The following transcript has been edited for length and readability. Listen to the entire discussion here on The Broadband Bunch. The Broadband Bunch is sponsored by ETI Software.
Pete Pizzutillo:
This episode of The Broadband Bunch is sponsored by ETI Software and VETRO FiberMap.
Brad Hine:
Hello everyone in Broadband land, and welcome to another episode of the Broadband Bunch. I’m your host, Brad Hine, continuing to bring stories, samples, and stats from the world of broadband. Today is day two at the Fiber Connect 2023 Conference put on by the Fiber Broadband Association, and we are on-site at the Gaylord Palms Conference Center in beautiful Orlando, Florida. Joining us today is a longtime leader who has helped shape a variety of technology businesses over the last 20 years. He was also asked to speak about a fast-growing piece of the broadband industry that has recently taken a huge spotlight in the tech world, namely artificial intelligence. Please welcome to the show, CEO of Actifai and Partner at Foundry.ai, Ned Brody. Ned, welcome to the Bunch.
Ned Brody:
Thanks, Brad. Thanks for having me.
Brad Hine:
Absolutely. So is this your first Fiber Connect, you’ve been to others in the past?
Ned Brody:
This is the first Fiber Connect for us.
Brad Hine:
Well, welcome. Welcome.
Ned Brody:
Thank You.
Brad Hine:
How’s your christening so far? Are you enjoying the conference?
Ned Brody:
It’s been excellent. We’ve met a lot of people, learned a lot, and made a lot of new connections. And hopefully, we can help some people here, and they can help us.
Brad Hine:
Yeah, great, great. Now, the topic that you spoke about when you and I originally talked about is one that’s gaining a lot of publicity. People are really interested in what it does. I think some folks have different definitions of what it is, and I’m sure it does depending on the context and the application of itself. So artificial intelligence, AI. So when we speak about AI, what are we really talking about? What does that mean? What does that mean to you and Actifai?
Ned Brody:
Well, artificial intelligence is certainly the most hyped term of our lifetimes, at least at this point. It used to be cryptocurrency, but it beat that out last year. So a lot of people get very concerned about the types of AI, the differences between them, and terms like machine learning AI, generative AI, and deep learning. It really doesn’t matter when you’re thinking about how you can apply AI to your business. And that’s really what we’ve been focused on for the last couple of years. Everyone or every executive in the broadband industry has heard about AI. Most of them have used it in the form of ChatGPT. And all are being asked by their other team members or their boards, how do I actually use this to drive some beneficial outcome in my business?
So when you think about artificial intelligence, the first thing is not to get too concerned about all these variations of differences. Think of AI as what you would do as a human being if you had three things — relatively unlimited access to data, the ability to process all of that in real-time, and then not make any computation errors. What you’re trying to do is emulate what a human being would do. But very few times in our lives do we have all the information available and we’re able to process it and we make a lot of mistakes. So that’s really what the definition of AI is. Think about it just as what you would do if you were close to omniscient.
Brad Hine:
So that’s easy to swallow. I appreciate that. So you spoke yesterday to a bunch of folks on a panel. So give us the context of where you were focused yesterday.
Ned Brody:
Yeah, so it’s really around this question of how I as a broadband executive bring artificial intelligence to help my company, my customers, or whatever key metrics that I have for my business. And I think at this point, we are the only artificial intelligence company that is focused on broadband natively. That actually matters, and the reason it matters is that all of artificial intelligence uses a concept of a training model to have an outcome. So when you think about artificial intelligence, think of whatever term you want to use, either inputs and outputs or a stimulus and response. The idea is if I say something or do something or ask something, what will happen on the other end? So we’ve all used ChatGPT.
And that’s typically, I have a question, and then there’s response written typically in English or created graphically that answers that question. That is trained on a relatively large set of text-based information. So if you ask a generic question, it can answer with an answer that hopefully gets pretty close to what you want. When you talk about business applications, usually you’re looking for a very specific answer. If I say or do this, what will the customer do? And you can’t really use ChatGPT to do that because it’s not trained on those sets of stimulus responses. So for example, if I offer a particular offer to a customer; and I say the following script, what will that customer do?
To train the AI to do that, you must have hundreds and hundreds of thousands of that instance. And that’s really what we came here to do, is to show how you can answer key questions like that in business.
Brad Hine:
Got you.
