How Fiber-to-the-Home Providers Protect Their Investments

Good morning, good afternoon, and good evening to everyone that is participating in this podcast. So first, I want to welcome my two guests today, and that is John van Oppen and Neil McRae. And first I'm gonna quickly do a little bit of introduction of our esteemed guests — but maybe I should say that I am Jezzibell Gilmore. I am with Kentik. I'm the GM for service provider business, and it's my pleasure to be talking about quality of service's impact on fiber-to-the-home service providers. And what are some of the known facts and reasons why it's important, and what are some of the not-as-well-known ones and the misnomers? So let me introduce our guests. Sitting right next to me is John van Oppen. He is joining us as a longtime broadband industry leader and network architect with more than two decades of experience designing and building large-scale fiber networks, including starting Spectrum Networks, a wholesale network, and later a senior executive at Ziply Fiber. He has been at the forefront of delivering high-performance broadband infrastructure across the Pacific Northwest and really all across the United States. You know, I think there's a very practical engineering perspective that he will bring to us on how QoS strategies will help ensure fiber-to-the-home networks consistently deliver performance that modern applications — from streaming to gaming to cloud services and remote work — demand. Next, but not the least, is Neil McRae, who is one of the telecommunications industry's most respected network architects and broadband strategists. With more than three decades of experience spanning internet service providers, global carriers, and national communications operators, Neil has played a leading role in the evolution of broadband networks, including fiber to the home, mobile, and next-generation IP infrastructure. Formerly the chief architect of BT Group, and now the chief technology and information officer at CityFibre. He has helped shape large-scale network strategies serving millions of customers. His deep expertise in network performance, customer experience, and service assurance gives him a unique perspective on why quality of service remains a critical consideration as fiber networks support increasingly demanding applications and ever-higher customer expectations. So welcome, John, and welcome, Neil. Let's dive— Thanks for having us. Thanks very much. Let's dive into the discussion, because really, this is not about me. This is about you guys' input on the discussion. So first of all, quality of service is really important. And why is it important is why we're here getting together to talk about it. But first of all, what's the opportunity in the market today for fiber to the home? Why is that becoming more important? I can start with that. I think one of the most important things to think about here is what customers want. Right? And I've been a long proponent that customers don't care if it's fiber or not — what they want is a seamless experience. And this means having a good Wi-Fi router is as important as having fiber to the user. Making sure that the user can hop on, the service works all the time. They want it to behave more like they're used to from their electricity provider. You don't turn the light switch on and have a dim light sometimes, a bright light other times. You want it to turn on and go fully on. And that's what people are starting to want from their internet providers. And I think both of the companies we work for are demonstrations of how the incumbents have failed to understand that change in dynamic. Both of these companies exist overbuilding incumbents because they realized that users wanted more out of their connectivity. And in the US, I sometimes jokingly point out that we don't have the world's most reliable power infrastructure, but my goal is to always beat the power company in reliability — absolute reliability — but also make it so customers don't have a reason to interact with us. Because as much as we think we're awesome as service providers, and we're a really cool bunch, our customers don't wanna talk to us. They want to sign up for service, they wanna put it on autopay, and never think about us again for the most part until they move. And if we can get it into that scenario where the service delivers every time they open up their computer, every time they hop on their Wi-Fi — that's the goal. And that means low latency, high availability, a smaller number of single points of failure. These are all advantages of fiber. Only the ends require electricity. In the US there's a lot of cable deployments. Cable deployments were notoriously — and still are notoriously — unreliable simply because of the number of active components. You can get to the asymptotic end of that, but there's always more electronics in that footprint than there is in fiber to the home. And simple does really win in most engineering scenarios in the last mile. But bringing it back to what Kentik's important for: users, if they want to watch — whether it's the first fight on Netflix one night, or the next night they want to watch something that ends up on Akamai or Fastly — they don't care about those details. They don't care about any of those details. They want to open up their streaming app and have it work like cable, where it pops up and starts playing. And that's what we as service providers have to remember we're delivering. And they don't want to call us when that doesn't work. They don't want to mention it on social media. Matter of fact, it always bothers me — you look on