Very much the accidental telecom executive, Gerry has started and led multiple ISPs both regionally and nationally. His broadband adventures started out of necessity after leaving Manhattan in 2011 for remote rural living on a Pacific Northwest island. Working with his local electric co-op, he led Rock Island Communications which has been one of the most successful cooperative-led broadband initiatives.
He moved to T-Mobile in 2018 and led the national development and deployment of its Fixed Wireless Access line of business. Prior to making his mark in the telecom sector, Gerry led global businesses in the financial industry, where he developed his passion for bringing meticulous research and a unique, boundary-pushing approach to data analysis.
Drew Clark: And we're live. Good afternoon. Welcome to the Broadband Communities Ask Me Anything series. I'm Drew Clark, editor and publisher of Broadband Breakfast. Super excited to welcome our guest today, Gerry Lawlor, who is the co-founder and CEO of Hexvarium. And we're gonna learn about who Hexvarium is, what Hexvarium does. And we're gonna see a demo, which is exciting. Demo Day here on the near kickoff of 4th of July weekend. Is it the start of a four day weekend? Not entirely clear, but 4th of July is just around the corner. Gerry, welcome to Ask Me Anything. So glad to have you on the program.
Gerry Lawlor: Thank you so much for having me, Drew. It's a pleasure to be here.
Drew: So Gerry, I have to start off with one of my basic dumb questions. And as a journalist, I get to ask dumb questions. It's important to ask dumb questions. What is a hexagon and why are hexagons important for mapping or broadband mapping, Gerry?
Gerry: So the easiest way to think about hexagons and there's a funny video I posted up in the chat if anybody wants to go look at it, it's a very, quite an entertaining but a useful video called Hexagons are the Bestagons. And it is effectively the most efficient way to geospatially organize areas that also maintain a much more natural relationship to each other. Like if you think about the world as a grid, the actual real world is rarely laid out in such a perfect way. Whereas hexagons can maintain as much of a linear relationship as they can, a circular relationship to its neighbors. So we use hexagons because, again, it's really efficient in covering space, but it's also very useful for us as we look at the optimal path in which to build. So we spend a lot of time looking at, okay, we see certain things in this area, but it's neighbor also matters, right? And you will see this in the demo as we start, as it kind of makes sense, but it's mathematical relationships is why we use hexagons effectively.
Drew: And what is it that makes a hexagon better than an octagon? I mean, a similarly regular polygon.
Gerry: What makes a hexagon more efficient is, I mean, you can push more to octagon shapes and you're not really gaining any benefit there. But the easiest is to compare between a triangle versus a square versus a hexagon. Triangles and squares have very poor neighboring relationships. And again, I've posted some links that people can jump in and see sort of demonstrations of why a grid is too simplistic. Triangles are too haphazard. They don't have great, they've awful neighboring relationships, particularly north-south as much as east-west whereas hexagons are much better in their relationship to their neighbors.
Drew: Well, I'll look forward to following that link.
Drew: I'm assuming Hexvarium is a play on hexagon. What does the varium in Hexvarium mean?
Gerry: It's a play on three things actually, hex for hexagon. Alvarium is the Latin for honeycomb or beehive. So they are, on our minds, the original builders of hexagons, given what bees do inside a beehive. And then the var, the V-A-R, the varium portion of the word, we derived 'cause we use a theory that is used in the financial markets called value at risk or var, right? Where we're constantly trying to compare the value of one area to another area and what's at risk when we start trying to make judgements, right? So it's kind of a mash of a few concepts put together and how we came up with Hexvarium.
Drew: Well, exciting. Looking forward to this demo, which we'll get to just now, but why don't you give us a two to three minute intro of how you got into this space. And obviously we had a piece about you, laying out a few of these details. But where are you coming to us from Gerry? That's an important detail...
Gerry: Yeah. [chuckle]
Drew: And what kind of got you clued into the vital importance of broadband?
Gerry: Yeah, I'm really lucky and I live in a very beautiful place called Orcas Island in Washington state about 100 miles north of Seattle. It's a chain, a part of a chain of islands called the San Juan Islands right up on the border actually closer to Vancouver than Seattle itself. Nearly 13 years ago, I moved here from New York City and with my then, my new wife, where my wife grew up in the state of Washington. So she knew this area. And 13 years ago we decided to do a 180 on our lives in New York. We were both in the financial industry working at the likes of Goldman and Bank of America and all these different institutions. And we moved here with no plan and decided to kind of reinvent ourselves. I was operating, I was the COO of a risk management technology company that was used in the financial markets, that was headquartered in Australia at the time, 2011 and '12. And I had really bad internet, I was dying on a couple of DSL connections trying to bond them together to get maybe a meg and a half of capacity. It was painful. So I was living very much the COVID life long before COVID in many ways. And it was really frustrating.
Gerry: And I found out my local electric utility, Orcas Power and Light, was building some fiber about a mile away from my house to connect a substation. And my current, my still CTO chap called Adam Kelly, who we worked together back then, happened to see, like recognize the fiber. I wouldn't have been able to tell a fiber cable from an electric cable, it wouldn't have stood out to me. But he is like, "They're building fiber down there. You gotta figure out how to get into that." So I started just digging into the subject and trying to understand like, how could I connect and what could I do just for myself, right? I just want a better internet for myself. And the utility came back with this mission, with this goal of saying, well, it'll cost you 16,000 to dig it to your home, we need a 10-year commitment to $300 a month for service.
