Welcome, everyone. My name is Alex Waters. I'm a part of the Communications Infrastructure and Telecom team here at BofA. Very thrilled to have Dan Peyovich, CEO of Dycom, at our conference today. Dan, welcome.
Good morning.
Any safe harbors you need to make before we begin?
Yeah, thanks, Ralph.
I see Callie in the room, so just making sure.
Keeping us on the call. Appreciate it. For any forward-looking statements I may make today, just make sure you consult our website for the safe harbor statement.
Perfect. Well, we'll get started then.
Yeah.
Yeah. Maybe just before we begin, for the folks in the audience and the people online, could you just maybe provide a little overview of Dycom and kind of what you guys do and the verticals that you operate in?
Absolutely, so we're a premier digital infrastructure services provider. We're in all 50 states across the U.S., doing wireline, wireless telecommunications construction. Our customers are the large carriers, you know, the AT&Ts, the Lumens, the Verizons, the cable customers of the world, and then we do work related to hyperscale or long-haul builds as well, so everything that's going on with the AI evolution, and a huge part of our work is service and maintenance, so that perpetual work that happens year over year, and as everybody knows, very active space, a lot of homes to get passed in the coming years, so we're excited and excited to be here to talk about it.
Yes. Very active space. So maybe just starting off where you ended there, kind of the bread and butter of Dycom is fiber to the home builds. Could you just start out and give us kind of the current state of what you're seeing? Obviously, your customers are increasing builds, a lot of folks in this kind of environment. So what's kind of the current state?
It's been fantastic work, Alex. I believe a lot of people know these builds have been going on now for several years. I think probably what we're trying to educate people on is that they're not as far along as people believe. Our customers continue to increase the homes that they expect to pass. Collectively, they've added 50 million incremental passings over the last 16 months. That's a very big number when you think about 140-ish million homes in America. And ultimately, we think that about 80% of the homes are going to get passed with that private capital. So if you're at about 75-80 million today, you add another 50 million on, you're getting to something like 125 million homes. And that's going to take a huge amount of labor forces, a large amount of equipment, and a ton of planning.
That's work we've been doing and doing well for the last few years in earnest. And again, you've heard our customers talk about it. It's a unique time when pretty much every single one of our customers is either affirming already high-speed run rates, affirming their commitments to pass a huge number of homes in their footprints, and then also incrementally adding that. And they're adding it in two ways. One, through their own consolidation, AT&T and Lumen being probably one of the most recent. So you're seeing all that stacked together. So you have this huge momentum behind it. And what that means is that those numbers continue to increase. And the number of homes expected to be passed every year continue to increase. I think industry estimates would tell you that about 10 million homes got passed last year.
So if you think about that developing over the next five-plus years, huge opportunity for us. One, that we're extremely well-positioned across customers. I think, again, people think about us connected to the large carrier customers. We have hundreds of customers we're working for every day. It could be a very small local co-op all the way up to the AT&Ts of the world. So huge opportunity. And I think probably the message I would like people to hear is that although we've done a ton of work there, there's still a lot of growth potential and opportunity.
When you think about the timeline, right? I mean, AT&T, their builds till kind of 2030. I mean, how do you think of, I guess, from here to there and then perhaps after that, the chance of overbuild, et cetera?
Yeah. So some of our customers have been very specific about their timelines. And absolutely, there will be builds that get done by 2030 for sure. When you think about the 80% or getting to 125 million passings on the private capital, we believe that still a number of those are going to carry well through 2030. And that's for a few reasons. One is not all the builds across all the customers are at the same pace. Two, these are highly complex projects. They're not, I know it seems, it can seem very simple, either dig a trench, put some pipe in it, pull glass through it, climb poles, string glass between them, but they're actually highly complex to do. You've got a huge amount of labor forces that are required behind that. You've got permitting, locating constraints that we continue to see in the space.