social media, you only get the complaints about the installs. But I feel like that's a victory, because if there's no network complaints and everybody says it's awesome once they get it, after you've dug up their lawn or whatever to pull the cable in, which can't be avoided — that's a win from a networking standpoint. Right? That's what we're trying to deliver: that transparent experience. And I think that's what brings all of these overbuilders together in my mind. I don't know what Neil thinks, but those are the things that bring it together to me, in comparison to the incumbents. Yeah, I think you've got a lot of points kind of hit on the head, John. Here's the reality. We are dependent upon the network in ways — back in the '90s when we were building the internet, we never asked... If I'd said to you in the '90s, "Hey, one day, every time you take a taxi, you'll be on a phone," you would have just laughed at me, right? The world has moved to a point where we are insanely dependent upon the network, to the point where if you ever go to work and leave your phone at home, oh my God, that day is just awful. You can't do anything, right? I think as network operators, we should congratulate ourselves, because on that basis we really have changed how people behave and how they work. I don't think that there's an industry that doesn't touch other industries as much as we do in telecommunications. And that's one of the reasons why I'm in it. And I think John, you've hit the nail on the head. We're probably — and this will show how old we all are — we're probably kind of going into the third or fourth decade of the internet, right? And I've got a couple of grandkids who just expect it to work. This notion of a modem, or dialing, or this thing that spins while you're downloading, or the status bar — they're like, "Granddad, what the hell is that?" Right? And they want to press a button and for it to just work. What does that make fiber important? One, fiber really brings that on-demand capability where you press a button. In fact, I've just introduced a kind of notion at CityFibre, which is: if I press YouTube on this remote control, how quickly does it come up on the screen? That goes back to old-school cable TV channel change, where you'd hear stories of people with smashed TVs because it took too long to change channels on the old cable. Customers and users have just got more demanding, partly because we've done a great job in showing them the art of the possible — showing them how well their streaming can work, showing them how well their gaming experience works, showing how well they can work and collaborate from home, run businesses. And without fiber, it's very hard to do that. So here in the UK, we were known as a fibre laggard. We were behind, and there was a great push to move forward. Actually, CityFibre is one of the companies that — because of creating competition and challenging the status quo — really helped turn the UK from a laggard into a leader. And in the space of the last five years, we've gone from almost no homes connected to fiber to about 29 million. And John started with my most favorite word: customer. Customers need this, they want it, and actually some of them are willing to pay a little bit more for certain value-add services. And we're not just trying to change the QoS part of it — we are also trying to make it easier to buy, easier to understand. And John made a point I couldn't agree more with: customers don't care, partly because quite a lot of customers actually don't understand this stuff. And again, I kind of echo what John said. So the proof in the pudding is: okay, we've installed, the customer's plugged in. And is there a big grin on his face, or is there a "this isn't working"? And what I think challengers like ourselves have done — I've kind of said, yeah, this whole journey, it has to be what I call the five-star hotel experience, right? And that's versus the deli experience. So you go into a deli, you order a cup of coffee, and you hear twenty people shout "a cup of coffee for table five." Whereas you go into this kind of top hotel, you order a cup of coffee and it magically appears. And what we're trying to do is have that latter experience. And by the way, not only does it appear, it tastes beautiful, and the whole experience is something that — again, to some extent repeating what John said — is kind of invisible. It's kind of like magic and this stuff just works. And you don't have to talk to us, you don't have to email us, you're just able to focus. And again, the key thing in that puzzle, Jezzibell, is measuring how well you're performing. And the work that we do with Kentik on that — the work that we do to really understand not just "is the network up," but actually "how is the network performing" — I believe increasingly will become table stakes for operators. Because customers just aren't willing to accept that their broadband, in this fourth kind of decade of the internet, isn't reliable and doesn't work in the way that it should. And again, for us, I think that's a great position to be in, because customers are setting the bar harder. The one other thing I would just add about the UK market in particular: we have one of the most competitive broadband markets in the world. We have kind of five hundred-pound gorillas that sell broadband, three of them sell broadband on our network as a wholesale player. They themselves want to offer more than what the incumbents have historically given them. And with that competitiveness, here in the UK the kind of average price for a one-gig service is probably around the $20-a-month