Ned Brody:
I can get much more specific about the application if you’re interested.
Brad Hine:
Yeah. Stats, numbers, stories, samples also. So before we jump into that, maybe let’s dip back into your career a little bit. So how’d you get started on this path, and where did you start? I know, obviously, you lived through the dot-com era. And so in the last 20 or so years, where have you been, and what have you experienced?
Ned Brody:
Well, my career was kind of circuitous. I came out of my undergraduate degree with something called decision sciences, which was a long time ago, unfortunately. It was really about half game theory and half computer science. So really trying to understand what human beings would do in situations, and we studied things called utility theory and what consumers have value for. Then, I went into strategy consulting for about a dozen years, mostly in the US, Europe, and Southeast Asia. Then, the internet boom came about in the late 90s, and I joined an early search company called LookSmart that we took public back in, I think it was ’99.
And I then did a series of roles in what I would call large internet. So I became the chief revenue officer at AOL. I ran commercial operations for Yahoo in North, South, and Central America, all at a time when the first real form of artificial intelligence was being used, which was programmatic advertising.
So instead of advertising being designed of you and I sitting across the table and buying something from you together, programmatic advertising said, “Well, I’m going to allow people to buy on an exchange and the optimization algorithms are going to be designed to put the right ad in front of the right person at the right time to get the best outcome.” That really was the first real installation of artificial intelligence. So I ran that at AOL and then did a lot of that at Yahoo before I started a new company called Foundry.ai, which was about six or seven years ago. The idea behind that was that everyone was asking, “Well, how do I use AI in industry?”
And we would go out and find companies that had a business problem that they wanted to solve or alleviate. If we mutually thought there was an application of artificial intelligence that could do that, we would build it. And if it worked, they could use it, and we would then create a new company. So we’ve created companies in healthcare, green energy, procurement, and workforce management. About three years ago, a regional broadband company approached us and said, “We’d like to figure out whether we can use artificial intelligence in broadband.” And that was the start. We built a prototype, which I can describe.
And it’s now been widely adopted in the industry. Two of the top 10 broadband companies in the country use it, hopefully, two more in the next couple of months. And then a number of smaller companies as well. The idea behind it is, for anyone who’s listening, when you’re thinking about the application of artificial intelligence, what you want to do is you want to look for places where a decision or even a lack of decision sometimes is being made over and over and over again on a repeated basis within your company and where that decision has a real economic impact upon your company, and it’s relatively unaided by data and analysis.
So when we worked with this company, this regional broadband provider, we looked at their operations and their business areas, their costs, and their profits. We said that it appears that the largest driver of profitability in broadband is what you happen to sell to a customer. It’s a really interesting industry because, oftentimes, companies talk about information asymmetry or information symmetry, meaning who knows what in a deal. In this case, what happens is, Brad, you go to your local ISP, and you say, “I want broadband. Or you might say, I want Wi-Fi, or I want internet.” Whatever it is that you describe it. The company knows nothing about you.
You probably don’t understand the differences between the different product offerings because, for most consumers, the difference between 200 megabits, 400 megabits, and a gigabyte is relatively opaque.
So you have a buyer and a seller both with highly limited information sets trying to cut a deal. And that’s never a good outcome because in marketing, what we’ve learned is something called the paradox of choice. That is when consumers are offered multiple choices, and they can’t really differentiate between them. They tend only to buy a price that suppresses conversion and suppresses ARPU in this industry, which as net ads have gone down recently because of interest rates and things like that. ARPU is the name of the game, so how do you drive that up? So if you go back to our definition of artificial intelligence, you as a superhuman, if you were the customer service representative on the phone call, what would you want to know?
Well, you’d probably want to know everything you could about the customer. So you’d want to know their level of wealth, which is a good correlative variable to their price sensitivity. So what will they be willing to pay for? You’d want to know things about their neighborhood. Are there lots of people who work from home in that neighborhood, and who need high-speed broadband? You’d want to know what the competitors can do because they’re the alternative to you. So what offers can they provide to that neutral consumer? And then you’d want to know some behavioral tactics around that like do they work from home? Are they a gamer? How many people are in the household? Things like that.
If you could gather all that together, then you could really make an informed decision and recommend something to the consumer. So we built a tool that does very much of that in its omnichannel, so it works in CSR channels, online, chat, and the like. What really happens is when you come onto a site, Brad, you typically give your address because you need to know that you can be serviced.