Gerry: And I'm justifying it in my head. I mean, when you live in Manhattan, you tend to pay for things that most other people think is insane. So I'm trying to rationalize this. And then they were like, "Oh, by the way, it's going to be dark. You need to figure out how to light it yourself." And that's when I was like, "What!" And I started digging into that, and I found out I would have to buy a wholesale commercial circuit from Seattle that would cost me $6500, $7000 a month or something. I was like, okay, something's wrong here. I've gotta, this is not gonna be sustainable. So I started just getting involved locally in the issue. And the issue was growing. And we had our incumbent legacy telco was frankly failing, and there were some serious issues and that just led a local movement that I ended up leading.
Gerry: We created a for-profit subsidiary of our nonprofit electric utility called Rock Island. Bought a local ISP, bought spectrum from Paul Allen's Vulcan Ventures, entered into a very unique deal with T-Mobile to build out a hybrid fixed wireless fiber network around these 20 islands. I mean, we were putting in submarine cable and everything in between. And the result of that ended up being so successful that T-Mobile asked me to come help them lead and build out their national broadband business. So the fixed wireless business that you see today, the genesis of it was here in this community. And I led that, build out and run that team for a few years. And then I left T-Mobile in 2020 to start Hexvarium.
Drew: Well, that is a perfect intro. Let's see your demo, Gerry. Talk to us through the process. I will mostly just sit quiet while you do it, and then we'll come back and continue our conversation.
Gerry: Yeah. Let me make sure I do this correctly. Ask this one. No, where's my browser going? Oh, I gotta do this correctly. Sorry guys. Toggle screen sharing window front. Okay. So let me know when that map pulls up. I'm...
Drew: Yep. It's pulling up now.
Gerry: Okay. So what we do, I mean the gen... There's two ways and I'll just, before I dig into all the details and start jumping around the screens and whatnot, can I, and is that clear to everybody?
Drew: Clear as it's gonna be. It looks good. Thank you, Gerry.
Gerry: Okay. So what we do is we've sort of two entrants into our process. So the goal with Hexvarium is, it's not just an analysis tool upfront. And it's not an engineering tool, it's not a marketing tool. In many ways what we... The way we look at it is we're trying to understand very quickly multiple factors of a given geography to get a quick sense, is there a good business case? And/or where are the challenging areas, right? And really get a very quick understanding of three key factors. Who and what is in a geography? What is it that I would need to build to support that geography? And what competition am I up against in that geography? And we can look at areas as big as states, entire states. I mean, nobody's gonna go build necessarily the entire state, but we've also ways that we can look at massive geographies to find opportunities at even higher levels than the detail I'm gonna show you here today.
Gerry: But that results in a very fast process of feasibility that takes most people quite some time. We can calculate out 20 years worth of risk in less than five business days, is our usual turnaround of this data. So what we start out with is very detailed information about every home and every business in the geography. So all the blue dots on this map is every individual address. All the multicolour dots are where all the businesses are and you can kind of get a sense. You can see the businesses scattered along main thoroughfares as you would...
Drew: Can you zoom in a little bit more, Gerry.
Drew: Just to give us a view of close up.
Gerry: And just to make sure Drew, can you see my mouse move over here? Because the barrier is different than my browser?
Drew: Yes. Yes.
Gerry: Okay, good.
Drew: We can see those tabs on the left. Thank you.
Gerry: Okay. Good. 'Cause the line is not in place. So I can zoom in and I know that this is Advanced Auto Parts and they're right beside Popeye's, Louisiana Chicken, and there's a plant in Scamby Plant next door and then a doctor's office and whatnot and jewellers. So we've a lot of information not only about where businesses are, but who they are, what they are, what types of businesses they are, and the industries they're in. We have a lot of data that gets into white collar, blue collar mix. All of these things start to help us inform demand and need, given the nature of different businesses. Like the analogy I always use, a Starbucks is very different than a local cafe. They're both coffee shops, but there's different risk elements to us when it comes to trying to support them within the broadband industry. Mainly, most Starbucks don't get local buying authority, whereas a local coffee shop does. And you've gotta make decisions when and how to sell to them. And we have the same on every residential location.
Gerry: We get into not only where every home is, but we also have a lot of data that tells us who lives in every home. So we know that this home is owned by Catherine Griffin. It is her primary residence, and her neighbor are Leonard and Ilian Uben, and they're a married couple. The house was built in '77, and it's a 900 square foot home. We know who owns every commercial property, every residential property, and a lot of detail behind that. So that also helps us in understanding family demographics and age demographics and income demographics. When we add a lot of that data together, we get some very precise who and what is in a given geography, right? And we use all of this data throughout our process. We've also developed a fiber design, a logical fiber design to every home and business in the country. So what we do at this stage of our analysis is establish a logical buildable path to understand the distances required and a relatively accurate, quick understanding of the mix of above ground, below ground within a given geography, right?
