Off-Nominal - 152 - Low ARPU Sub

Episode Date: May 9, 2024

Jake and Anthony are joined by Caleb Henry (Director of Research) and Justin Cadman (Co-CEO) of Quilty Space, to talk about their financial analysis of Starlink.TopicsOff-Nominal - YouTubeEpisode 152 - Low ARPU Sub (with Caleb Henry and Justin Cadman) - YouTubeStarlink Webinar AccessQuilty Space on X: “Happy National Space(X) day! Get a peek into our Starlink financial analysis. - ~2.7M global subscribers, ~95% consumer/SOHO. - 60% are U.S. based subscribers, with international growth surging.”Quilty SpaceFollow CalebCaleb Henry (@ChenrySpace) / XQuilty SpaceFollow Off-NominalSubscribe to the show! - Off-NominalSupport the show, join the DiscordOff-Nominal (@offnom) / TwitterOff-Nominal (@[e-mail]) - Spacey SpaceFollow JakeWeMartians Podcast - Follow Humanity's Journey to MarsWeMartians Podcast (@We_Martians) | TwitterJake Robins (@JakeOnOrbit) | TwitterJake Robins (@[e-mail]) - Spacey SpaceFollow AnthonyMain Engine Cut OffMain Engine Cut Off (@WeHaveMECO) | TwitterMain Engine Cut Off (@[e-mail]) - Spacey SpaceAnthony Colangelo (@acolangelo) | TwitterAnthony Colangelo (@[e-mail]) - jawns.club 🐘Off-Nominal MerchandiseOff-Nominal Logo TeeWeMartians Shop | MECO Shop

CLS and go for main engine start. Hello, happy Thursday. It's a four-up here, Jake. We've got a surprise guest. A surprise guest. A surprise guest that brought a sword nonetheless. That's what I'm good for. Just perfectly located in the frame, too.

Like, it could not be better. Right by the logo. We can like screen cap that and make a tweet out of that. Like, here's the off-nominal logo and a sword. All get together. So good. We know it's going to be a good show. This is sprung from, this actually was the first show where the guest list was completely pulled together by one person on the show bullying another into appearing.

So I went out to lunch with Caleb and said, you should come on this show immediately after you unveil this big financial thing that you've been working on for months. And then he erupt justed into appearing with us as well. So this whole thing does feel like an MLM in a good way. MLM in a good way. Okay. Yeah. All right. All right.

I can go with that. So we're going to talk about Starlink. I just say you see that Justin won't be bullied into anything else because, you know, now he's prepared. Oh, yeah. It's true. He brought some backup. It was a message to you, Caleb.

No more surprise podcast appearances. But you, too, and I assume plenty of others at Kualty have been working on a huge project that we're excited. Jake personally, excited because he's a subscriber to Starlink, so he's probably going to try to peel back the curtain a little bit. But yeah, I'm excited to hear about what you've been working on and how you've been working on it, because I find the process of it particularly intriguing. So thanks for hanging out with us.

Yeah, happy to be here. So, you know, like you said, we've worked on this for months. And, you know, I think it took such a long time in no small part because SpaceX is a private company, which means it's not easy to get financial information about them. And it just involved a tremendous amount of homework on our part to come up with reliable estimates and figures, you know, as best we could for a lot of the key metrics of the company. Yeah. I'm excited to unpack those and find out what's what, because I don't understand what's going on. Jake needs help with satellites. sometimes I look at like you know like so there'll be like periodically there's like a news article

that's like Starlink is probably making money or Starlink is like underwater like you know there's like every imaginable thing that comes up I'm always reading them I'm like yeah this makes sense and then I read the next one that's like the complete opposite I'm like oh this makes sense too yeah so I don't really know I don't really know what's real and what's just made up so my favorite subplot is when SpaceX this is I'll know when they made it when they become the buys ISP because everyone always hates their ISP, so I was always plummixed. That's why they would want to be that. So at some point, presumably, they become the person that everyone's complaining about. And so I don't think that's in your report, but we'll find out. But anyway, did you...

It's kind of funny. I heard yesterday, I was at a reception at an event here in the Tampa Bay area, and someone referred to it as the... I think they said it was the nine steps with Starlink. First, you admit that that it exists and they kind of walk through every one of those excessive steps in the process, you know, and then toward the end, it was, well, it kind of followed the usual trajectory of those steps. So, admit that it exists. That's, they're going to be a nod to date. It's a knack for taking on businesses where that's the first step. Yeah.

That seems to be their thing. I mean, there's definitely still companies that won't. acknowledge, you know, we listen to a lot of earnings calls and some of them are like, yeah, we're being challenged by Starlink. You listen to the Telasat ones and Dan Goldberg will spin this into a positive and say, look, we are fighting them for customers, but it's just proof that people want low Earth orbit internet. And then they say, you know, segue into the future light speed. But for others, it's, it's the competition. Yeah. Well, I don't know, Justin, if Caleb gave you any prep work for this

because that was his job. But did he tell you to bring anything to drink? I have a drink. I do. Look at this. What is what you're cooking with down there? Athletic IPA. Run wild. Yeah, it's good. Yeah, it's a good, good non-alcoholic proof. It's hard to find

a good N.A. I've heard a lot of good things about them. I don't think that's the first one that's appeared on this show. I feel like, hasn't somebody have one of those? Yeah, lots of people have. We've had this conversation where they're like, they used to be not good. And like now all of a sudden there's been like a little bit of a renaissance with them. And they're actually like producing some pretty nice non-alcoholic bruise. Yeah.