mark. Such is the competition. And it's fierce. As we've been in this build phase, we're now in this exploit phase, and one of the ways you exploit is by having amazing quality of service. For sure. And you know, both of you super well said. And as a consumer myself, right, where I have to be consuming broadband services for a slew of things — not just watching Netflix. I mean, I'd love to be able to just watch Netflix all day. Whereas streaming media for my son is more like going on YouTube and playing video games. But we all have different recreational demands. But I work from home. Right? When I'm not on the road going to meetings, I work from home. And there's high demand for fidelity in the content that I'm creating, like what we're doing with you right now, or that I'm consuming, or that I'm utilizing to do work. And I'm not alone, right? Just think about how many people today, in today's world, work from home. If they don't work from home a hundred percent of the time, they're generally in the office like two days a week. And most people say "return to the office" — they mean return to the office for some number of days a week. And for the rest of the time, they still have the flexibility of working from home. If you don't have the level of service that you're expecting, how do you work from home? Well, you don't realize that it has a problem until you see the problem. And so one of the things that's stuck with me, watching as ISPs have gotten better — especially the ones that are competitive — when there are outages, you see really interesting responses from customers about how they can't work. You see how much it impacts their lives. We had an outage of about a hundred houses because of another provider building literally in front of their houses and going right through the duct bank that fed them. And they were like — and you can't fix that, but that's the only kind of outage you can actually explain. "Hey, you know that backhoe crew that was out in front of your house? They're the ones who broke it. We were there too, trying to put it back together." But the feedback from people was what was so interesting to me. They were effectively missing a day's work. And they've gotten used to a service that was so reliable that they didn't even think about it until it wasn't there. And to your point, Neil, about we've done a good job in some ways — I think that was one of those demonstrations of how we've done a good job in some parts of the industry, making it so transparent that until it's gone, they don't even realize everything that depends on it. And that's a good thing. I was struck watching the last six years going through COVID, and the amount of changes in traffic flows and sort of SLA requirements that we saw was fascinating to me. And that would have been very hard to navigate without the Kentik system watching our traffic flows. We were trying to produce, you know, who do we have to add capacity to sooner rather than later, and where — and then figuring out how that was gonna fail over in certain scenarios when sites go down. Because the outage you can explain to people is the backhoe in front of their house. It's very hard to explain the backhoe, or a building fire fifty miles away, taking out their house. That does not make sense to them logically, and shouldn't happen in a good infrastructure design. But part of that failure analysis and preventing it is to expect failures. And I watched a great failure happen two weeks ago where we lost a whole peering site due to a pretty catastrophic set of router failures, and I couldn't find mention of it anywhere on social media. And it was doing hundreds of gigabits, crashed in the middle of the peak of evening. Oh. And nobody noticed. And you don't get those very often, but proper planning and analysis and metrics will really help you make those failures be a non-issue. Absolutely. So I mean, agreed. Actually, this last weekend we had one of some of the places where we build — we use cabinets for the plant just because of the geography of where we are. And actually we had an RTA completely demolish the cabinet, affected a few of our customers. But what was incredible: we heard from the field team where the locals were trying to help. They were bringing coffee and tea and "Hey, can we help get this back online?" They were understanding that something like this is hard to manage around. This was kind of a bit more of a rural community where, until we arrived, things like YouTube were almost a dream. Some of us have been in this industry — can you imagine the first time you click on YouTube and you watch a video? I mean, for me that's ten, fifteen, twenty years ago. For some people, they're actually just being able to do that now as we roll this fiber out. But in this scenario, our customers — or rather our customers' customers, because we're obviously in wholesale — they were really keen to do whatever it took to help us build this. And even, in kind of a previous life, we were building to an island and the fishermen actually helped the cable ship navigate the waters, because they could see the difference that fiber broadband would bring to them, and anything that they could do to make it happen quicker, they would actually do. And I think back to your point about homeworking. We have a bit of kind of pendulum swinging. Actually, my wife, who works in the neocloud industry, she works from home five days a week. She simply can't work without broadband. And the minute there's a blip on Teams, I've got a ping saying, "Neil, what's going on with the network?" Because that collaboration — one of the greatest things