets at the right time for them.

Kannan Venkateshwar

Okay. And then I guess, from sales process perspective and all the agencies that you sell into and so on, is there a set of milestones we should be looking at? And how long does this process take in terms of conversions from the point you start it?

Scott Mair

Yes, so the sales process is such that we sign the state agreements. And then underneath that, we have to go work with all the agencies and sell FirstNet products and services to them. And so that -- there is no set time line. Our sales cycle could be as quick as 2 or 3 weeks, as we saw happen after the hurricanes. Performance always brings an opportunity to sign more people up, and we saw some of that from some of the recent disasters. Other times, it's 6, 7, 8, 9, 10 months. It could be a year. And it's -- the agencies are starting from one place, and they have to plot out their path from where they are and the services and technology they use to a FirstNet solution set. So good success, and we have a full-time dedicated sales force working that. The network build and the capability and performance really matters, and so that sells itself.

Kannan Venkateshwar

And then when you think about the impact of network quality as you go through this FirstNet build-out, as you deploy more and more spectrum, how is that playing out? And what should we expect in the coming years in terms of network quality improvement?

Scott Mair

Sure. So the beautiful thing around FirstNet is it's really a reason to go touch the towers. And so as we're touching the towers, we're implementing what we call our 5G Evolution program. And what that means is when we touch a tower, we're actually upgrading that tower. We're putting a lot of spectrum to work. In fact, between end year '16 through end year '19, we'll increase the spectrum that's deployed by 50%. So we're putting a lot of spectrum up, and that really enhances the quality of the network, both speeds and reliability, and we're seeing that. We're seeing that not just with our internal metrics, but our external metrics that we're getting from third parties. So the performance of the network, so far, is really improving. And when you look at external GWS, the Best Network award that we received in September is really a milestone for us that tells us that we're on the right track, that our network is outstanding today. And it's going to get better as we continue on the program. And that's -- GWS, they drive over 500 markets across the U.S. and over 1 million miles. So it's a really, really robust process that they use, and they look at what is important to the customers.

So that's a good indicator. The capability we're putting up on the tower is really one that leverages LTE advanced capabilities. And what that means is carrier aggregation, improved modulation techniques and 4x4 MIMO. All yield a much better experience for our customers. In fact, those towers will support 400 megabit-per-second peak speeds. And I always like to say, kind of rule of thumb, 10% to 20% of that is what an average customer will see, so 40 to 80 megabits per second is the experience. And that -- right now, we'll have that complete in 400 markets by the end of this year, over 175 million people. And we'll be nationwide next year, by the middle of the year in 500-plus markets. So that FirstNet trigger then allows us to go touch the towers and put those capabilities up and really, really improve the experience for our customer base.

Kannan Venkateshwar

That's great. And on the 5G front, if you could just update us on the process there and what the time line there looks like in terms of deployment.

Scott Mair

So when you look at 5G Evolution, going into 5G, right, so faster speeds and capabilities. While we're out there building, we're also deploying radios that are 5G upgradable with software on the towers. So we've been doing that through much of '18. It's an entrée into the next generation of technology. So I'll start a little bit with standards, right, and kind of -- because that leads us to why we're here right now at this point in time we are. So back in 2017, 3GPP accelerated standards for hardware. And that enabled all of the hardware manufacturers to do product design, fab work and such and get a head start because that's always the long pull on development. Middle of this year in June of '18, 3GPP released the final standards for that first iteration of 5G. And that enabled things to get locked down, software developers, all the OEMs to begin their work.

And so between end of June to what will now be end of this year, right, December this year, we're finally going to see commercial. We've said we'd be first in the market with a mobility 5G solution in small parts of 12 cities, going to 19 cities in the first half of '19. That commercial interval of six months, to me, is absolutely amazing, from standards being built and finalized six months later. In an infrastructure business, having -- getting that out in six months is amazing. So our product that we're going to be using -- that device ecosystem is pretty important. So the product we're using is a MiFi product. It's the Netgear Nighthawk 5G mobile hotspot. That will be an entry device. And the device ecosystem is really what drives adoption, obviously. You got to have something to use it. And the way the time lines are shaping up, there's news this week, right, on what we'll see in the first half '19. So we'll begin to see some 5G-capable devices in the first half of '19. And I think in the second half of '19 and early first half 2020, we start to see that ramp. We start to see the scale devices come into play. We see scaling of networks in '19 and '20. And so what we're starting now is important. The ecosystem is developing very nicely, and it's coming together. And so you'll see '19 ramp into 2020 in terms of 5G rollout.

