Uniti Group Inc. (UNIT) Earnings Call Transcript & Summary

[Audio Gap] with Uniti. My name is Greg Williams. I cover the cable, fiber telco and wireless space at TD Cowen. I'm joined here with Kenny Gunderman, President and CEO of Uniti. So Kenny, thank you for joining us.

It's good to be here, Greg. Thanks for having us, as always.

Sure. Before we discuss GenAI, which is on everybody's mind today and yesterday, a question we get a lot is your valuation unlock from where we sit here, specifically -- as we consider the [ some ] of the [ parts ] valuation that we would ascribe to Uniti, the company is still trading below those levels. Just curious what do you think will anticipate help Uniti to get to appropriate valuation, I think closing the deal would be one, but just help us out there.

Yes. So we've recently put out the proxy for our transaction and there's a lot of interesting background reading, the narrative leading up to the to the merger announcement, which shows a tremendous amount of strategic activity by us over the past several years. And a lot of that activity validates what we've been saying all along about the intrinsic value of our assets, including premium multiples compared to where we trade and frankly, compared to where we previously traded. So we've always felt undervalued thus working hard at strategic initiatives to try to unlock that value. So that's number one in a good backdrop. I think number two, with respect to the catalyst, we need to keep executing at Uniti. I really believe we have one of the best fiber businesses, one of the best-performing fiber businesses, and so continuing to show good execution is key because that's what allows those higher intrinsic valuations and continues to attract that strategic and financial interest in the assets. Number three, and probably more importantly, we need to execute at Kinetic and we have to educate the market on Kinetic, which is the fiber-to-the-home business at Windstream. And because that business has been private for 3 years and in bankruptcy before that, I think the market is cold on that business. And so we've certainly heard that in stage that we need to educate the market more. And so we've been doing that. And I think that's going to be one of the catalysts, is continue educating and continue performing. Thirdly, we need to relieve some of the technical pressures on the stock. So Greg, you mentioned the fact that closing is still a ways off. I think in reality, we think it could be sooner, and we're going to have more to say on that soon. But that's something that takes care of itself, the passage of time Secondly, we need to work on collapsing the dual debt silos, and that's something that we're actively working on, and we'll have more to say on that as well, I think in the near future. And then thirdly, the shareholder vote is right around the corner, and we still feel confident about that. So as we start to knock off some of these sort of technical pressures on the stock, I think those are all going to be catalysts. And that's way before we even closed the transaction. And then I think upon closing of the transaction, we need to execute. We need to integrate the businesses, get on with our synergy realization. And I think when that happens, the market will start to recognize that this business is really at the crossroads of two key themes in our industry right now, including conversions of fiber-to-the-home and wireless, and secondly, hyperscaler demand. And I know we're going to talk a lot about that. But we're very excited about telling that story. And frankly, as we've proven in the past, if we don't get the valuations in the public that we expect, we're going to go back to being active on the strategic front as well. So that's still part of the road map.

Right. Got it. And hopefully, we can -- have time to talk about both the convergence on the Kinetic side. And but then let's talk about enterprise and GenAI demand on the fiber side first, the fun stuff here. Lumen came out not too long ago a press release noting $5 billion secured on GenAI-driven fiber deals, another $7 billion in their pipeline. Maybe just talk about Uniti's opportunity and your thoughts specifically to that deal and how that can really translate into the things that you're doing?

Yes. We definitely noticed that deal and have followed very closely the market reaction and both good and bad fallout from it, and we've fielded a lot of questions about it. And I think there's some -- probably some misconceptions about it, and there are some things that the market maybe is focusing on, that's not correct. And I think, number one, for Uniti, the hyperscaler demand is definitely real and it's growing. So at the beginning of the year -- and it's ramped up very quickly. So at the beginning of the year, we had roughly a dozen opportunities in our funnel for hyperscalers, and now we have close to 90. So within a very short period of time...

A dozen to 90.

To 90, ramped up quickly. And we started to ramped deals with the hyperscalers. We've done deals with the hyperscalers in the past, but we've now started to do deals on a more regular cadence. And so I think one of the things that the market seems to be trying to do is put all of these hyperscaler deals into a single bucket...