Gerry: Which are really the big drivers in cost estimating and trying to understand what have I got to build? Where am I going to be building it? Am I looking at very small drop densities? Am I looking at how much we even get into, how many state roads versus federal roads versus county roads am I building on? And that there's levels of risk of what I have to build and the makeup of it in a given geography. And we also have a competitive analysis that we perform in everywhere. So every hexbin here, we score on a 10-point scale of competition where we look at the services being offered. We look at it all through, like everybody else, through Form 77 and all the other data sources. So we get an understanding of who and what providers are in a given area, what technologies they're using. We look at pricing data as to what service, product pricing they're using between both promotional and on an ongoing basis.
Gerry: And then we look at actual speed tests within that hexbin to compare what's being marketed to somebody and being offered to a consumer versus what they're actually achieving. And we distill all of that on various weighted methodologies on a 10-point scale, 10 being better for a new entrant versus one being more competitive at better prices, achieving higher speeds and whatnot. So it gives us a very quick way to understand the differences between one neighborhood and another in a community. To give you an idea, each hexbin at this level of our analysis is 0.2 square miles, right? So the typical, you're talking about a few hundred residences per hexbin is the level of detail that we're getting into. Ranges in the most densest areas, maybe above a thousand, usually we're looking at hexbins below 500 locations, right? And then all of that data we use to distill down into what we call our hexval model, which is this, and each hexbin is its own set of 20-year financial analysis.
Gerry: So we are looking at the entirety of the CapEx expected to be deployed inside that hexbin, a level of take rate that we expect to be able to derive from that hexbin. We look at income levels, what prices we think you can achieve, the ARPU that you can achieve, and all of that gets boiled down into its own financial model that we can look at payback periods and customer lifetime value of each individual hexbin, and then how they are in aggregate and together, right? So in doing all of that, it then allows us to start asking questions about the area. And over on the right-hand side here, you see all of the numbers to assume that, okay, if I was to build this entire geography, it's Santa Rosa County in Florida, Northern Florida, if I was to build this entire county and we average 36% take rate across, which is what we estimate across the entire geography, I was to build 100% of the backbone, 100% of the middle mile, and 36% of the drops of the last miles, I would approximately be spending $164 million per 79,500 odd locations.
Gerry: And in this area, we see a 50/50 split between above ground and below ground, but we see that as a negative CLV over 20 years. So then we start to ask questions. Okay, well, some people wanna know if I'm just a pure investor, I wanna know just maximum return. I'm a private investor and I wanna know optimal return in the area. We start looking at, so we start filtering. Okay, so where is, and you'll see as I start to filter out, we're never just removing the bad areas. We're always maintaining a logical buildable network because in many, like you're gonna, no matter what fiber network one builds, you have to build through bad areas per se to get to good areas in many cases. You cannot chart cost the areas to get to more dense areas and whatnot.
Gerry: So we're always maintaining a buildable path through this analysis, right? Which gets back to the use of hexagons and the relationships and the sort of the logistic science that we've applied in that. So we get to ask, okay, let me find optimal return. Okay, how far do I have to go? So as you can see here, as I'm moving down, you can kinda see the filtering coming down that road as I make my way back into these denser communities. And now I'm starting to see, okay, I've dropped from 164 million to 81 million and I've significantly dropped my average cost per passing and I'm starting to see higher CLV levels. Yet I've still got the majority of the passings within the build footprint. So it allows you to quickly see the extremes.
Gerry: And I know here the maximum return is right around there. It's right around 11.3 million in customer lifetime value is what we estimate out of this area. And that gives... So it gives people the immediate scale of, well, where's the optimal, where are the challenging areas? How can I look at edge out strategies? If I was to build this, we have ways I can pull up other demonstrations where we can start adding in fixed wireless tower locations to this design to see that if I build this much fiber and I start adding certain wireless locations to it, who can I pick up logically within coverage areas of that fiber network using a variety of different spectrums and technologies? The standard we look at, which is what most folks are using is the Tarana technology.
Gerry: And we can apply that spec immediately to this method and quickly see, okay, if I deploy here, I can pick up X customers. So it allows us to very quickly start asking really smart questions about an area that then as you move through our process, we start getting into more depth around what type of network should you be building? Should you be building centralized split versus distributed split? Should you be building the same design in every hexbin, both sides, like H patterns, I patterns or U patterns of fiber? How do you not strand capital in areas if you're not as confident in your take rate in one hexbin versus another? These are all the types of questions we now start to ask in our process as we go through it. And then ultimately the goal of the platform is it stays involved even when an individual starts putting money to work.
Gerry: We connect into the various project management systems that folks use and all the CRMs that people use, the Salesforces of the world and whatnot. And we're constantly watching how good is your build doing. And is it aligning with where marketing and engagement is coming from and how is that stacking up to the assumptions that we made? It's more akin to the day when I sat on trading desks, doing trades in the financial markets. It's not like I just locked my assumptions in that moment in time when I did a trade. If I'm looking up my portfolio of risk, it's constantly changing. I made a decision to transact, but then I've gotta constantly assess that and reassess that risk to make adjustments. Effectively the same concept is applied here.
Drew: Gerry, keep the screen open for a few minutes.
Drew: I've got two or three questions just on this. Do you wanna show us another part of the demo too, though?
Gerry: I mean, there's tons more stuff. I mean, we've done digital divide analysis across the entire country. I can dig into that, which gives us some pretty useful insights. We've got ways of being able to filter massive areas to find opportunities, all sorts of different things. This is one of many capabilities.