When you take that across everything and then you say, not only do we have all those fiber to the home builds to go do, which are only increasing and expected to increase, you've also got all the hyperscale or long-haul middle-mile work. That's really just getting started. So you have to add all of that on top. BEAD is finally starting to get some color and context behind it. So you have to add that on top. And we just think that there's a ton of opportunity coming into the space. So naturally, as part of that, some of the fiber to the home is going to go past 2030 overall.
I want to get to data centers and BEAD and the wireless side of the business, but maybe can you talk about the maintenance portion of your business? I know you guys came out a couple of quarters ago and kind of quantified it for us investors. I mean, just where do you see that business growing? I mean, I know you've talked about this in the past of every new home passed is a maintenance opportunity. Could you just dive into that a little bit?
You took my tagline there. That's exactly right. First, it's a priority in our business. Service and maintenance has and continues to be a priority in our business. I do think that makes us unique. There's plenty of competitors out there that their preference would be just to do the project work. We really want to lean into the long-term relationships. The service and maintenance work is very difficult. It's very complex. You have to have people at the ready, even on Labor Day weekend, to go address any issue that comes up anywhere in the geographies that you cover, which for us, again, being across all 50 states, is a large part of America, both rural, suburban, and urban. So it's a very difficult thing to set up. There's a ton of fleet, equipment, and personnel that you have to have at the ready.
It's been a focus in our business for a very long time. And for as long as we've gone back and looked at the numbers, it's always been more than half of our business. And we continue to make it a priority to ensure that it stays over half of our business. Now, that number can modulate, but the important point is that it underpins the rest of the work. What it does for us, other than being able to have these deep relationships with our customers where we're servicing them on a daily basis, and then when the builds come up, it positions us well from a relationship standpoint. It also positions us from a people and capital equipment.
So all of that is at the ready, and we can leverage that, whether it's into fiber to the home builds or whether it's into long-haul, middle-mile routes, or whether it's into what's going to come from BEAD, because we are in many of these rural locations too. I think when you take all of that together, it just sets us up well for the rest of the demand drivers, but it also just keeps this bedrock foundation that allows us to have that recurring revenue over time.
And within the maintenance portion, I believe, is the locating business, which.
Is a part of it, correct?
Which is a large kind of recurring business for Dycom. Could you maybe just unpack what that locating business is and potentially quantify it of how much of that 50% or greater than 50% it is? Because there is a recurring nature to that, right?
Yeah. And for most of our locating business does end up in the service and maintenance category. So the maintenance business for us is typically around 6%-7% of revenue. That is a business that we've been able to grow. And I think it's a lot of the same messaging that we have on the wireline and wireless side. What Dycom brings is, other than our footprint, is we're really leaning into the customers to provide a level of certainty and really raise the bar in the industry that we think differentiates us. Our customers have shown that they want to consolidate vendors. They want to have vendors that prioritize that certainty of delivery, that certainty of execution, and prioritize the long-term relationships over short-term. And so they've been continuing to consolidate vendors.
And they've also, as we've talked about on some of the earnings calls, we've been awarded new markets, markets that we weren't in, but somebody else was in and was struggling to perform. That all plays right into the utility locate business as well. Many of those are the same customers. There's additional customers outside of our core customers too, but they're all looking for a similar recipe, right? They want that certainty. They want the relationships. And so we've been able to grow that business and see continued growth opportunities ahead of us.
The scale of that business, I mean, obviously, Dycom operates in all 50 states. Is that a subset of the 50 states of the locating business?
It's across many states. Obviously, it doesn't cover our whole footprint, but we do cover a lot of ground.
Okay. All right. Maybe switching to the data center portion of the business. I mean, obviously, we cover the data centers here. It's a very big growing aspect of the industry and big for you guys as well. I mean, you guys quantified it with 2Q earnings and kind of gave a $20 billion TAM over the next five years. I mean, maybe a couple of questions in there. Could you maybe give us the building blocks of that TAM or what kind of services those are? And then perhaps why only the five-year outlook for this business?