Very difficult to put in a box.

In reality, they're very different, and they're very consistent with how we sell to all of our other customers. So for example, in the second quarter this year, we sold a deal to a hyperscaler, all dark fiber, all intercity. The dark fiber lease, which is MRR and EBITDA, but -- and it positively affected bookings, no CapEx and very high margin. So a terrific deal. And so I'd like to say that it hit all the vanity metrics.

No CapEx, [ so does the ] lease-up on existing fiber...

All existing fiber, all lease-up, a terrific deal, hit all the vanity metrics like bookings, revenue and EBITDA. But then in the first quarter of this year, we did a deal with a hyperscaler, which was building greenfield fiber in a market, preexisting market for us. And it came at a high NRC, so we typically talk about yields on anchor deals of 5% to 10%. This yield was well in excess of that, and it was greenfield fiber that we now have the opportunity to go lease up on a recurring basis with our own fiber. Terrific deal and we're going to feel the benefits of that deal for many years into the future because we've laid fiber that we're now going to be able to lease-up, very consistent with our anchor lease-up model but it didn't hit the vanity metrics, right? It didn't hit bookings. You don't see it in MRR or recurring EBITDA anytime soon. but that's still a terrific deal. And I think that's probably closer to what Lumen did.

That's the misconception that you speak of it's a -- but not hitting the vanity metrics to use your words.

But that doesn't mean it's not a terrific deal. And ultimately, that's a deal we would do all day long because it -- if it comes with a lease-up model on the other side, and these deals...

That's key, right? I mean you're building these things out of the Tier 2, 3, 3 markets, even Tier 4 markets because maybe that's where the power is for these in data centers. The risk is, can you lease them up? Or is it just going to be one strand or a data center to nowwhere?

That's exactly right. And we've seen those one strand to nowhere deals, and we haven't done them. And I think that's my third point, which is I think as a fiber industry, we're a lot smarter about deals that we take on, and we're a lot smarter about not overbuilding each other. And so there's a lot more discipline in the industry, which is a good thing. Last point, sorry, I'm going on. But I think I've heard a lot of debate about the AI use cases and how that's important because if -- are those use cases are really going to drive demand or not? Is it really sustainable? I think the reality is the demand is more hyperscaler-driven than it is AI. I think the hyperscalers are investing in this infrastructure and they're even saying publicly that if they don't use it for AI, they're going to use it for something else. So AI is just one of the use cases of many that I think the hyperscalers can use. So the incremental TAM that we're seeing don't attributed just to AI, attributed to the hyperscalers. I think that's a significant difference. And very last point, we're in this really lumpy period where we're investing in the learning modules and the big proprietary data centers. And so you're going to continue to see lumpy deals and deals that don't necessarily hit the vanity metrics. But once we get to the inference phase, I think that's when businesses like fiber-to-the-home, fiber-to-the-building, even fiber-to-the-tower and small cells are going to start to see that benefit because that's when that edge, low latency streaming or online inference is going to be key. And I think that's when you start to eventually see the MRR effect.

Right. Okay. So when you move towards the inference, that's when you get those maybe lease-up opportunities, and make up for, for lack of a better word. Maybe more of a technical question. Just curious how many connections you [ grow ] into to a data center? When we speak to the data center products, they want to see 3 or 4 networks going in. So -- but Microsoft just signed one large [ amount to ] Lumen. And in your case, in Huntsville, maybe not speaking of that one specifically, but in general, like how does that work? So are you the only one in Huntsville because that doesn't make sense to what I hear about the hyperscalers wanting a couple of providers in.

I think that's a good question. I think the key is it's not necessarily multiple carriers, it's multiple routes. So for example, if we can be and are the sole carrier in some of these data centers, but we're selling multiple routes. So redundant -- and that's -- you get how redundancy works. I mean if there's 1 route that's down with us, there's a second route that's up for us and that redundancy works. So that's particularly true in Tier 2 and Tier 3 markets because typically, there are not multiple carriers going into those markets, and that's really where we are. But there are other markets, including the NFL markets, where that redundancy does tend to come from multiple carriers.