Kannan Venkateshwar

And when you think about the applications for 5G, I guess there's been some debate about whether it's fixed or mobility, which leads the way. But if you can talk about how you are thinking about the big use cases and how you see that evolving over the coming years.

Scott Mair

Sure. So from a -- from our standpoint, I mean, fixed wireless, the technology works. It's not a question about technology, in my mind. We've done fixed wireless in our network on LTE as part of our Connect America Fund 2 commitment from the government. We've been doing that for two years. And so we know the technology. We know it works, and it works for the purposes intended, which is real broadband. 5G, the same, right? We've been doing testing for 2 or 3 years, as have others. In fact, we've had fixed wireless up and working in Waco, Texas for more than a year on 5G. Not standards-based at that point because the standards weren't complete. But we know the technology. The challenge is the use case and the economics, right? So where does fixed wireless work? We're focused on mobility. And the reason we're focused on mobility is if the fixed wireless use case technology works, it's the economics of where it does it fit, and it fits in certain places. Dense urban areas, near-in suburban, I think there's application there. Difficult to build and you can bypass the last 200, 300 meters, 400 meters for that. But it's -- it doesn't work everywhere.

So from an economic standpoint, if you're going to be there, if you're building small cells, which is what the fixed wireless 5G case is, those small cells can do mobility or they can do fixed. And in fact, they can do both at the same time. So opportunistic? Absolutely. If we're there, we build small cells primarily for capacity. It's a very, very effective tool for capacity. That capacity could be allocated towards fixed or to mobile. And again, we believe the mobility use case is probably the right place to be spending our time and effort, but it doesn't mean that there won't be places where fixed wireless is the right tool and the toolkit at that time. So it's not a technology challenge. It's kind of a deployment challenge and then also the economics around it. Other use cases, I think the use cases for 5G are just now developing. There's going to be many, many more to come.

We see initially enterprise businesses as being the area where the entry will be first. We've thought about partnering with a manufacturing firm, and I really believe that manufacturing is going to be a key capability. When you look at a factory floor, it's real-time telemetry, real-time analytics. You have factories that now need to be more nimble than ever in terms of being able to reconfigure for product changes very quickly. The cycle times are shortening. 5G enables real low-latency connectivity to all of the metrics that matter and data that matters. What's my supply chain look like? What's my stocking and logistics look like for that robot or that machine? When do I make -- need to make sure I have the stock there so that I maximize the throughput? Visual inspection, defect identification. The earlier in the process I can identify the defect, the more productive my line is. So that's one space. Certainly, other places in terms of augmented reality, virtual reality, low-latency applications, IoT. I think there's going to be plenty of IoT use cases where low latency matters, both in businesses as well as in consumer, right? Gaming is clearly an opportunity. Sports and entertainment is an opportunity both in the business side. Augmented reality for training, right, and we're beginning to see that already today. Very, very effective training tools. So I always like to say just like with a CPU right, the faster the CPU, the more capabilities that it has, applications and use cases get developed to take care of and makes use of the capabilities offered. And we're going to see the same thing in 5G in a pretty big way.

Kannan Venkateshwar

And the other big focus area for you, of course, is the broadband, the fiber broadband build-out. What's the current status of that build-out? And what are the milestones we should be looking for over the next, maybe, year or so?