I think that we've taken communications from is, you know, two people on a phone to a group of people having a conversation like this, or people working to build something. And again, through COVID — this one can put tears in my eyes when I saw this — but we had musicians literally using Zoom to play together. Again, if you'd said that to me in the '90s, I would have laughed and said that's never going to happen, right? But this is the thing that, when I step back, is the network provides — in a very quiet, not loud way — an ability for people to do things that in any other way can be difficult, hard, or actually just simply not possible. And that isn't just about running a cable along a road and digging the road up. That's difficult. It's easy to get bits of that wrong. It's also, if you plan well, John — I think the point you made — it's actually quite easy to get it right. All those things demand quality. And I actually learned this during the Olympics in 2012. BT — or, at the time, we were the host network operator. And you suddenly start to think about the network in very different ways. And I remember having a few discussions about this where, you know, Usain Bolt has a hundred-meter run that literally lasts nine seconds. Done, it's done, right? And in the world of networking back then — and I remember talking to Avi about this problem, actually, back then — we had networks that we were polling every fifteen minutes, or even some even thirty minutes. So how do you know anybody saw that event? You don't. And actually that's what started us, and the industry as a whole — those sorts of events, and the fact that more and more people were watching them through broadband. It started that journey of telemetry, of real-time visibility of the network. Those seconds, those microseconds are super important. And having the knowledge that the network can perform — you've done your testing, you've done your failovers and you've tested things — but also, even in the situation where there is a problem, that the network can react very quickly. And I think we as network operators, we sometimes are challenged with being seen as kind of the dumb pipe. More and more, the pipe is intelligent. More and more are we putting in capabilities so that everything else on the internet looks great. The pipe's only dumb when it works right. You know, I think people like the dumb pipe idea, but to get a good dumb pipe to them is actually very complicated and very involved. Oh, extremely. Requires, to your point, the monitoring and the real-time visibility. One of the things that — I think you're hitting spot on — has been both a fear and a motivator is that coffee talk the next day with users. Right? You never wanna be the provider that failed to deliver. We've had some interesting calls on our side when the first big live events went out and CDNs failed in some interesting way. And ones where we avoided it, because as a carrier that doesn't really do any wholesale, it's all direct on our IP network. We watched a few other parties have complete meltdowns, and the one that made the press a lot was the Tyson fight, the first live thing— Yeah. It was really a winner for Ziply at the time to go through that without having a congestion event. And the coffee talk the next day — somebody who was at Netflix at the time, and pretty senior-level person, lives in the territory and used Ziply at home. And I particularly appreciated that, because the feedback was not random coffee talk. It was from somebody on the CDN side using it and saying, "It blipped twice, and I know for sure both of those were us, because those were the ones that I knew about internally. And otherwise, it streamed fine." And I'm like, that's— Yeah, I mean, I think that — to be, you know, that's kind of the... For us as a wholesaler, that's also part of the proof of the pudding for us. So one of our wholesalers is Sky, which is part of Comcast in the US. And Sky, their motto is all about being different, and they're very — they're known in the industry for being very selective about who they work with, how they work with people. And it is very much a collaboration. But I kind of mentioned this in the warm-up in the green room earlier on: Sky have this "Believe in Better." It's their kind of model. They try to do everything better than the last time. It's a constant improvement. And today we have a million connections in CityFibre. So we have a million live customers on our network. Congratulations. Yeah, with a celebration. And we had to, as part of that with our team, we had some of our big wholesale customers explain why we hit this milestone relatively quickly as quite a young company. And that's all to do with that quality of experience. And as more and more media companies move to IP and move away from traditional broadcasting techniques, that for them is super important. Sky is a TV company that kind of does a bit of networking, is how I might put it — or they might see that differently. But their goal is to provide great content in a very competitive content market right now. And for that experience to be — you know, the coffee talk to be about what happened in the program, rather than the fact that you can watch the program, right? And that's super important for them, and they're a great partner for us and their customers. When I joined CityFibre, I went out with a field team just to go through an install to understand what happened. And seeing that Sky TV come up like instantly, seeing the kids on their iPads playing Fortnite and Roblox and it just works — that smile factor in that moment from "we've installed it" to "it works" is incredible. And that's where, for me, when I see that happen, it's definitely a moat. But I kind of look to the future in terms of this quality of service, Jezzibell, and I kind of start to think: okay, more and more of the things that we do are going to get more and more dependent upon the network, either through AI or through other things that we start to depend upon. So this isn't a journey that has an ending for us in this industry. And something that we've done in this industry literally for the last 150-odd years: we have to continue to push to keep making this better and better, and keep enabling those things that make a difference in the world, work better — bring digital inclusion to people that aren't able to embrace it, so that they can be all they can be, and they can be creative, or they can work from home, or they can school from home. And as we start to, what I like to say, automate the mundane, the network is there ready to take that workload. And I always remind people that there'd be no AI if it wasn't for the internet and broadband. Right now, that's ultimately what's enabled us. So I think the future is exciting for us as an industry as well. And again, I say this to my team: if you're not measuring it, you're kind of making it up. A lot of the work that we do with Kentik is being able to — we really understand this, we measure it, we can have a deep understanding of what's going on — so that when we make decisions, they're driven by data as opposed to kind of rolling the dice. I think data is the North Star these days, right? So whether or not you can add things up in your head and figure out where it should go in a topology is not as relevant as being able to back that up with the data that shows what's gonna happen. Because those modes of failures — the networks get more complex and support more users in more places with more exit points, which is all the things you need for reliability — mapping all of the design pieces of that and measuring the performance levels gets orders of magnitude more complicated. When you're planning growth and you're modeling in your actual traffic, the number of permutations actually gets so big so fast that, you know, this gets cut and it goes here — you can predict that on a ring, but anything past a ring topology you can't easily predict that by hand. And I think that's just made more complicated with the move to the cloud, where certain peers are very important. And, you know, we find problems on a regular basis. And I prefer to find them before the customers do, before it's impacting. Every once in a while you'll get a customer-reported one, and I'm always excited when it's somebody else — you know, it's like, the customer report came in, oh, nope, it's the other end that's the problem, not this. And those are fewer and farther between the better you can manage that data, the better you can detect it, and you can look for trends and failures. And the Kentik piece — actually around traffic monitoring — was super important in that COVID period for us as a layer three provider, because we were trying to figure out where to put our attention. Because it was growing so fast. We could grow it, but only if we applied our growth resources and engineering staff to the right tasks in the right order. So we were uncharacteristically letting things get to, you know, eighty percent utilization and stuff, but we were doing it strategically using data — say this one needs to go get upgraded now, then this one, then this one, then this one — because you were limited on equipment, personnel, and velocity. Because we were growing... I don't know what happened on the CityFibre side, but our legacy footprint basically doubled in usage in three months. Wow. That's incredible. And when you see those sorts of growth numbers, to hold that reliability up, you have to double it too, because you might have eaten some of your redundant capacity. You need to understand what you ate in that growth and go augment it on the back end, so that then when the failure happens six months from now, you have a quiet failure instead of a loud one for customers. Because nobody wants the loud failures. And that's all kinds of customers. I can think of retail customers and small biz customers having very similar aspects. But in our space we also covered a lot of cell carriers and backhaul to other towers and things like that. And it's a big part of the business. And that "trust but verify" got to the point where we could verify that we were trustworthy enough where people stopped asking. And that was like this step where it turned, for me, watching it — where we went with several of the cell carriers from "prove that you met the SLA" to "oh, we didn't ask you for a year, because you had two outages and they were single sites. And one of them we created." You know, and you go to that mode where the packet delivery gets so good where, in an emergency, you're on a call with them and it's like, "How can we help?" That's phenomenal. And I love what you said about making strategic decisions for the business leveraging the data. Right? Because I think most of our audience here listening to this podcast are interested in how you are leveraging quality of service to help you grow your business more effectively and creating value for not just your customers. And I think that's number one thing, right? We're all customer-centric companies, so we all focus on the creation of value for our customers. But what is the value that you create for the company, in addition to just providing the service to the customer with this level of quality of