So we wanted to be really clear about the $20 billion TAM. There's been a lot of conversation related to what is the potential impact, whether it's related to AI, the hyperscalers, or enterprises continuing to move to the cloud. We have a very unique insight there. We've been spending time with the hyperscalers over the last several years. Obviously, we've been spending time with our carrier customers who are doing a lot of this work for the hyperscalers. And across all these conversations, we've gotten a really good look at what needs to get built across the U.S., what needs to get built. And I want to talk a little bit about timing and how connected and disconnected it is from the data center expansion.
But we have insight to a lot of those routes, whether they're replacing existing routes, upping the capacity, lowering the latency, or building completely new routes. Sometimes those are for redundancy. Sometimes those are just new routes going to new data center locations. So we have all this information that we could look at, and we're able to size and scale what that total opportunity set is. We also included the inside the fence work we've been talking about. So that's taking fiber from the right of way, pulling it into a data center campus, and then interconnecting the data centers. Smaller portion of the $20 billion, but it's included in there. There is nothing in that $20 billion that's outside of the service lines that Dycom executes today. So we're not including supply of the fiber because that's not typically what we do.
We're not including anything on the power side as well. So that purely is addressable into the work that we do today. I just want to talk a little bit about why it's connected and disconnected from everything that's going on with the data center and AI space. What's unique about this long-haul, middle-mile work is it is highly time constrained. So I talked about complexity of the fiber to the home build. The complexity goes up many, many notches if you think about long-haul and middle-mile networks. Why that is, is because you're going through metropolitan areas. You might be going through downtown Chicago. You might be going under the Mississippi River. But you're ultimately connecting a route over a long period that's got to be connected end to end.
So you have to balance across all of those different municipalities, all of those different terrains to have continuous workflow, continuous operation, ultimately continuous connection, ultimately for hyperscaler, but certainly for our carrier customers who we typically contract with. To orchestrate all that across different permitting environments and all those variables is highly complex. Physically doing the work, because we're talking about very dense bundles of cable, very complex. The traffic control, everything around it is highly complex. What that means is, ultimately, for the hyperscalers, this is a long period of time to go spend all of those dollars. It's a long period of time to get ultimately what they want to build or what they see that they need to build today to get all of that done. So this is a time-constrained effort.
Though the revenue dollars and potential are big for our space from a CapEx perspective for them against $400 billion and rising CapEx related to AI data center spend, it's very small. So the time is really what is creating the need and the urgency, so yes, will there be more when AI continues to develop and there's more data centers? We believe so. Will there be more as power constraints get cleared out and they can go build in new areas or continue to build in existing areas? Absolutely. But in the meantime, there's just a ton of work that has to get done to really get the capacity up and the latency down and the redundancy in for what they purely need today, so you have all of that coming together, and that's why we say that, one, we think this is a kind of once-in-a-generation build.
Two, it's going to go well into the next decade. We see this as being 10-plus years. And then to your question, Alex, the reason that we said five years and painted a really detailed picture around that is because we wanted to be specific about the addressable market that we see today. We have really good clarity and insight to what's going to happen in the next five years. Beyond that, we would have to start making a lot of assumptions. Now, we believe that after 2030, the numbers are only going to go up. The need is only going to grow. All of our data consumption only goes up year over year. We don't see that ending anytime soon. So all science points to yes, so to speak, after 2030.
But because we had such specific data, we wanted to really be clear about what we think that is over the next five years. And let me just finish with, we do think that's back half loaded.
Okay, so back half 2028, 2029, 2030 is the way to think about it?
Yeah. I think it'll start. We've been doing overpull work with Lumen, longtime partner, fantastic opportunity. We've been doing that work out in the field for seven or eight months. It's just now kind of getting up on plane. These projects just take a long time to get going. So we see 2027 is really being when things can start from a revenue opportunity being more significant and then absolutely 2028, 2029.
Okay. And then are there—and you spoke a little bit between the differences between the overpull work you're doing with Lumen or the existing routes versus the new builds. Can you maybe just talk about those two differences between the traditional fiber to the home a little bit more in terms of how that work is? And then what specific geographies are you kind of operating in for this data center, for these data center builds?