Scott Mair

So with our fiber build-out, fiber underlies everything we do, whether it's wireline or wireless. And so fiber matters. And so with our FCC commitment from the DIRECTV merger of 12.5 million living units, that's mid-'19, we need to be completed with that. And we'll hit that. We'll absolutely achieve that. We're 10 million complete through the end of the third quarter. Machine is up and running, and we're doing really well at building that capability. So we'll be at -- in fact, by middle of next year, we'll be at 14 million homes passed. And because we also have a deep fiber footprint, we'll have another 8 million businesses that we pass. And those 8 million -- so that gives us 22 million locations that we can sell to the building, to the premise, fiber-based services. So it's a great inventory for us. And we'll continue to evolve both our in-region and out-of-region footprint for fiber. We have -- people think of us sometimes as just the in-region franchise footprint. That's true. Not just the fiber we're building now, but we have fiber deep into our network because of our IP broadband and IPTV services. So that's already fairly deep there. But out of region, we have fiber in many, many, many cities as part of our network. And so we continue to invest in both in-region and out-of-region fiber.

Kannan Venkateshwar

As you go through this planning process of 5G and you're doing the fiber deployment, how has the planning process changed overall, given that you have multiple objectives with fiber today versus what it used to be maybe a few years ago?

Scott Mair

Yes, it's changed a lot. And it's really brought about by the availability of databases and data for buildings, locations, where are sales opportunities and such. And so the tools of yesteryear for a fiber engineer were to bring as much data -- gather as much data as they could. Because we've always done integrated planning, but it hasn't been as robust and as much automation and tooling as we see today in AT&T. So a fiber engineer of yesterday would say, "I have a customer order for a fiber-based service to a business that I've got to put in 2,000 feet of fiber." And so they build from point A to point B down main street, and they provided that fiber service. Today, when I get that order, a fiber engineer, we'll take a look at, okay, what's the ROI for this job when I optimize it with the automation? And the databases and tooling that we have, the fiber engineer literally can push a button and say, the best solution for this is I can route it down this street, up that street, around the corner. I can make my way to that business at the end. I could still provide my fiber, but now I'm passing 500 additional MDU units. I'm passing another 20 businesses that I can sell into.

I'm passing two cell sites that I'm paying someone else transport and backhaul for, where I can now put it on my own network. I know where I'm going to be building small cells in the future. We can plan out that. We know where we're going to be. I can route that fiber. So now I've optimized the route. I've optimized the sizing of the fiber. And my ROI goes from what would be a good ROI to a great ROI because of the tooling and automation and integrated planning that we're capable of doing now more than we've ever been able to do. And that is a key difference, right, because otherwise I would be missing opportunities. And so that tooling makes all the difference in the world, better solutions. In every one of those jobs I do, I'm that much closer to the next opportunity, the next business that I want to sell into. Where I might have been 1,500 feet away before, now I'm 500 feet. The incremental costs with incremental marginal time that I need to spend to get to that next business, it's a beautiful thing if you like to engineer networks.

Kannan Venkateshwar

Another focus area that you mentioned earlier was, I think, software-defined networking, and that continues to be a focus across the entire industry. If you could you help us understand what's being done, so far, on that front and what the objectives are. What are you solving for that's out there?

Scott Mair

Yes. So we've been on this path for the last 3 or 4 years. It's really nice to see the industry really adopting software-defined networking. We've made a lot of progress. We said we'll have 75% of our network completed or virtualized by the end of 2020, and we'll make that. We're 55% end of last year. We're making good progress this year. Two things. We started at the customer edge, and we're finishing -- we're focused on the core. So those were where we have been. The next great opportunity is the radio access network space. The evolution of 5G networks are really edge compute-based. They're also software-defined-based. And so the next 2 or 3 years, we're going to see road maps that then also virtualize one of the areas that was later in the road map, in the wireless access space. So really good progress. We're seeing benefits both in CapEx efficiency as well as OpEx efficiency. In fact, when you look at OpEx in AT&T in the operations and technical space, if you will, we've seen 15 consecutive quarters of year-over-year operational expense reduction. It's really nice to see us being more efficient. And again, the tooling, the automation, the virtualization gives us the opportunity to do that.

Kannan Venkateshwar

I guess we have just 10 more seconds. I don't know if there any questions in the audience here. All right. I think we are out of time.

Scott Mair

Excellent.

Kannan Venkateshwar

Thank you.

Scott Mair

Thank you very much. I appreciate the time.

Kannan Venkateshwar

Thank you. Bye.