At that moment, we pull about 2,000 different variables of consumer-friendly information, so there’s no privacy violation anyway. But we now understand things like the value of your home; how many children you’re likely to have; who the competitors are to your home; and what services they can offer. Then, because we’ve done this now a million times, offer an acceptance pair or stimulus and response, we’re able to predict and say, okay, if this person lives in a $375,000 house, it has four bedrooms, they’re fully employed, they most likely have three children, there’s one DSL provider with the following array of offers. There’s one cable company or broadband company with fiber in the ground with the following array of offers.
And they tell us they work from home and watch a lot of over-the-top television but are not in fact a gamer. We can develop the ideal offer and the key reasons why that’s relevant for that consumer. Those are contextually tied to that information. So it might be … it’s one thing to say, “Oh, if you work from home, you should have high-speed broadband for video conferencing.” It’s another thing to say, “Oh, 87% of the people who work from home in your neighborhood have one gigabit. Maybe that’s what you should have.” Because keeping up with Joneses pitches are extremely relevant to people and good sales techniques.
It might say that this is actually going to save you $120 versus competitor X in your neighborhood. Or if you have an MVNO product, it can look up and determine what network that customer is on calculate the savings in real-time, and say, “Oh, if you happen to be a Verizon customer, we can save you $720 a year on your wireless bill with our new wireless product.” So all of that data actually is available. And if you can right-size that product for that consumer, it has real benefits. We’ve now done this with, as I said, two of the top 10 companies in the country and many others. The average improvements have been frankly staggering for the companies.
So the conversion rates have improved by up to 30%. The average revenue per user has increased by an average of 12%. The sale of secondary services like mesh Wi-Fi or home security has doubled. It turns out that now that we’ve had customers for over two years, the average churn has reduced by 8 to 11% at the 18th-month mark. One of the theories was, “Oh, if we’re raising ARPU, then I must be seeing increased churn.” Not if you actually rightsize the customer, you actually see decreases in churn. So all of that has led to really, really large improvements in customer lifetime value. One of our customers that’s closer to maybe about half a million subscribers, on average, saw a 38 million EBITDA improvement a year.
That’s typical. So it’s been about 20x ROI, not 20% ROI, but 20x ROI. So it’s been really interesting to work in the industry because it’s a very fragmented industry. There are a lot of customers for our product out there. They’re also trying to figure out what to do with AI. And we have a tool that every company has now recommended to someone else. So everything has been word of mouth. The reason you asked me before, is this my first conference here? It is because we just built a product three months ago that can actually go down to Greenfield. Beforehand, every one of our customers was 100,000 subs and up.
We needed to build products that could be more self-sufficient and operate in a world where a company didn’t even have an eCommerce flow. It’s really hard to put AI out there if there’s no commerce flow, and you can only do it on the CSR side. So one of the things we’ve just built is an eCommerce flow in a box, essentially. We call it Actifai Digital.
For smaller providers so that they can utilize our artificial intelligence in hopefully increasing ARPU conversion, improving churn reduction, or reducing churn, sorry. Lastly, improving the sale of things like mesh Wi-Fi and the rest. That’s really what we’ve been doing and that’s what I’ve been talking about here.
Brad Hine:
So that’s really exciting, by the way. It’s cool because I hear a lot of conversation on the expo floor and in and outside some of these conference rooms and sessions, a lot about the model around starting with a greenfield, building up, how do we build up. There’s the operation side, and there’s the business side. We get bogged down with the operation side sometimes. Sometimes we don’t want to own a lot of things on the operation side. We want to automate as much as possible. On the business side, we have this great model, but we want to make it active … I almost said Actifai, which is a great name for a company. We want to Actifai all those metrics in the business model and try to put them to work to be automated in some way, so we can see our market.
We can essentially look at all our locations, all that data that you mentioned pulling in, 2000 bits of data that a CSR or an Omni-channel application can see as you’re pulling up a location or an address for a subscriber in that area. So what I’m curious about is how you get started. I know there are probably some academic definitions for AI. Are there different types of AI you guys use that you verbalize within your business internally?