So I'll answer the last question first. Because we're across the entirety of the United States, we know both the subsurface conditions. We know how to work in these municipalities. For us, where we do the long-haul, middle-mile work can be anywhere in the U.S. And we have work that's across multiple states and multiple geographies. So there isn't a limitation there. And certainly, we have an appetite to be across it all. Related to the fiber to the home work, the comparison I use is fiber to the home is a little bit of a shotgun approach where a customer is looking at a certain number of passings. And you can go across multiple neighborhoods, multiple municipalities to add up and aggregate those passings. Very different than long-haul or middle-mile where you're just connecting point A to point B or point A to B to C.
So over here, you have more variability. So even though it's highly complex, you can kind of go with a broader approach to make sure you're going to hit the numbers. Here, you just got to be completely dialed in. You got to be exactly precise to make sure that you can deliver and execute on that route. The actual equipment and people that it takes are similar, just higher level of complexity and a much higher level of coordination, project, and program management. And again, I think opportunities where we can differentiate. Permitting is much more complex because, again, in some of these situations, typically in a neighborhood, permitting continues to be a challenge and issues, and we can talk about that if you'd like, but even more so if you're talking about being in the downtown.
And you brought up the inside the fence work. That's a newer, quote-unquote, opportunity for Dycom. I mean, maybe why weren't you completing this work beforehand, and what's the customer kind of take been since you guys started offering it?
It's a natural evolution for us. We've been talking to the hyperscalers for years as part of that conversation, as part of doing some of these long-haul routes. It naturally led to what's happening once it gets from the right of way into your data center campus. What's happening for connecting it now that you need a larger fiber count that you're getting out to the right of way itself and to that vault? What are you going to do to take that into the campus? What we're finding with the hyperscalers is very similar with our carrier customers. They're looking across all of their different campuses, all their data centers, and they're saying, "We have a number of vendors, still highly regionalized, localized, and who does the work?" Variability in level of execution, level of certainty they're getting, and so our conversation started from the same place.
Again, where we think we differentiate is, what if we talk about a high level of certainty, a high level of accountability, and we only have to learn it once, and we can translate that across multiple geographies? That's really where the conversation started. We're very pleased that we've been awarded different campuses in different states where we can go do that. That's great work because there's work to do today, but also as they continue to build in data centers, as they continue to upgrade the networks, they're staying power. Then we did talk about a separate award that we received this quarter, which is just pure maintenance of their own facilities. I do want to be specific. Excuse me. That is different work than what we do for our carrier customers. It's not work that competes with the carriers.
This is very specific for the hyperscalers.
Okay. And just in terms of the timelines, I mean, obviously, the long-haul builds for data centers, they take time to complete. I mean, for the inside the fence work, is that mostly completed, I'm guessing, once the data center is almost complete or completely built?
Yeah. So there are some where today they just want to upgrade the network. So the data centers are there. The campus is there. Just go. There are others where they're continuing to add to the campus. So as things come online, we'll be there for a long period of time, potentially, Alex, continuing as they bring new data center in, creating that connection. As they bring higher density, higher count fiber in that's not there yet from long-haul, middle-mile, then we'll come in and bring that out into the campus. So the answer is it kind of depends, but it's a little bit of both.
Okay. Maybe just switching to BEAD. I know we've been talking about BEAD for a couple of years now, but.
Five years.
Yeah, exactly. For a long time. I mean, can we just talk about where we are currently, what your controversies have been like with the states? And I'm assuming some customers are getting plans ready. Programs seem to be much more fiber than maybe people feared heading into the new administration. So I'd love to get your take on that.
We're excited. First, it's not included in any of the outlook that we've given or any of the future talk that we've had about the business. BEAD is not in there at all. At the same time, we're finally getting to a point in the program where everybody's getting more clarity. For a long period of time, we said that we thought two-thirds, maybe 70%, would be fiber en masse across. Tons of moving parts that have happened over the last four years around that. What we're seeing today as the states are starting to come out and talk about the awards, that two-thirds of that has been fiber, has been either fiber, or I should say there is some HFC in there as well now. So wireline side with the cable customers getting to bring their plant in as well. That's all a huge positive.