Right. So you can be the 1 carrier, but you are -- maybe you have 2 or 3 different points, we could reach that data center?

Correct.

Okay. Got you. Maybe just talking about the margins that you see on these higher upfront NRC sort of deals. Typically, I would have thought that when they pay upfront, you've got to put that back out there in terms of CapEx in the data centers. And I'm just curious if that's the case or not, you actually [indiscernible] some of that?

It's a good question and yes, the answer is yes and no. In some cases, when we're asked to build new fiber or to build off of existing fiber with laterals, then yes, there's new construction and the IRUs or NRCs, whatever you want to call them, are used to offset that capital spend. And so therefore, you get into this discussion about yields. And I'll tell you that, whereas in the past, we've always talked about those 5% to 10% anchor yields on new builds. What we're seeing is substantially better than that on these hyperscaler...

But that is DCF from the upfront cash then?

No, that's a yield on the capital that we're putting at work, also burdened with the OpEx that we're putting to work on day 1. So that's the initial cash flow yield. There are other transactions where if you're selling existing infrastructure and there's no CapEx, then those are obviously very high-margin transactions. And so if you have a large embedded investment network, then there's a lot of those opportunities and that's the case for Uniti.

Right. And going back to the comments about 90 deals in your funnel, I'm assuming -- help me with the size of those. I mean any Huntsville deals like that? Are they a lot smaller, just trying to understand the nature of these deals.

Yes. They're across the board. We don't have any $5 billion deals, and I don't anticipate having any of those. I think those are more of the unicorns, at least from our vantage point. These are more tens of millions, hundreds of millions of dollars of deals. And so there are definitely deals like the one you're referring to out there and those are growing, but there are also just those regular [ Wave ] deals. We're selling [ waves ] to hyperscalers. We're selling intercity dark fiber connectivity to hyperscalers, so not inconsistent with just our regular [ Wave ] business. So if you were to ask our sales team, if they were here, they would not attribute this demand to AI. They would just attribute it to customers that are buying fiber capacity, both lit and dark at a much higher velocity than they have historically, including going back to just last year.

Got it. And so we talked about the economics and the nuances of dark fiber related to GenAI. I'm curious about the pricing environment in GenAI on your lit services, particularly the [ Wave ] business. Will pricing increase, will decrease, no incremental change from what we're seeing in GenAI on waves?

It's very route-specific. I think on dark fiber is a big part of our business. I mean I prefer dark fiber overlit, frankly, higher margin, less CapEx, longer-term contracts, more predictable, lower churn. Dark fiber pricing has stayed very stable in the industry for years. I mean if you go back 10 years, people are still selling $1,000 per strand mile. And that's roughly what we're doing today, plus or minus, depending on the route. On lit services, it's different because there are many routes that -- if they're lit, almost by definition, they're more competitive. And so when you've got competition, you've got to respond to the market. We tend to stay away from the more competitive markets. Our whole -- one of our true notes of our thesis is to target those Tier 2 and 3 markets and routes. And so when you have those, you have a little bit more pricing power. And I think in addition to that, there's a pretty steady cadence of upselling bandwidth, right? So waves are now -- they're not 10-gig anymore. They're not even 100 gig. It's now 400, and we're even looking out 800 and probably terabit plus. So similar to products in the past, if you're able to offset pricing compression with bandwidth upgrades, you can pretty much stay neutral to positive.

Right. Maybe same or higher ARPUs, but a lot more units.

Yes. And by the way, I'd say just tying that to your question about hyperscalers, I've said this before, but the hyperscalers are -- they're very smart at buying capacity. A lot of the people at the hyperscalers are telecom people, they know the industry, but they also have more capital to put to work and they tend to be less price sensitive and more focused on quality.

Get it done fast and...

Quality of network, speed to market and predictability of speed to market. So if they say they need something in 6 months, they mean 6 months, not 9. And so I think they want to work with people or companies that have a proven track record of construction.