Drew: Very good. Perhaps we could come back to something a little later.
Drew: But let me just ask a couple questions here. You mentioned the 36% take rate figure. Now, I don't see that on the right. You answered my question I had about what is on the right here. And great, great discussion of this, but can I tweak that take rate assumption and where do I do that?
Gerry: We, so in our backend, yes, in our model, we can go in and start setting different take rate assumptions. What we do in our next phase of our analysis... So we call this our detailed market analysis in what we call our go to market analysis. We derive multiple scenarios. One thing we don't like to do is peanut butter spread a take rate assumption across an entire geography. A lot of people we find walk into their models and their methodology, and they've got one ARPU number and one take rate number that they apply equally. We think that's quite problematic. We can set up a scenario that shows people what those numbers look like, and how it can get applied to the same geography, but we can also very quickly prove why we think you can only get 25% take rate in one hexbin and 45% in a different hexbin.
Drew: All right. Now, back to our hexagons. What is the hexbin versus the hexagon?
Gerry: The bin is literally just the... The hexagon is just the, like a flat shape on the geography. And the hexbin is where we start adding content to an understanding. So basically the individual height of each of these hexbins right here that are represented by different data and we use colors to represent their differences as well.
Drew: So, how do the geography space relate to the... I mean, like obviously a cellular network is built, when we look at maps of those, they can get smaller as we get more and more points, nodes in the network. Is this analysis capable of being shrunk to smaller hexbins?
Gerry: Yes. If people are familiar and they can do a quick Google, and again, I posted it in some of the chat, there's a standard that came out of Uber called H3. It ranges, there's... I forget the exact number of hexbin structure that they have, but I think they go down to two, three meter hexbins globally, and there's a table of various focus. We use, if we're doing statewide or multi-state analysis, we use what's called the H7 level. This is H8, and then we go to H9, which is 0.04 square mile size hexbins. So when you think about, this is 0.2 square miles, we go down to 0.04, and then basically at 0.04, you're locking at individual homes and businesses beyond that, you're down to that house, that address or that business and that location.
Gerry: At that point.
Drew: Well, this is awesome. Let's go to some of the questions we've got piled up and let's go ahead and turn our shared screen off at least for now so we can see a little clearer. So, here's a question that we've got from Nat Nguyen. So you have figured out the where, the how, and now you're adding the third dimension, the when to the equation. What are the main challenges that you are facing when adding risk timing to this equation and just continuing, how do you reconcile the spatial and temporal aspects of data sets to generate geospatial and time series analytics? There's one more question there, but let's start with those two. Okay?
Gerry: Yeah, that's a pretty meaty question for sure. And I did type up an answer to it, and a lot of it is, again, it's why we... And I hate to use the term, but it's the best one that I can come up with. But when we profile an area, the very first thing we start to look at is do we have an understanding of demand or need based upon the profile of a given location? Is it an older, younger? Like the obvious. We get into the fact that we have data on household count that tells us this is an elderly couple versus a younger family, we start making demand assumptions and timing assumptions around them. Now, it's not to say that the elder couple won't buy a fiber service, but they'll probably take longer making the decision.
Gerry: So when we start seeing a makeup of a hexbin made up of different demographic factors, we apply time values to each of those so that then in turn, if we're applying the individual metrics to every home, we then aggregate to the hexbin which then aggregates to the entire geography. That's one way that we just, we look at who is going to be your ideal customers in year one and who's in year two and who's in year three. It's not to say you're not gonna get them, it's more when are you going to get them? We see a lot of operators make mistakes looking at neighborhoods because they're wealthier and they're younger and they've got a higher demand. But then we start looking at competitive factors.
Gerry: And it may not be, you know, broadband desperation is easy to find, but broadband apathy is harder to find, and frankly, a lot of people are lazy in making change. So again, we look at, depending on the individual competitive analysis we see in that area, we stretch out timing different than other hexbins where we see different competitive factors, and all of this we're constantly kind of boiling together to assess not only, here's what you have to do to be successful but when should you best do it. And the harder part of it is then when we apply back building logical path, because again, you can't build islands, you can't just cherry pick the perfect places in year one, you have to make some tough decisions and build in areas that may not be ideal to get to areas that feel ideal in year one and so on.
Drew: Gerry, it was more than 20 years ago that Scott McNealy then CEO of Sun Microsystems had the phrase, you have no privacy, get over it. And I was actually present when he said that, and look, we all know this data, these public records, there's so many data sets is available, and yet when you just like casually toss off, hey, here's this random person here, this random neighbor, and all of the assumptions and inferences you get about that, let me just ask that question quickly about, like what is the state of kind of "privacy" in this world we're in where so much data is available? I wouldn't expect that people in the geospatial industry not to use what's available, but anyway, do you have any thoughts on that?
Gerry: I mean, while I've purchased and I buy a lot of data and we use a lot of public data, we've about, I think I'm growing 350 different data sets that we bring into this model, everything that I've procured is, like the data that I just showed you is data that you could go find at your local municipality because it's all public property information and...
Drew: What are some of those primary data types and sources that you use in your mapping and risk analysis?