That's all a huge amount of opportunity. That still has to get fully developed. I think it's important to remember, and we've been talking to the states for a long time. We've been talking to our customers, the sub-grantees for a long time. The states previewed their plans with the NTIA. They didn't just drop them on the doorstep and run. So we believe that getting through these next paces, there's also obviously the new administration wants to get fiber built, right? Wants to get capital deployed. So we think there's a lot of momentum that's going to carry it in. And we do think that we could see revenue opportunities as early as second quarter of next year. We're already in conversations with our customers. Some have even come out and talked about the awards.
Probably a big point that happened in all of the change as they released the new NOFO notice of funding that came out with the new program requirements, there was a number of changes that happened. Labor workforce requirements, how the states can spend the money, that now it's got all the deployment related. But a big thing that happened there was they made it a lot more appetizing for our larger carrier customers. What we've seen so far in the states that have announced, which is about half of the states, is highly weighted towards our larger carrier customers. The AT&T, Comcast, and Brightspeed are the top three by total awards for fiber today across the states that have announced. Of the 10 top customers that have received the most fiber or wireline, eight of those 10 are customers that we're working for every day today.
So it's really favored moving towards customers that are larger scale, certainly favors us. We do work with hundreds of customers, so we have relationships throughout. We do think that there's going to be a lot of opportunity there. So yes, we think that we're finally at a place that it's getting to be a little more understanding about how the program's going to go. We're already talking to customers about prospective builds. Another point that I would make for us with existing contracts with many of these customers, the opportunity is that we don't even have to have a new contract, that we can really be just moving from one area, driving however many miles to an adjacent area to start once they do get the funding in place. And that's why we think that 2Q of next year is an opportunity.
Okay. So 2Q next year, we begin to get some funding. I mean, do you think the lion's share is probably in fiscal 2028 for you guys or fiscal 2029?
Yeah. I think it's going to be one thing that has happened. It's a little bit of a Kentucky Derby now. So you have all the states where before they were going to be more modulated in how they released. Now everybody's going to be lined up at the start when the year turns over at pretty much the same time, other than maybe Texas who got an extension. It's going to take some time for the supply chain to work through that. It's going to take some time for all these to get ramped up. It's going to get time for people to work through permitting and how that's going to work. So yes, I think you're going to see a lot of ramping in next year. I think you'll even see ramping, but good activity in 2027.
But again, there's certainly going to continue to build over the lifecycle.
And then just thinking about the $42 billion, how much is a TAM for Dycom? Is it for the labor side, right? Is it half of it, would you say? Or what are your kind of thoughts on that?
So there's the $42 billion. There is the match that has to come from the sub-grantees. So if you can call that round number $50 billion-ish. Again, a few things have happened over the last four or five years as we've worked through BEAD. One, homes have continued to get passed out in rural America through other programs, right, through cooperatives, local cooperatives building out their plant. So the number of actual passings has gone down. So I think that's going to bring the number down some. The other thing is that there's the other programs, the non-deployment programs that were part of the $50 billion. Those can't be done anymore. At the same time, there's obviously a big cost focus the way that BEAD is set up now. So the number's going to certainly come down from the $50. Where does it end up?
I think we're going to learn that. Hopefully, in the next few months, I think we'll have much better insight. It's still going to be a significant opportunity. Is it $20 billion? Is it $30 billion? I don't know yet, Alex, but we do think it's going to be a material number to spend. And again, remember, that's going to be in a relatively short four or five-year timeframe.
So we've kind of talked about the fiber to the home side, data center opportunity, BEAD. We'll get to wireless. But just in terms of kind of your employee count, right? And I mean, do you feel like you're well-positioned to capture all these opportunities going forward? Or how are you thinking about it on the labor side?