Got you. And you noted on your earnings call that 40% of your funnel is GenAI. Is that the 90 deals in the funnel. So is that good a good extrapolation that...

Yes. We've gotten that question a couple of times. So what we said was 40% of the bookings this quarter were from AI. We didn't actually give a specific percentage of the funnel that's related to hyperscalers. I would say it's not quite that high, but it's growing. And I would also say that it's also not -- part of the way we talk about the funnel publicly is just MRR, which means bookings and it doesn't capture some of those deals like I described earlier that don't hit bookings. So if you...

It is upfront though...

Right. So if you included just contract value in addition to MRR, that number would probably be higher.

Right. And like I said earlier, a big focus would be lease-up deals, right? So in order to obviously get higher margins and low CapEx...

Right.

Okay. And then you also noted on the earnings call that wireless was picking up a little bit in the second half. To what degree do you expect to pick up as we consider the weaker 2023 going forward spending? And is this a small or a gradual pickup? Or is it going to come to a meaningful degree?

Yes. So first, I would say -- and I heard this on a previous panel, which was great, by the way, that none of us fiber companies are betting our future on hyperscaler demand. We didn't build our businesses expecting that demand. This is definitely incremental to the TAM. So I think that's something to keep in mind. And to that point, as a wholesale provider, we pride ourselves on being pretty diversified across hyperscalers, wireless, international ISPs, fiber-to-the-home, carriers who are buying backhaul and others. And to that point, wireless has been down. Wireless was down in 2023, but you really didn't see an impact on our business. We still grew at those -- that solid mid-single digit. And I won't say how far down bookings were in 2023, but they were down. And what we said at the beginning of the year was, we thought bookings would start to tick up this year, and it would -- we really start to see it in the second half of this year on wireless. And that's what we're seeing. And so I think by the end of this year, we'll probably booking -- wireless bookings will probably be up around 10% over 2023 year-over-year. And a lot of that -- it's still speculative, but the early indications that are typical of business are out there, which are wireless carriers are talking to us about what they need. They're talking about 25 gig versus 10 gig. They're talking about that they're starting to put out RFPs with multi-site locations, including existing sites, new sites. So those are all like leading indicators that suggest opportunity is coming. Certainly in the second half of this year, and I expect probably a bigger pickup in 2024.

Interesting. So it's RFPs for multi set, but you also mentioned maybe shoring up some of the fiber-to-the-tower. So 25 gig sort of spectrum is that...

And even new sites. So [indiscernible] and we're even seeing a little bit of activity on small cells, which I've -- we've always talked about small cells as probably not being a huge part of Uniti's business at least in the near term because we're in the smaller markets. And we said when small cells come to those markets, we'll be there and be ready. So we're starting to see a little bit of that, but I think that's probably still a little bit off into the future.

I was curious if you think maybe AI could be a driver to the Micro Edge and maybe we do finally see small cell proliferation.

I think it definitely is, and it's one of the reasons why -- back to what I was saying earlier about that inference phase. I think that's when having the fiber-to-the-home, fiber-to-tower, fiber into the building is going to be key because that is when latency matters, that is when speed -- latency matters, reliability matters at the Edge. And I think that's going back to what I was saying earlier, that's been consistently true of the hyperscalers business model for years, right, not just about AI, but other business cases. And so I think on a go-forward basis, those types of businesses that are more MRR-driven are going to benefit from that cycle.

Got it. And on the earnings call and even today, you again reiterated the M&A further, I guess, in the fiber space and even fiber-to-the-home if we want it for right convergence. But we all read the S-4, and we saw that there was no shortage of discussions happening. Maybe help us with the various M&A scenarios that would be attractive to you as you sit here today?