service? Well, to me, I think that the metrics are very simple to watch. First, if you have unlimited capital, you can do lots of cool things, but you don't have a business if you spend unlimited capital. So I think it's important to differentiate: making it as good as it needs to be to meet the business objectives, but not spending so much money you make no money. So the things we looked at on the retail side — and I would wager Neil has the exact same metrics. But churn. Every customer you churn is another one you have to replace with marketing. And since we own the full stack of that, that was a big deal. Right? And so if you could lower churn even one percent, it made a gigantic difference, because that was like getting that many more net adds every month. Tech support. Nobody wants to call tech support. A certain percentage of times people call tech support, they cancel — because they get frustrated and they cancel. The users don't care what the problem was. They don't want the problem, and they want to try something else. Yes. And tech support has another aspect of it: it costs real money. Oh, yeah. Taking those calls costs money, and they cost satisfaction scores. And so I think our business North Star was pretty simple. It was going to this mode where we want to interact with the customer as little as possible unless they want to. So make it — if some of this is software — make it easy to upgrade your service without talking to a human. Make it easy to order install, show up on time for the install, dispatch software, things like that. Moving into the network monitoring and maintenance: if there's an outage, even if it's planned in the middle of the night for a customer premise — we gotta move a cable that feeds a bunch of houses because they're building a new salmon culvert— Pacific Northwest, that's what happens. Absolutely. And you let everybody know that there's gonna be an outage two hours between, you know, one AM and three AM. They're doing fiber cuts to do a culvert under this highway. Okay. Tell the customers. Communicate. Don't make it so they have to call you and ask. Right. Yeah. Don't have unexpected situations. If you do have failure, communicate immediately. The other thing that we found was, you know, network intelligence — and on the backbone part, Kentik is very helpful for this. And then we actually wrote a bunch of our own software on the PON side to help correlate outages across ONTs, so we could figure out quickly, without customer interaction, where the common points of failure were in an outage if something got dug up, for example. And those are the investments we're really starting to see in the network. Because the software investments — both on the backbone, things like Kentik; in the service assurance, Kentik again; or in the last mile, which is usually an in-house thing because ISPs have inventory systems that have to all be integrated to do that — those places are where we're seeing lots of executive interest on investing, because those drive down all the things that make metrics worse. And we've always looked at this a bit like a land grab. You wanna be the first somewhere, and then you wanna be awesome so they never switch, because that makes it infinitely harder for the second provider. Because if people aren't having trouble with the first one, they simply don't look at the second one. And I think that's where in the US cable really messed up a lot. They got such a negative opinion from the customers that when fiber started overbuilding them, we'd see double-digit percentages of the customers in a market we built move the first month. Wow. And when you're seeing that, that is a sign of dissatisfaction with the previous provider. "I don't know these new guys, but they're not the old guys." And so what you want to do is take that excitement and keep it. Yeah. I mean, maybe just to kind of end where we started, with the customer. I think when things go wrong — as they do, right, the RTA that we had that took some of our customers down, or maybe a power outage, or things that are kind of beyond reasonable control in many cases — communicating to customers is super important ahead of time. It's always better to tell them before they call you. Couldn't agree more with that. The one other thing that really frustrates customers that I've seen is where you send an engineer where the customer knows the engineer isn't going to be able to fix this. You do a truck roll, which is expensive, timely. You know, this isn't an issue here, this is an issue somewhere up there, right? And having the right tooling to identify that, and enabling our upstream wholesale customers to tell their customers that actually we know what the issue is and we're doing this and we're doing that — I think it's always better, and being as proactive with communication as possible. And also ensuring that your team's informed. Increasingly the broadband network's being used for things like healthcare pendants — for informed people that need to know that if they press this button, someone's going to come and help them. And if there's a fault on that broadband line, or there's some issue, when you send a few people out, to say, "Hey, by the way, on this service, this is an elderly customer," or "a customer that needs special care" — so you may have to tailor the process to that. And actually, here in the UK there's been some great work done here actually by Sky that we're plugging into, to ensure that those sorts of customers that need extra support, but also need the knowledge that everything works well for them, is increasingly important. And I think my goal in life is