We believe one of the things that differentiates Dycom is our approach to labor. We've got a huge workforce. It's highly dispersed around the country. We have crews sometimes that are one person, and we have crews that are maybe a dozen or 15 people, but our average crew size is a very small number of folks. How you set up to attract, retain, train, make safe that workforce is highly, highly complex, and it's something we've been doing and working on for a long time, so we do believe that we can differentiate in our ability to both attract and train and retain labor and have them deliver at a very high level. I talked about the level of certainty that our customers have come to expect from Dycom, so we do think that sets us apart. We've shown in prior quarters, we've had organic growth of over 20%.
So we've shown our ability to grow our workforce considerably over those periods of time. All that said is we talked about all the different drivers coming in, talked about the fiber to the home builds continuing to ramp. We talked about BEAD. We talked about the hyperscalers. That's something you have to be incredibly proactive about. You got to have a core strategy and discipline about your labor proposition. And again, this is where we think it makes Dycom different. One, if you look at our leadership, and I'm talking across the organization, myself included, the majority of our leaders started in the field.
If you think about attracting new talent to the organization, new talent to start at the very entry-level position, we think that that ability that people can see a career path, and we spend a lot of time talking about it and training through it, that you can quite literally get promoted all the way up through the CEO of Dycom is a pretty unique and powerful message. Two, and I will tell you, I'm a huge advocate for the trades, huge advocate for getting folks into the workforce. As an aside, huge advocate for getting metal shop and wood shop and auto shop back into high schools. All of those things, I think, can help to feed the future workforce. But we do see that we get a lot of feedback from the new generation that they don't want to be sitting at a desk.
And so the question is, how do you take that and operationalize it? How do we craft our story, craft our message, craft our opportunities so that we can make sure that we're bringing the right people in that have staying power and that can deliver? And I'll give you an interesting anecdote. We have found that people that play a lot of video games can excel at fiber splicing. So how do you take those skill sets that people have, and how do you have that conversation the right way with somebody so that they understand what they're walking into? So all that to kind of wrap around, Alex, we feel like we're very well-positioned. We've shown our ability to grow there. Very proud of our workforce. Right now, we're spending a ton of time with our training programs. We're building new training facilities.
And then within our numbers and what folks don't see, it's about the leadership. It's about the training. So the levels of folks that are managing people, we're spending a ton of time right now continuing to develop those, which allow us to grow the folks that are out working with their tools every day as the work comes through because there's going to be a lot of growth required.
My seventh-grade wood shop teacher will be very happy with what you just had to say. I mean, maybe just on the equipment side quickly. I mean, how do you feel? We talked about the employees for the equipment. I know you guys don't own the fiber that you're deploying, but the large chassis trucks, the diggers, everything like that, how do you view kind of what you have right now versus what you need?
The same way that we talk about our customers. We want to be long-term partners to our customers. We want to lean into their organizations and understand their needs. And we operate the same way with our supply chain partners. So our equipment suppliers, we consider to be very close partners. In fact, in a lot of cases, we're working with them and their R&D teams on how they can improve the equipment to make it more effective, more safe, more productive out in the field. And all of that lines us up to be ahead of the curve. I can tell you, even when times were very tough coming out of COVID and getting equipment, we were able to stay ahead of it. Our commitment and our goal with our customers is that we don't ever want our labor or our equipment to hold up their builds.
And so we spend a ton of time working to stay ahead of that. We do buy a lot of our equipment, which gives us resiliency across when pressures do come up. And it gives us a different level of relationship when we need to rapidly accelerate or buy more equipment. All that to say, we feel like we're in a really good position. The supply chain's in a really good position today to be able to meet the needs of the growth ahead.
Okay. And then maybe just on the margin side of the business, 2Q margins were very strong. They've continued to kind of increase over the past, gosh, year or two. Can you just talk about perhaps the cost management efforts that you guys are doing? I mean, even the revenue capture, as well as where should we kind of expect margins going here in the near to medium term?