Yes. Well, first, I think on conversions -- I know there's a lot of debate about whether that's real or not and whether it's important or good for the industry. I think some of our biggest customers talk about it as an important theme and the others -- other of our big customers, not necessarily. And so I think the body of empirical evidence is still developing on whether it's important or not with respect to customer retention and upselling customers or just selling to new customers. But one thing I think is indisputable is that there is a substantial amount of investment being made in fiber-to-the-home. And whether you think that needs to be bundled with a wireless offering is not really the question. The question is whether you think fiber-to-the-home is a good product and whether or not it could be in the road map for some of these large strategics who are looking to develop their conversion strategy. And I think the answer to that question is emphatically, yes. And so at Uniti and Kinetic, Kinetic is developing or has developed a wireless offering along with its fiber-to-the-home offering. It's a bundle with AT&T. I think if they were here, they would tell you it's more of a retention strategy as it is opposed to attracting new customers, so more retention. But I think with respect to the strategic -- and maybe to your point on the M&A, the strategic value of that business to me is substantially enhanced today than what it was when we signed this deal and certainly what it was 6 months before the deal. And I think 6 months from now, 12 months -- 24 months from now, the strategic value is going to be enhanced even further.

Sure. And I guess it doesn't even matter if there's empirical evidence waiting for convergence or not. I mean, at the end of the day, if 1 or 2 of the carriers are -- cable companies, et cetera, are going down that path. It sort of forces the others to do so almost whether you like it or not. I mean does that change your outlook on the Kinetic assets in any way? I mean, I think you kind of said that it's fiber-to-the-home regardless. But now maybe it's [ cable ] to the home on steroids at this point?

I think it is. I think it's -- look, before we did our transaction, and as we were contemplating it, we certainly felt that it was a strategic asset, some of the background reading in the S-4 validates that there's been interest in that business for some time. So we've been convinced of that for some time. I think what we've seen over the past few months just is further validation of that. I do think that it's also one of the -- to your point about on steroids. It's one of the reasons why we think expanding that build and getting to 60% plus coverage of the footprint with fiber is important. We think that's the right model. We think the returns are there, and we also think that makes it a more strategic asset and a better performing asset on a go-forward basis.

Right. And then how about M&A on the enterprise fiber front. Any updates on keeping Uniti Fiber within a QRS? And how do you think that will unfold?

Yes. So just as a reminder, our view of the new business going forward is we would have Kinetic, the fiber-to-the-home business, the Fiber Infrastructure business, which is our current Uniti business plus Windstream Wholesale and then there would be the Managed Services business, which is Windstream's current enterprise business. In that lateral business, we've been public about how that's not core to our fiber infrastructure strategy. It's a good business, cloud-based enterprise product business, good strategic value to it, but it's just not core to our fiber infrastructure business. I think the fiber infrastructure business and Kinetic actually go well together synergistically. We've been very candid about how the fiber-to-the-home business needs that middle mile and it needs that backhaul to the core. And in fact, some of our biggest customers over the past couple of years have been fiber-to-the-home providers buying backhaul from our wholesale business. So we think there's a lot of synergy to keeping those businesses combined, and that's our base case. But at the same time, I think there's potentially 2 different strategic homes for those businesses. And I think over the past again, 12 months, I would say the interest in wholesale and enterprise fiber has also been enhanced and hyperscaler demand and AI discussion has not dampened that enthusiasm, I would say.

Yes, I'm kind of curious if the GenAI fiber discussion is actually maybe put some of these assets in play? I mean does it put...

Well, I think it does. I think our -- today, we've had a full schedule of meetings and this is the first conference post the announcement of our transaction with Windstream, where the meetings have been dominated by questions about our wholesale business and our enterprise fiber business and hyperscaler demand in generative AI, whereas previously, it was fiber-to-the-home. And so I'm not saying there's a shift necessarily, but I do think, as I said earlier, I think the new Uniti is really at the crossroads of those 2 themes, convergence -- fiber-to-the-home, wireless convergence and hyperscaler demand, which is additive to the fiber TAM. So those are both good things for this new business.

Right. And then that third piece, you mentioned Managed Services is noncore. I mean you were exiting Managed Services even before. So now you've sort of inherited it. Just maybe an update on Windstream's Managed Service business and when or how you can look at possibly sell it?