that before I go on the next journey, that everyone that needs to be connected is connected. And that is, you know, a big scale challenge, a big data challenge. But it's also an ambition that will really make a big difference in the world and in people's lives. And that's one of the great things about working in our industry. I love what you said, because it's the passion that we have for people that drives our passion towards the quality of service that we provide to our customer base — hence drives, actually, business value that we create for the stakeholders of the business. Right? This is not just the investors that we talk about. It's the employees that are part of the company, right? And the customers that we impact their lives, as you talked about, and how our employees feel about the work that they are doing. People wanna feel proud. Wanna feel proud. If you want good customer experience from employees, you want the team to feel proud of the service they're delivering. Yes. A hundred percent. I think the best — at least in the US, and I think BT is not too dissimilar, so we'll just name them too — the ILECs and big incumbents had a reputation for... even their employees knew they didn't care about the customer. And in some period... and they've come around in some cases, right? But that was the nail in the coffin for some of these businesses, and why we were enabled. And John, they've had to, because the challengers that have come in have shown that actually there is a different way of doing this. There is more customer-friendly, that shows that we are an industry that can deliver and can be relied upon. And actually, what I always hear from customers even more — or maybe more so on the business customer side — is they really want us to do more for them. Other services: security, voice, dot-coms. And they see some of the changes that we're making, and the trust with these customers is high, the stickiness is high. And the best way of winning business is being right. If someone's recommended to use you, rather than you having to pay that marketing effort to attract customers, so much the better. And one of my personal goals is that a big chunk of customers come because people recommend us. And that for me is — back to your point about the team — when you go out, when you go at a dinner party with friends, and they hear you're from this company, and then you get nothing but a whole lot of pain, as opposed to "I come from this company," "Oh, we love you and life was great and you fixed this" — that's where I want to get to in our industry. And I think we're on our way. I think we are lifting what the norms and the table stakes are, and I think that's a great thing. Yeah, I had a really interesting experience that demonstrates exactly that with an employee, which kind of brings the two together. So when we were looking at buying the bankrupt ILEC in an area — we ended up doing it — and I met an employee who gave me a tour when we were looking, and they said they put their badge away when they go to the bar in the small town, to make sure their badges were off them. And so they weren't branded. And two years later, I met the person. He's like, "Oh, I leave the badge on now." Yeah. I love it. And you know, if that is an indication of the change you can manifest by providing a service that the team believes in and is willing to recommend themselves. I love that. And I love that your successes involve Kentik's participation. Right? To me, I'm proud of what I'm doing at Kentik, because I feel that we are providing a tool for you to give your employees so that they can do their jobs better. You're helping us keep it boring. That's what I like. So I would say that the job in engineering is we're trying to keep it boring and exciting. Excitement should be the shipment that didn't come from, you know, X or Y vendor. It should be the new thing we're building. It should not be the operations. And I think Kentik is a really important part, at least for us, of keeping that operation boring day to day. And people talk about being excited about networking. I love networking, but I never want exciting operations. My favorite operational disaster was — we had a big windstorm a couple years ago, and we were on a conference call because we knew the windstorm was coming in and we were worried about it from a facility standpoint. And by the end of the two-hour conference call, forty-seven sites had lost power and every generator had started. And the whole team was sitting there like, "What should we do now? I guess maybe we should go around and just check on things and see if any trees came down in our parking lots or something." You know, we were doing a great job of keeping it boring. That's fine. And to me, that is the most important part about network operations today: doing our very best to keep it boring for our customers, because they just wanna flip the switch. They just wanna hit start on the video. They wanna get on the Zoom, and they want this to work. Well, with that, I just want to say that we are at time, and I want to thank both of you. Neil, all the way across the pond in London— All the pleasure. —and John here in Bellevue with me. Thank you both, and thank you all for everyone that's out there listening. We appreciate you paying attention and understanding the level of effort these networks are putting into providing quality service for their customers, and how we are doing the best we can in keeping network engineering boring. So thanks, and have a great evening, day, morning, wherever you are. Yes. Thank you. Evening here. Have a great night, guys. Thank you.