Gerry: We're across the board... We naturally have everything from the census, everything from the ACA, the American Communities data sources. We have data from Ookla, Opensignal, data from Precisely, Lightbox, CostQuest. I've got geological data from the US government. I've got aerial data from a variety of different sources where we're starting, 'cause I see one of the questions popped up about above ground and below ground, we use a lot of street view data to get an opinion around what does the infrastructure look like in a given area. And we're starting to experiment with...
Drew: And that's your assumption, that's your prediction of this percentage aerial, this percent underground. That's not based on any one else's. That's your kind of...
Gerry: We take in... We're, that is a really tricky one to really get accurate. And we're part... We're working with a couple of new entities that are focused solely on that effort of can we identify very clearly, not only is there a pole there, but there's even entities that are trying to get ideas of what would make ready look like on that pole versus this pole? And there is a push in that direction where you will actually get down to, I know where infrastructure is and I can then put a risk assessment on that infrastructure. Because in many ways, again, there's this assumption that, oh, above ground is the best way to go. But the reality of it is it depends on who owns it. It depends on the make ready. And in some instances, both the time and cost of that decision you might be better off going below ground even though it's more expensive per foot. And it's a matter of being able to stress test that and understand it in our mind.
Gerry: And we look at every, we have a lot of data that tells us the exact width of the right of way. That we know, okay, this is the space you can be in, and then what is in that space? Because again, going below ground is always a speed element. If I can, it might cost me more above because I have to deal with a utility or a different owner. But am I gonna be better off because there's just so much other stuff, water, gas, below that's gonna slow me down, right?
Drew: Right. Gerry, we had Jim Stegeman, the CEO of CostQuest on Ask Me Anything about six weeks ago, and I think you, and he and some others from Ready were on a panel at Broadband Communities about this subject of mapping. And one of the things Stegeman said in the Ask Me Anything was, okay, now we did the location data for the FCC, the fabric. But the data about providers is not ours. And so I just wonder like, how do you kinda think about this disaggregation of the data, the data about providers, the data about risk, the data that you're pulling in from other sources. Like what's the way you think of your data sources as inputs and what's the output you're providing to customers of Hexvarium, Gerry?
Gerry: The inputs never stop changing. We keep adding to it. There is no panacea to this data problem. We don't see any one data source as the most important to us. It's really our methodology is how we've brought that data together and how we can make a lot of very accurate assumptions in very small geographies and understanding. The biggest lift in our regard is the fact that we have a very strong opinion and understanding of very small areas. And then that allows you to make better assumptions on that area versus the blending that goes on, which basically removes the true understanding of the risk you're taking in building out a massive infrastructure that's always gonna be capitally intensive.
Gerry: And there's always a miss. Like generally I can't stand in the industry where people use, it's all about just building the cheapest feet past the most amount of people. I think that's like, hopefully we are moving beyond that. Now that we're getting into areas where density, like when you spend enough time looking at this country, it's not a dense country and there's nearly as many people living in edge suburban to rural America as there is living in metropolitan urban America. And even urban America in this country is not a very dense place in the scheme of things relative to other cities around the world. So really understanding the geography and then applying, and I think the biggest part missing in everybody's model is an understanding of the customer. They tend to see it as an engineering problem or a competitive problem. Like people can figure out what competition I'm going up against and people can figure out, I can build something in a given geography and make it work mathematically or make it work physically. But that doesn't guarantee you're gonna be successful in the returns that you need.
Drew: Right. We've got a boatload of great questions live in addition to those that were posted on the page. I'm gonna ask one from Gary Lee Fry. Have you looked at the state broadband maps to see what the differences are between what current ISP states are speeds versus what the real or actual speeds? This is a huge issue and this issue I've been kind of talking about, arguing about for 15 years, more than 15 years. And again, it's just, it's sickening the way the FCC has totally put the burden of proof on the customer or the state, to say you do not offer broadband at this nonsense speed of 100 by 20. I'm mean, there's so few that are, or anyway, now I'm on a rant, but Gerry, talk to us about these, the actual versus the promise speeds and the state broadband maps and how they're doing on this.
Gerry: Yeah. I mean, like everybody suspects, but we've been able to prove to a decent degree the problem is far worse than is ever led on, and I'm pulling something up here to be able to just show you...
Drew: Sure. Go ahead.
Gerry: Really quick. Let me, and even on the definition of 25/3 and 120, so let me reshare my screen here and get back to... Share Chrome tab.
Drew: While you're doing that, I'll just finish Gerry, Gary's question here. He said this came up the first round and several state grants with federal funds, ISPs gave the max speed, but with the real world use, the speed dropped to where the area would qualify as unserved or underserved, but the ISP obviously wants to keep them out. So talk to us now as you drill into the...
Gerry: So what you're seeing here is actual speed test data achieved green bean, anyone that achieved above 25/3 and red, where they were not achieving above 25/3. I've zoomed into the Midwest. And you can very quickly see, okay, 25/3 seems achievable between Atlanta and Chicago and the Midwest of the country. But when you start bringing, am I achieving 100 max?
Drew: What did you just do? What did you just do as you cross?
Gerry: All I'm doing is I'm looking across two years worth of every speed test that Ookla has ever received.