Ned Brody:
Well, so artificial intelligence, there’s no really great way to categorize the different types. If you think about numbers, you can say okay, numbers are even or odd. Or they’re prime, or they’re non-prime. Or they’re like lots of ways to look at numbers and categorize them. You can talk about artificial intelligence versus machine learning versus deep learning. Some of those definitions are … they’re not arbitrary, but they’re not really useful. So for example, the difference between deep learning and machine learning tends to be the number of nodes that the machine is using to come up with its answer.
So I think definitionally three or more nodes means it’s deep learning versus machine learning. Now the average person really doesn’t care. It doesn’t make any difference.
That’s the place where the people we hire as data scientists care greatly about that. There are things about … there are different types of predictive models like random forest versus others. And what you end up doing as a data scientist is you take the problem you’ve got, which is I want to predict this using this dataset. Then, you try different models and different data sets. And there’s something called parameterization, which is how I figure out what variables should be in a particular analysis. So one of the tools we provide to our customers, we haven’t talked about it, but it’s kind of a side one is … the very first question is this home serviceable? Because home pass files are really dirty.
So if you think about how you can use geospatial artificial intelligence to identify whether a house is serviceable or not, pretend it’s not your home’s pass file. You’ve got, let’s say 12 Main Street and 16 Main Street are serviceable. And if you looked at a map and calculated the angle between 12 to 14 and 12 to 10, you would see that’s a straight line. So the probability that the cable or fiber line runs in front of those houses is really, really high. As a data scientist to create that, what you’d say is, “I’m going to put a variable into my model called the difference in degrees between the homes.” And that’s like a decision you make as the scientists, but that’s really in the weeds stuff, right?
What’s important is more how do I actually create value? So we’ve looked at many things. We now have tools for predicting churn and stopping people from churn. Another case is upselling. So the classic problem for let’s say someone who’s overbuilding is, “I just put fiber into the ground. And my customers can now have gigabit instead of 200, but it turns out they’re not upgrading. So what do I do?”
So we just ran our test with one of our larger customers, and they took CSRs in the downtime. And they had them call out to customers who now had a better physical plant in the ground and could have gigabit. And they got a certain return on those calls. What we did is we simply looked at things like usage, data, demographics, and competitive situations. And could not come up with dynamic sales pitches. So for example, let’s say that you use three gigabits on average a day with a relatively tight standard deviation around that number, but once a month you use 80 gigabits. It’s very likely that you’re a gamer downloading the newest patch or version of whatever game you like.
When you’re calling that particular customer, and if the customer actually has enough of a disposable income to actually afford that gigabit and things like that, your pitch might have the second bullet point be, “By the way, games download in four minutes versus 27 minutes,” if that’s the right numbers. I don’t know.
So we did that, and you can imagine other scenarios that you could pitch other key selling points. And we saw an 87% increase in conversion. So it’s all about using that piece of data, and I don’t really care whether it’s a random forest model or whether it’s deep learning versus machine learning, whether it’s … the T in GPT stands for transformers, which uses attention as a type of technology. None of that really matters. It’s really about figuring out what the problem is, finding the right data, finding the right model, and then proving that you can actually get a better result. So in all of our cases, we’ve been able to AB test.
We’ve been able to say that half the CSRs get to use this, half don’t. The web traffic, half does, half doesn’t. And you can actually see the delta in improvement. It’s very different than selling some other products because, at the end of the day, you’re offering cash flow at a significant discount, right? If 20x is the number, then it’s 5%.
Brad Hine:
Awesome. Awesome. So you just talked about the upsell use case. I’m sure subscribers or non-subscribers are calling for a very different reason. So I would assume you have different package approaches that basically take a use case, plus an algorithm, plus the dataset, and you’re going to continue to develop these things. Where does it end? Is it when people come to you with problems that you just start solving from scratch?
Ned Brody:
Yeah. Every single modular feature that we’ve developed has been the result of one of the customers coming and saying, “Can you do this?”
So you need a couple of things. One is you need the data. Some of which we can bring; and some of which we need to source from the company for operational issues. We’ll explain in a second. Then you need to test it, and then you need to find a way to make it actually effective from a human perspective. So the myth of AI is that it’s going to replace human beings. There is no AI that we’ve put in the market that replaces anybody. It augments what people do.
So CSRs, we haven’t replaced any CSRs. CSRs just have gotten a lot better at their job. The typical thing we hear is, “Oh, I used to think of my CSRs as order takers. Now they’re actually selling.” That is because you’re taking the best possible thing they could say and you’re giving it to them in real time. So when you talk about modules, we’ve built ones for retention; we’ve built them for upsell; and we’ve built them for new customer acquisition. We’re starting to work on things that are now operational.