When I moved into this seat, we made a few changes and a few statements. So one, we gave a revenue outlook for the year because we wanted investors in the street to understand what kind of opportunity set we had. And also understand that within quarters, sometimes there can be parts and pieces that move, but they don't always have anything to do with what the longer-term drive and need is. So we said, "Here's a revenue outlook." We also said, "We have opportunities to improve margin, and we have opportunities to improve cash flow, and we're going to focus on those. We're going to invest in those and create a core strategy around improving them." You saw the improvement on both of those. You've seen those develop over the year, but you certainly saw them come through in Q2. Very proud of what our teams have done.
Specific to margins, some of that does come through as operating leverage. And as I've talked about on the calls, strategically, we will reinvest operating leverage sometimes to continue to build the business. Other times, it's going to drop through the bottom line. So that will continue to happen as we grow. The secondary and important part is I talked about our dispersed crews. I talked about our crew size. We continue to see opportunities out there about how we can get better. And we can get better in a number of ways, right? You can always improve safety. That is the first and foremost thing that we always talk about. We've made a ton of improvements, and you're never done there. So we're always looking to improve safety. What comes with that? When you improve safety, quality gets better, production gets better.
We have a focus on quality, and we certainly have a ton of focus on efficiency. We have a number of different ways that we measure it, and we've set goals and guideposts and expected outcomes with all of our operating companies. We spend a ton of time. I think, again, one thing that's unique to Dycom, we're working across all 50 states. We're working across many, many customers, the majority of customers. We create mountains of data every day, mountains of data, and so the question is, how do you harness that data and operationalize it? How do you not just take it and be backward-looking, but make it be forecasting in what you do or what you need to do? Since our work is unit-based, very different from a lot of companies who might be percentage of completion, ours is unit-based.
So we get real-time view into how a project's performing. On the very first day, we can see in the systems that we've built out in the field that are completely digital, that our teams are using on the front lines, we can get real-time information on how a project is starting. And then we can adapt to that. We can look across the enterprise and across contracts and look for bright spots. Who's performing the best, right? Who's being the most efficient? Who's got the best safety records? All of those things. And we can take and bring those across the organization. So we talk about not having to learn things 30 or 40 times. So intense focus in that. How do you bring AI into that kind of data set? And that's something that we've been working on for years now.
We built, and I think I've talked about this before, we built a large language model that's reading tens of thousands of pictures to make sure that our buried depths meet specifications. That would take a massive team of people, and this goes to margin improvement, right? So you can either have a massive team of people, or we can have one large language model that's looking at this and making sure that we have a level of quality and only bubbling up things that need to be addressed. That's really the tip of the iceberg from an AI standpoint because of the data that we create, and I can tell you, we've got large teams working on this today, and we think that we're going to continue to improve that.
Ultimately, what you're trying to solve for is you're taking somebody that's out working with their tools, that's very expert at what they do, and we have to give them the right kind of information, the right kind of tools to also improve efficiency over time, so again, very proud of where we've gotten to and very impressed with what our teams have done, but we think there's continued opportunity for margin growth.
Okay. And maybe just thinking about the competitive landscape just for the service providers like you guys, I mean, can you just talk about, A, why does Dycom win versus perhaps some of the other publics and even the regional privates that you compete against?
It is a competitive space, to your point, Alex. It has been competitive for a long time. Our customers are highly sophisticated in the way that they procure work. As I've talked about on the calls, I think where we differentiate is our goal, our goal absolutely is to raise the bar and raise the expectations of our customers. Our goal is to deliver a level of certainty that our customers don't see from competitors in the space. Do we do that every day? There's always work to do. But we believe that we've done a good job of proving that if we tell you we're going to do something, we're going to do it, and where we do have missteps, our customers know that we're going to do everything in our power to overcome that and do even better.
We're going to continually raise the bar on ourselves, never satisfied. We think that gives us a competitive edge. We think that our approach to labor and our labor forces and the level that we can deliver with those forces, that that gives us a competitive edge. The knowledge that we have across customers, across geographies, and taking all of that information as we go look at new bid opportunities, we think that that gives us an edge to be more competitive and really knowing what the cost base is. So when you put all that together, I think it's the old proof is in the pudding. You see the growth that we've had, the continued growth opportunities. I talk about on the calls how we continue to get awards for new markets. I think that's a really important data point.