Yes. It's -- as a reminder, we're very infrastructure-focused and product-lite. We like to focus on connectivity. So Ethernet DIA, Internet waves, dark fiber, so really simple connectivity products. We do some Managed Services, but it's more sort of I would say, lite touch, and it's definitely value accretive, but it's a very small part of our business and it's smaller than it was, right, because we've exited most of that. Windstream -- and it's all -- and our infrastructure business is all on-net or predominantly on-net. So we're big believers in owning the underlying infrastructure and getting owners' economics on that connectivity sale. Windstream's enterprise business is different from that strategy. It is a -- it's largely off-net, and it is cloud-based, UCaaS, SD-WAN, security, firewall, again, good businesses and cash flow accretive businesses, very little CapEx associated with it. EBITDA is flat to slightly growing. And so there's good strategic optionality to keeping that business as a part of the overall Uniti and having it be accretive from a cash flow perspective. But we also think with respect to having a fiber infrastructure story that's simpler, they're strategic alternatives for that business. And there's definitely some strategic interest in those assets, particularly the cloud-based assets.

Right. And then just shifting gears to Kinetic specifically, you mentioned that it's been private for a couple of years and through a bankruptcy process and you want to come out and educate investors on Kinetic and what it is and and the attributes. Maybe you can help us today and highlighting some of the things you do want to go out there and educate the investors on Kinetic.

Yes. So -- well, I'll start with some of the vanity metrics, and I don't mean vanity in a bad way, but I mean the ones that people tend to focus on. So number one, it's more of a rural footprint. So similar to our strategy, Uniti are focusing on Tier 2 and Tier 3 markets, build fiber first, build fiber early and you have a right to win for many years into the future. Kinetic is doing that. It's a more Tier 2 and Tier 3 footprint with less competition. And so the percentage of overbuilders in their markets has hovered around 15% for 5 years now. So once they started building fiber-to-the-home back in 2019, overbuilders have largely stayed away from those markets. So I think that's an important fact number one. Number two, ARPU is pretty healthy, almost $90 per month ARPU, and it's been growing consistently. We get the question, is that too high? Is there some pricing pressure?

At times you see that, let's say, Cable One that's more [ rural ], so they sort of command a higher price ARPU and now on a decision they're making on their own, their ARPU is getting hit, and that's a question we get as well as was the ARPU growth sustainability at Kinetic?

I think it is. I think there's -- so when Windstream reports that number about 70 -- the true broadband ARPU is around 70, which is not substantially higher than most, right? So the other things in there are cable modems, a little bit of voice, a little bit of residual video. So I don't think there's as much downside pressure as people could be concerned about. And plus, in these more rural markets where you have candidly less competition, you do have a little bit more pricing power. So I think that's a good story that people just need to understand. And then thirdly, cost to build. Kinetic is building at $650 per home. That's lower than the industry average, and you can debate what that average is, but it's still low and attractive. Historically, Windstream has built a lot of fiber to get fiber to the node that is really just investing in the Uniti network. So it's building some of that long haul, building some of that middle mile and ultimately to the node, that they're now taking advantage of because they don't have to build all of that to get to the home.

Right. All the central offices and nodes are tied up with fiber Ethernet. So [ Kinetic is the ] shorter last mile.

Correct. And we talked about that on earnings. We talked about how over -- almost 100% of nodes that are broadband-enabled have that 1 gig connection to the node and 10 gig connection to the central office -- or from the central office. And thirdly, Windstream several years ago put together their own internal build capability. So back when we were all debating labor shortages and could you get the teams in place to actually execute on the build, they made a very strategic decision to develop their own internal build -- in-house build capacity. And so that not only gives them the ability to control that build, it gives them the ability to fund it at a lower rate. And we estimate roughly 20% to 30% savings on building it internally. So anyway, You put those things together, more rural footprint, attractive and growing ARPU, attractive build capability and build rate, there's a lot of potential in that business. And I think, historically, I wouldn't say Windstream has been capital constrained over the past few years. But I think on a combined basis, that Kinetic business can now benefit from the owner's economics that I mentioned earlier. And so I think on a combined basis, we're going to have the ability to really expand that build.