Gerry: It's a massive amount of data and we've distilled it down to individual hexbins and where that speed test came from down to, again, 0.02 square miles, and you start to see are people actually getting it or not? So this is every... So you can see that the delta between, okay, people are achieving 25/3, how big is the problem at 120 if that's going to the new definition? Yet we all are in the industry that we should be even way beyond 120 as a standard. So you just see the size and scale of the problem. This is not... Look, this is why we believe and I've dug deeper, we can, and I've got some proprietary stuff I can't show here right now 'cause it's way too controversial, but we can actually get down to what people are saying versus what they're achieving quite accurately. And there are immense gaps.
Drew: All right. Let's go back to the questions and un-share the screen there. This is from Manny Vernon. I hope I got that right. Gerry, can a customer bring in and overlay their own data sets like the customer's demographics, cost to acquire, cost to serve and retain so that they can estimate CLV more accurately? Unpack that a little bit and explain that?
Gerry: Yeah. We, as part of our go-to market analysis, so when we perform this detailed market analysis for our customers, the next step of our process is to engage with them and learn. Okay. In many cases, they already have some existing network or they have an understanding or a set of assumptions about a community. And the reason we don't want any of that upfront is we wanna establish what we believe is just a clean baseline about the geography. And then in our next phase of our process, we start comparing. And that's where we get into scenario analysis, right? And you can bring your own data, we can consume that data and we start comparing it in various scenarios, right? Again, I would say our data is really no different or our output is no different.
Gerry: They're just good assumptions. And assumptions are just that, [laughter] right? They're just assumptions and we can make them in what we believe to be as good as possible, but at the end of the day, it's never guaranteed. And that's why we want to continue to analyze an area even after you go to market. Because so much of it is about, can I make sure a little problem doesn't turn into a big problem? Can I see where demand is coming from fast enough to move my capital there rather than continuing to plow capital down a path that isn't achieving the return that we expected today? Not to say you won't go back and build it, but how do you time where you should be putting your capital and when, and you need to be on top of that. We wanna get that down to at least a seven day process where I know I'm putting capital out and I can receive revenue in within seven days. That's our goal.
Drew: Okay. Sorry, I just, I'm not necessarily following you. Could you just explain that seven days, what do you mean seven days? For you to do what in seven days?
Gerry: Yeah, you're spending money, you're out building, right? The one we...
Drew: Yeah, I'm an ISP, I'm building...
Gerry: Yeah, you're an ISP. You've decided this geography, and we've done all our analysis. You feel really good. You start putting money to work, you start going out and constructing, generally the timeframe from which people are spending money to the time they actually start connecting customers is way too long. It's north of, in many cases six to 12 months in many cases. The efficiency we wanna bring with this capability is, can we squeeze that time down?
Gerry: Right? And ideally get it down to about seven days.
Drew: So like stepping back at this broadband tsunami we're experiencing, right? The federal government, President Biden himself with Vice President Harris and Secretary Raimondo made the announcement four days ago of the amounts that each state is gonna receive of the $42.5 billion in the BEAD program, Broadband Equity Access and Deployment. And this kicks off the process the next six months, the states are gonna be actively doing their plans with their money. Now, okay, so how do you see this playing out? What is the role that ISPs are gonna play? What's the role the state broadband officers are gonna play? And what do you see kind of, what you are doing fitting into that? And what's the results you hope to get out of this?
Gerry: How I see it playing out is, I think like most people, there's gonna be 50 unique scenarios that everybody's gonna struggle to get their head around given how each broadband office or each state sets up their process, some states are advanced, others are way behind, and I think we will see everything in between. I'm worried that frankly there's gonna be a lot of capital given out to entities that don't deliver because they don't walk in with a clear understanding of what challenge is in front of them, both the state as well as the individual operators that start receiving funds. I believe when we've looked through our data and our analysis, it's easier to identify the true unserved. The challenge with the true unserved is gonna be getting to them and the economic argument that I need to make as an investor and operator in order to make some prior investment to set myself up to take advantage of the unserved dollars that I receive, I think that's gonna be one challenge. And then the other challenge more on the underserved, I think is gonna be far greater that because it's so Swiss cheese throughout the country and in pockets, you're gonna have to come up with a lot more private capital to surround an area that you might get public capital.
Drew: Right. Right.
Gerry: Right? And I think the mix of that is gonna complicate people so much that they end up avoiding a lot of areas because they have to make an amount of private investment to augment this public investment. And I see that as quite challenging unless they can identify it, right?
Drew: This is a great segue to Jace Wilson's question about, talk about the origin, which you started a little bit about your history, but speak a little bit about what led you to found Hexvarium and what's your vision for the company? What's the long term, particularly in light of this broadband build out?
Gerry: So we started this out with the original view of actually being like many other over... We operate, we own and operate network in an area in California, south of San Francisco, in the Peninsula. And myself and my other founder, Mike Farmwald, similar to me, just started putting his own money to work and built out a fiber network in his community. And when we met, we created this bigger vision and our original plan was let's go to the market and raise money, and use this technology that we were designing and starting to develop at that time and use it as a proprietary tool to express the investment thesis of a fund. Where, let's say we raised half a billion or a billion dollars, which, you put that money to work maybe over a five to seven year period, you're connecting a million locations around the country in various geographies that we could identify opportunity and put that capital to work and we just go build network. 'Cause we believe in, not only the need, but the economic outcomes of making smart decisions. When we went to market to raise that fund and to talk to the infrastructure investment community, when they saw the capability, they were like, "The entire industry needs this capability. Because we as sponsors and investors of operators are not getting the feedback loop that we want to see. We get a lot of very baked assumptions about what our money is going to do, and then we see a very different result and we're very frustrated by that."