So how do you avoid truck rolls? How do you avoid incoming calls? So simple examples, if I know that I can read the IoT data, the Internet of Things data coming off of the modems or the DOCSIS data, and I know that there are a number of drops occurring for this set of houses and the phone rings from one of the houses, I could answer that with my IVR and say, “Fred, it looks like we’re having a little bit of trouble. We’ve already dispatched someone, or we’ve proactively reset your modem,” or whatever it happens to be that you want to do, that avoids that 100 and something dollar truck roll and that $10 call.
So if you think about the P and L of any broadband company, there are just tons of things in there from truck rolls to customer service sites to bad debt to how do I raise revenue, and how do I do upsell and cross sells? And you just walk your way through and triage. You start with the ones that have the biggest economic impact, and you make it all the way down eventually. So that’s what we’re doing.
Brad Hine:
That’s great. It truly is fascinating, and it’s great to see you apply this to this industry that I know has been trying to automate things all over the place just to be more efficient, save costs, and all those things. If somebody wants to find out more about Actifai or reach out to Ned Brody, how can they contact you?
Ned Brody:
Sure, you can go to the website, which is actif.ai or you can write to sales@actif.ai, and we’ll be happy to get on the phone and show you a demo. There’s some stuff on the web, obviously. But it’s one of those things that when you see it, you kind of get it. So I’ll describe a little bit of what a CSR sees.
So in a typical world, using a standard set of software of OSS/BSS, you’ll probably see a bunch of screens. And the customer service agent will be walking through. When they use the Actifai tool, either within their billing system or separately, the first thing that comes up is a map of serviceability, which shows every home that’s also a customer, the speeds that those people are using who’s not a customer, the density of that, so that they really feel comfortable that they know this is serviceable. In fact, we reduced calls to engineering by over 50% from CSRs because typically they don’t … they either get something back from their tool that says, this is not serviceable, or we’re not sure. Or they don’t believe it, primarily in rural areas.
And that’s great that you reduced that by 50%. Then you ask a couple of questions, and what pops up isn’t another form to fill in. First, we show a picture of the house of the consumer. We say this is how much that house is worth. It has four bedrooms and 1600 square feet. They get a chance to personify that person is on the other line. What CSRs do, if you’ve had the opportunity to listen to lots of calls, is they try to figure out who’s on the other end of the phone. They listen for anything they can. They listen for geographic accents. And they listen for clipped military speak.
They listen for anything that they consider a clue to who that person is. The problem is they always use those as reasons to go down market because they hear things they think of as synonymous with someone who’s not willing to be able to afford anything. No one ever says, “Oh, that person sounded really wealthy on the phone, or it sounded like they had lots of kids on the phone. Or it sounds like they’re a heavy gamer on the phone.”
All those reasons are things to have higher speeds and higher ARPUs, but people don’t perceive those from conversations. So you need to give clues to that CSR as well as tell them exactly what to offer and give them those bullet points of what they should say. The most common thing we hear from CSRs is, “I never would’ve had the guts to make that offer to that person, but the tool made me do it. And I was shocked they took it because they do.”
If you can provide a relevant offer with selling points that are actually contextually relevant to that individual and resonate with that individual, they will buy. And that’s really what we’re doing. So it’s not rocket science. There’s some hard math behind it and some data gathering and some figuring out those things. But once again, it’s what you would do as a super salesperson if you had all that information at your fingertips and could say, “Yeah, okay, now I know who this person is. Yeah, what am I going to do?”
Brad Hine:
If you were all-knowing. Yeah. So to piggyback on that, as you’re putting these packages together and refining this process, you’re also probably refining your own algorithms as they are collecting data and seeing the success rate.
Ned Brody:
Yep. So we talked about this before. The hard part of this is that you have to build a training set. So if you think about this industry and its fragmentation, there are some companies that are huge and have 10 million subscribers. Then there are companies that have down to 4- or 5,000 or 10,000 subscribers. So you need to have a lot of instances of, “I made this offer and I said this, and here’s the outcome,” in order to build the model, you need hundreds of thousands to actually get to a position where you can with confidence, predict what’s going to happen. And of course, the problem is that in that combination of stimulus and response, everyone records the response.