But at the end of the day, it comes down to the relationships we have with our customers and what our employment proposition is. And we believe that our strategy is working well. Always room to improve, for sure, but we think that we'll continue to be able to compete.
Okay. Maybe last one for here, and then I'll open up to the audience if they have any questions. But just thinking about kind of our whole conversation here, right, between the fiber to the home builds, BEAD, data centers, I mean, it seems like next year is shaping out to be a very solid year for Dycom, especially with one of your largest customers really accelerating that traditional builds. I mean, any kind of guardrails or guidelines of what to expect heading into calendar 2026 that you could provide for investors?
Yeah. As we did this year, we will give a revenue outlook for the full year as we get closer. Not going to give that today, but I would point to.
I tried.
Yeah. Yeah. No, I appreciate it. I would point to the back half of the year. You have a ton of momentum in the business. In Q2, record revenues, record EBITDA for Dycom, for the history of Dycom. Very excited about that. That rolls into a back half of the year where we've shown, given an outlook for Q3 that's robust, we've given you the full year outlook for Q4. What that implies as you go into next year, and certainly based on the dialogue that we've had today and dialogue in the industry, is there continues to be significant growth opportunities ahead for Dycom. We believe that we're well set up to be able to capitalize. We believe we continue to show that. And our goal is to be clear and transparent, right? I mean, it's our job to build trust in the investment community.
It's our job to build trust with our customers. We're always going to be straightforward on what we see. Right now, what we see, to your point, Alex, is our customers are highly invested in passing homes. They're highly invested in blanketing a majority of their footprint with fiber. You have other things like the AI and data center evolution, these long-haul and middle-mile networks that are really just getting started. And then on top of that, now I think we're going to finally see something for BEAD. We have the wireless work, which continues to perform and perform very well. That will give us a great position for next year, an opportunity to flex off of that. So all that said, we're confident in our ability to deliver for this year. And we've got a lot of energy and excitement around what next year looks like.
Maybe just to your point on the wireless side of the business, the Black & Veatch acquisition was, gosh, around a year ago here.
Just over.
I mean, that has performed very well the past couple of quarters. Can you maybe talk about what the work you're doing for there? I think you've noted in the past that it's not necessarily a pull forward of activity, or maybe perhaps a little bit of a pull forward, but you've still identified some opportunity on the outside here.
Yeah. It's a great customer. We worked with for a very long time. It was a great opportunity for us to marry what we've done internally and the solutions that we've built with our own wireless business with what we did in the acquisition. And so we were able to very quickly add to that business, retool parts of that business to be able to take on and deliver at a high level for that customer in the overhead equipment replacements. What that's meant is that, one, we can go faster. So there has been some pull forward. Two, the scope has increased as well. It's performing very well. Very pleased with the acquisition. Great to have those team members as part of our team. We see another two and a half years of that work continuing. And it sets us up well.
I think our strategy around wireless is to have it be large enough that we have a great footprint that we can flex into as densification comes in the coming years. And on the other side is, unless it takes a little bit longer, it's not so big that we can't backfill that very quickly with some of these other demand drivers. So we really feel like our strategy is solid there and feel well positioned.
Okay and maybe just in terms of the densification point there, I mean, have you had conversations with the carriers about them starting to accelerate that? I mean, as a tower guy, I'd be upset if I didn't ask you.
Yeah. There's always conversations with having their customers, a lot of which that I can't talk about when I'm on stage. I think there's a lot to be developed in exactly how densification is going to play out. I don't think that it's tomorrow. But again, I would also tell you that we have plenty on our plate for the next couple of years. And I think it just sets us up well. Ultimately, as I talked about with these other drivers, data consumption goes up. So over time, data consumption continues to go up, likely going to be more densification equipment requirements.
Okay. Perfect. Well, I think that's a great place to leave it. Dan, thank you so much for being here. It's great to see you.
Alex, appreciate it.
Awesome. Thanks.
Thank you all.