Gerry: "People are coming in and saying they can build for $X below per passing and we fund on that and we fund on how capable we think the management team is, but the reality is very few are in control of what actually plays out over time and those returns are not fitting our investment thesis."
Drew: This seems to go to the risk point that you've mentioned...
Drew: Over and over again, and we have a question from Tara Whipple about what measures are the most important in determining the risk of a broadband investment? You might have talked already about this, but...
Gerry: Yeah, I touched upon it. Really to us it's the combination of people, the thing you gotta build and the competition you're up against, and all of that is influenced heavily by the geography, the realities of the terrain, is heavily influenced by the type of people, the types of businesses, the growth, the expected growth that's gonna happen in communities. And then you're looking at competitive factors and then how to stress test that over time. Because even if I, in that sample that I showed, $164 million investment to build out that entire county for the sake of argument, you're not putting $164 million to work on day one. It's an investment that is only going to change over time. So how you manage and stress test making that investment has to be able to keep up with the inherent change that's going to occur. And in this cycle of what we do, we've all witnessed most of us here on this call who deal with any customer when it comes to delivering services knows that if I give you a better broadband connection today, within two years, you're gonna need something double in capacity.
Gerry: And I mean, it never, so it's like you're constantly just trying to chase, and I think people like, they look at this in two locked moments in time where they make assumptions today and they basically hold on to say, well, it's gotta maintain forever. And that's just not the case.
Drew: What's the role of co-ops in this space? And talk a little bit about whether co-ops were key to what you did on Orcas Island.
Gerry: Absolutely. I mean, that's where it all began for me was with my local co-op, [laughter] with Orcas Power and look, I generally think, one, I'm very proud of what co-ops have done across the board. Like I think they've really, here's the one thing I've always said about co-ops. I love how the fact they don't call anybody a customer, they're all members and they treat everybody like a member. And I think that's a really important concept that should be brought into this more broadly, but even beyond co-ops, I think electric utilities are a critical part of solving this problem nationally. And that's where I always go back to, no one balance sheet can bear the burden of this problem. Not even the federal balance sheet can bear the burden of this problem. $46 billion is not going to solve this problem in my mind. This is a shared risk, requirement, especially as you get out of major metropolitan areas.
Gerry: And it's not just money, it's also existing assets, like a utility may not have anything to necessarily do with the deployment of broadband or the management of customers on that broadband network, but they can monetize their assets in better ways. And I think that changes the risk profile. Same with municipalities and other people that have infrastructure that, you know, state bodies. Like I've spent a fair bit of time with the Washington State apparatus looking at both the DOT and the broadband office and the Department of Commerce and I'm like, the DOT has massive long haul cable capabilities down interstates connecting major areas. Why can't we open that up and make an avail of it? It's an asset you can bring to the table. And that's... And to me that's got a value that you have to be able to monitor, you have to be able to calculate within a multiple pool of various investment types. And it's not just dollars from, we're too siloed. And that's what frustrates me. And unless money starts to mix together, I think the problem is...
Drew: Are you a financial player? I mean, do you see yourself playing in that space? What opportunities does Hexvarium itself offer for broadband co-ops?
Gerry: Yeah. Our goal is to see our platform analyze 50 to 60 million locations around the country, and we charge on a per location basis. That's the size of the opportunity that we see. Again, our goal is to manage the money. We're not gonna be better than some engineering firm. We're not necessarily gonna be better than some piece of consultancy that you receive. I actually think we are better in many ways, but that's a different point. [chuckle] But generally our focus is on, are you putting money to work whatever the makeup of money is in a given geography? And that can be made up of many, many different scenarios. But are you putting that money to work in the right way over time? That's the the view that we wanna stay on, and that's the area that we remain focused on.
Drew: Gerry, we've got a lot of geeks on this call and I'm getting some reaction and encouragement to do another demo on the digital divide. So let's take another five minutes here.
Drew: Go ahead and share your screen on demo two. Okay?
Gerry: Yeah, this is, let me, where's my San Mateo? Yeah. So this is a digital divide analysis of two counties in California, San Mateo County and Santa Clara counties, basically everywhere south of San Francisco down to San Jose. Two probably wealthier of the wealthier counties in the country. What you see is the higher, the darker the hexbin the more BEAD requirement is necessary effectively. So you can tell on these fringes out here and out here, these are very much falling into the unserved areas. And they're, as I said, I think they're easy to identify. Where you start to see other more challenging is where you get underserved are all of these pockets within dense geographies. And what we've calculated out is in each hexbin, again, we like to score things, we've created a scoring method that calculates out the proximity of every home to a public institution, like a library, the current competition, the things that that customer can receive today, the prices in which they can receive those services, and then the amount of SNAP or social welfare utilization inside that hexbin.