Right. And on your first point about being rural, I mean, I think penetration could be even higher in some ways. So what do you have for terminal penetration targets at Kinetic? I mean, typically, to be 45%. But in rural areas where you're sort of the only game in town in some of the markets, It could be way north of that. I'm just curious if you have...

I think -- I don't want to give a number because Kinetic has not communicated about that publicly, but I think your general -- your general direction is correct, I think.

Right. And to what extent do you anticipate Windstream participating in the BEAD program? Are your hands full at this point? And would you need a partner to go out and build some of these BEAD homes?

I don't think we need a partner. And I do think Kinetic and we will participate. It's really -- we get asked the question, why are we doing this combination with Windstream now versus some time in the past or some time in the future? One of the reasons is we wanted to approach this BEAD opportunity on a combined basis because I think this is a -- by definition, BEAD is targeting the smaller, more rural markets, which is exactly where Kinetic's footprint is. And so Kinetic has characterized this as a once-in-a-lifetime opportunity for that business. I think that's a good way to think about it. I think there's a lot of opportunity there. Our hands are not too full. This is something that we're very focused on, and I think there'll be more to say about that in the future.

Got it. And then the other way of playing the convergence game would be for Windstream to tack on an MVNO instead of maybe a larger strategic taking out Kinetic. So I'm just curious where Kinetic thoughts are in terms of exploring an MVNO opportunity?

Yes. So I mentioned this earlier, but they have actually put in place a -- it's not an MVNO, but it's a co-selling opportunity with AT&T on the wireless side. So it's a wireless bundle where they're giving customers a discount if they bundle with AT&T Wireless. It's -- I think they view it more as a customer retention tool than it is attracting new subscribers as a way to sort of counterbalance the cable cos who may have an MVNO in their market. And they just started that at the beginning of this year. And I think the early returns on it are pretty -- are good. It's performing in line with what their expectations were with respect to retention and there's actually been an uptick in customers who are taking that service.

Right. You noted Kinetic brought everything in-house or for the most part, and that's helped with the cost because my next question is, is labor and handling labor with upcoming BEAD challenges and then, of course, supplies. Maybe you can talk to both of that within Kinetic and then at Uniti. I guess the second question to that is, how do you think about the in-house. Sometimes there's pros and sometimes there's cons to it. And wondering if you would actually do some of that on the enterprise fiber side as well.

Yes, good question. So at Uniti today, we do about 30 -- somewhere between 20% and 30% of our own construction in-house, and the rest is obviously outsourced. That works really well for us because if we have a true really strategic build that we need to hit a time line. We've got an important customer. We can deploy our own resources and get it done. You know you're going to get it done and get it done in a high-quality fashion. It also gives us a really good read on the market. So we know what labor costs. We know what materials cost, and so it's a way for us to keep contractors honest for lack of a better word. And it gives us -- and we're big enough and we do enough new construction that we've got contractors in our network that know us. We work with them regularly. And so we go back to them. If we go back to them on a regular basis, they know there's business there. We tend to get moved higher up on the list from a priority perspective. So we've never had a shortage of labor on the Uniti side, and that balance of 20%, 30% in-house rest outsourced works well. The challenge on in-sourcing too much of it is the economics work great as long as you keep those crews busy full time. What kills you from a margin perspective is if you've got crews spending a day or 2 a week driving from location A to location B, that's lost time.

So that's the penalty of being rural, but also maybe just better planning could be done?

Correct. It's good -- but as long as we're executing in our general footprint that works well for us. On Kinetic, I would say they have been able to keep those crews busy virtually full time. And so the economics have worked well, and that's why it's contributed to that low $650 cost per build. That's going to continue to be the case as we go through expanding this build through the BEAD program and otherwise. And I think there may actually be a time when we need to bring in external contractors to help supplement that build. But I don't think any time in the future, they will -- there will be a time when those crews [ aren't ] largely fully utilized.

Right. And the FBA just recently put out a report saying that we're going to need a 58,000 fiber workers. So I guess in-house could work to your benefit.

I think it has so far, and I don't see that changing in the near future.

Got it. With that, we're about out of time. So thank you, Kenny, for your time.

Okay. Thank you.