So everyone records what someone bought, but no one ever records what was said or what was offered. So you need to first start doing that, but if you’re selling a thousand or 2000 or 5,000, let alone 100 or 200 new orders a month, think how long it takes to get to a couple of hundred thousand instances.
Brad Hine:
Right.
Ned Brody:
You just can’t do it. So our rights with our data are that we have the right to use every instance, no matter what company it was used for, then anonymize that data and scrub it so it can’t be tied back to any one company or individual and use that as the training set for every new customer. So even if you’re a small customer with say a couple of thousand or 5,000 subs, you now get the same model as someone who’s got four million subs. And that’s kind of the democratization that can happen when you have that ability to use data across companies. So that’s really an exciting thing I think for smaller companies in this industry.
Brad Hine:
Absolutely. Like I said, talking about business models with all these folks and trying to figure out how to Actifai it.
Ned Brody:
I like that you’re using it as a verb, a lot.
Brad Hine:
So what’s next? Right now your company is very young.
Ned Brody:
Yeah, so we’ve been in business now just a little over two and a half years, I think and it’s been a great two and a half years. As I said, we’ve managed to get to and hopefully go on four of the top 10 broadband providers. The biggest thing we did was identify the need that we couldn’t service smart broadband providers, but that’s a lot of where the innovation is. So we’ve now launched as of four months ago, a product that if you’re greenfield, you can take our eCommerce platform and it will do everything that you would expect an eCommerce platform to do, including the application of all this artificial intelligence to predict what that customer will do, but make the best offer to them and the key selling points.
So that’s a big push for us is really getting a lot smaller, medium-sized, tier two and tier three, I think they’re referred to in the industry on.
Brad Hine:
Right.
Ned Brody:
We’ve been really gratified. One of the questions we always get is, well, how do you integrate into my OSS/BSS system? So kudos to CSG, which is the largest billing platform in the industry. We asked them to integrate. They took a look and they said, “Actually, we’d like to license it.” So they now are selling it to all their customers. So in the process of rolling out to all CSG customers right now, and hopefully more of those BSS and OSS platforms in the future, so we’ve got some great channel partners. We have new products, that actually can get to the smaller end of the market and really provide, I think, a necessary need or fill a necessary need, and we’re building out new modules from other things, the marketing funnel to operational pieces.
So that’s really what’s next for us, but it’s just a different part of the marketing world. You talked about how people approach their businesses and the business side as well established. And what we found is there are tons of companies out there who help you make the phones ring or people come to a website, but once people get to your CSRs or once they get to the website, there’s almost nothing.
Brad Hine:
Right, right, right. Yeah.
Ned Brody:
So that piece is actually even more important because if you can change your conversion rate by 15 to 20%, or if you can change your ARPU by 10 to 15%, it’s a whole different game in terms of your metrics.
Brad Hine:
Right.
Ned Brody:
That’s really where we’re going. We’re trying to do everything we can to help broadband companies and consumers find the right products and services in a world that is changing as fast as broadband is.
Brad Hine:
Well, I appreciate your place in the market right now and in making everybody smarter so to speak and making better decisions. We know with all the money that’s coming into the market, we all worry about whether is it going to be spent efficiently and on the right things. So I would love to get back with you in the next three, or six months and kind of check your progress, hear some more stories and figure out what you guys are doing, if there are some other challenges you guys are trying to get over the hurdle on, I would love to hear some of those stories from you and I’d love to invite you back.
Ned Brody:
I’d love to. And first of all, thank you for the opportunity and it’s been a great conversation. For anyone listening, if you have problems that are different than what we’ve described, we’d love to hear them. Anything that we think or that you think, because we’re not the experts, anything that is a problem for a broadband company that is a problem for multiple broadband companies, is something that we should be looking at. And you all know more than we do, so we’d love to hear from you. It doesn’t have to be, “Can we buy something from you?” It can be, “Can you help us solve this problem?” And we’re as curious as you are. We’d love to try to find out.
Brad Hine:
Great, great stuff. Thanks Ned. Thanks for joining today and as we end another episode of the Broadband Bunch; I want to thank everybody here at Fiber Connect. And thank you to our team here at the Broadband Bunch. And until next time, Ned. We’ll be talking to you soon, hopefully.
Ned Brody:
Thank you so much. Really appreciate the invite.
Brad Hine:
Great. We’ll talk soon. Bye-bye.