Gerry: And that allows us to kinda understand the need as it relates to the digital divide or the BEAD process, right? And we can calculate that out accurately. And it's scenarios like this where I see, like we did our own, we operate network in this geography and here in the middle, and we've got poor neighborhoods adjacent to what is probably one of the wealthiest neighborhoods in the world. And right next door is a pretty underprivileged neighborhood, yet we wouldn't be able to get BEAD dollars for that area. So how do you make decisions in trying to understand where that money can be put to work and where it can't be? And, the nature of the problem in every area. So, it's a lot more nuance than I think people understand. And I think that's gonna create the challenge for a lot of folks.
Drew: So, a couple more quick questions and then we'll get to the close out here. Manny Vernon asks, right now most BEAD dollars is going to large national ISPs who have proven experience. I'm gonna contest that question, but...
Gerry: Yeah, so not accurate.
Drew: How can Hexvarium help local co-ops to get BEAD dollars? So anyway, what's your reaction to this point?
Gerry: Yeah, look, we can help. We don't, again, we work with some of the big ISPs in the country as much as some of the smallest and down to even some tiny municipalities, and I don't think there's, I don't think it's gonna be as black and white as people feel or as people assume. I think the big guys are generally not going to pivot in the way people assume they will. I still think they will drive the higher level metrics that they demand, that will continue to push them into areas where a lot of BEAD dollars are not going to be applicable, and I think they're gonna struggle as they've always struggled. BEAD money is not new. Like receiving such funds is not a new thing, frankly, that's, I ask myself that question every day, why do we continue to have this problem?
Gerry: The industry has spent trillions of dollars over the past, God knows how many years, a lot of it coming from UFC funds, to CAF funds, to the American Reinvestment Act, like billions upon billions of dollars. Why do we still have the problem? It's as much an operating problem as it is a capital problem. And those two things are not very good at finding a home together. And that's why I think a lot of other entities need to come to the table, such as utilities, such as municipalities and other operators that we're also starting to see that consolidation that's going to occur. We still have thousands of ISPs in this country, that's not sustainable. And I think there's gonna be consolidation in the industry as much as there's gonna be new entrants, new types of players gonna continue to grow. And again, I don't think it's just a public money conundrum versus just a private money conundrum. It's gotta be put together.
Drew: All about blended, blended money, blended capital. Let's spend the last couple minutes we have together talking a little bit about Orca Island. Okay? So what's it like on Orca Island?
Gerry: It's beautiful. [laughter]
Drew: I'm sorry I've never been there. And you started out, your company started out as I think Open 5G and...
Gerry: Yeah, that was a name I inherited, which I quickly got rid of, but yeah, that was... But Orcas is a very unique place, the San Juan Island is a very unique place.
Drew: But let's talk about how we get Broadband there. Like what, literally, where is it coming in? Do you have Fiber connections and where, just like a little bit about like what's it like there and how we get that kind of better broadband to Orca Island?
Gerry: Yeah, we operate a... Here we operate a 400 gigabit connection in six fibers inside a submarine cable that goes 10 miles at about 500 feet below the sea, back to a place called Anacortes back to the mainland. And we have diversity once we hit the mainland and whatnot, but we built about 45 miles of undersea cable, usually mixed with power and telecommunications, we've built 40 plus, 41, 42, now becoming 5G cell sites in the community. And we've deployed active e-fiber to the majority of homes. I'm sitting here with a symmetric 10 gig connection on a remote island. And everywhere I go on my cell phone, I have five to 600 megs in my pocket everywhere in a very remote community and we had to figure that out, that was the challenge.
Gerry: The challenge was we have a very low density, very challenging topography, the prior ISP we built is called Rock Island. That's very much for a reason. And 90% of the fiber we built out here is below ground because it's a beautiful place. People don't wanna see infrastructure. So we had to do it differently, which also added to cost and challenge. But one thing we did very successfully, and this is the unusual part, everybody partook in spreading the risk. We raised directly out of people's pockets. And it still continues to this day, well north of five, probably even north of $6 million at this point, where people were cutting checks themselves to whether they were a homeowners association, funding infrastructure around their neighborhood or individuals with the longer driveways. We created a financial model that allowed people to understand the real cost to them and what it took to get to them. And how they could share the risk with their neighbors. And one thing I would love to see happen on a broader basis is, how do you turn individuals in communities into your best salespeople? 'Cause at the end of the day, that's the best person that's ever going to wanna push the need in a given area at a very local level.
Drew: Well, this has been remarkable and wonderful, and even the questions are coming in. Our time has expired and we've got a great 4th of July coming up. And don't forget that in two weeks time we have an Ask Me Anything with Christine Hallquist, the Executive Director of the Vermont Community Broadband Board. And on July 17th, we've got a special live Ready or Not, that Scott Woods will host with Gigi Sohn. Okay? Gigi Sohn, who really needs no introduction, but has become very active, always has been very active, but specifically active in the American Association for Public Broadband which is super exciting. And then we've also got a, Where's The Funding episode with returning guest Chris Peritz on July 19th. And I might as well mention we've got Valarry Bullard, the Director of the New Jersey State Broadband Office in four weeks time. So lots to come up in July. The month will not ease up. Gerry, it's been such a pleasure to have you on the program for this hour.
Gerry: Thank you very much.
Drew: Take care and have a great weekend.
Gerry: You too. Great time, everyone. Happy 4th.