Yeah. Yeah. Yeah. Yeah. I did not. I've, I've got my, uh, my superior here. I just go ahead and, and, uh, some of suju. So there's a, there's a store nearby that has a whole bunch of,

different Korean sojus and I have a big fan. This passion fruit one. You brought one of those to my house once and it was delicious. Yeah, I forgot about that. I don't know what was that one, but I know it was one of those and it was delicious. It's about the same as wine in terms of content. It's like a glass of wine. Okay.

What you got, Jake? Rusty beer. I forgot about that. So I'm rocking just a local brew. This is Petito here. in Yucatan, just a regular logger. It's kind of like their everyday beer, but I had one in the pre-show. I'm playing it fast and loose today, two-beard day.

So I had one of the pre-show, and I don't know what was up with it, but the cap was all rusty. I don't know if you can see this, but there's like, there was like rust inside the cap. And I don't know what's going on with that. I drank it anyway because, hey, there's only so many days in your left in your life, right? It's shortening them. Yeah, it's a little rust in the water. It can't be so bad, right? Jake, I thought last week I drank my last star base beer,

but I found that the last one had actually just been moved, and I have a lucky launch day logger. Nice. So I don't know if this is technically the Starlink launch day. Is that accurate? The Starlink financial analysis launch day? Or was this just the day with the sessions? When did this happen?

Yeah, so this was the day with the session, but in a way it was like the launch day for the general public. We made it available and our subscribers, I think, knew about it beforehand. But we wanted to also bring it to more people's attention because it's something we get asked about incessantly, honestly. It's on the mind of just about every satellite communications company and then even some operators outside of the SACCOM world. So I figured we might as well.

Do you do a project like this not because you actually want to know the answer, but just because you want to be able to have a link that you can send to someone when they ask you that question and you don't have to answer it? I mean, saving time is a big deal around here. Yeah. But no, I think we chose to do this study. And it was to answer some of the questions that we kind of started this off with. Is Starlink making money?

Are they not making money? Does this business model make sense? For decades, this is an industry that has gone through fits and starts with the idea of putting large networks of satellites in low-width orbit. And there's a whole list of the ones that failed either permanently like Tel-Medesic or went through their own Chapter 11 process and then emerged in varying stages of health, la, iridium and Global Star. And so then we have this new wave. and the question persists, does this make sense? People get really excited about the technology, but financially from a business standpoint,

what's going on here? So I think our conclusion, which I can let Justin expound upon, is that Starlink, in fact, is making money and is doing so sustainably, believe it or not. Yeah, yeah. I mean, that's kind of, you know, it's one of the topics

that it just, it's in every meeting these days. You know, it's a discussion around Starlink. And sort of like, you know, what you were saying earlier, one discussion, oh, they're making money hand over fist. The next one, it's so they're at the verge of a liquidity crunch and bankruptcy soon to follow. So we kind of set it upon ourselves to try and get to our own conclusion. And that became a labor of love over the last few months, collecting information. talking to different people, doing our own homework and analysis to kind of come up with where we came out.

And, you know, hopefully it helps to put some of those questions to rest, particularly for other people in the industry, right? I mean, they're having to make decisions. Do they want to invest more in X or less in Y? And if you're doing it based on faulty information or if you're kind of making the presumption that this is a flash in the pan and it'll be gone tomorrow, you know, it's helpful. to have somebody standing behind you and saying, look, here's kind of the way this business operates. Here's how they make money. And, you know, factor that into that decision making. Yeah, yeah.

This was kind of an inevitable project, I feel like, for you guys to take on. But it's also weird in that I feel like you had to be cognizant of what phase of the Starlink project overall we are in at any given moment. So I'm curious how, what the decision is like, for when is the right time to do this sort of large-scale analysis because we went from like really rapid initial rollout to revisions to now like an incalculable amount of launches that are

happening per year. There's like always a Starlink launch happening every minute. So, you know, is the rate of change on what the actual state of Starlink is? Is that a driver of when this kind of analysis makes sense or does it feel like you could repeat this on a certain, you know, interval to catch up with new versions? What is that? How do you understand that? The other way to ask that is like, does this report mean anything in six months? Right.

Like, how is this now your job to just be the Starlink financial people? I think it's something that we will stick with. Certainly if our subscribers find continued value in it, the idea was to take a crack at doing it now and if people receive it really positively to continue. developing this and maybe every six months unveil an updated version. You know, with publicly traded companies, we have models at Quilty that we update every three months because that's when they disclose information. And so for this, the process is obviously a bit different, but I suspect that there, I suspect strongly that there will be recurring value because it's not a question that's going

to go away for the industry. And so we, we have sketched out some ideas on how to expand that to include future parts of the Starlink business, and especially as they become more significant in time. This was, we wanted to just carve out Starlink, so not looking at launch, not looking at anything else that SpaceX is doing, no crew stuff, and just selling Internet to people, how much does it cost to do that? How much do you make back? Are you making money?

That was what we did. I think as long as the industry is interested in it, then it's something will continue to pay attention to as analysts. Yeah. When you say that it's sustainable, like, what does that mean? Like, are they making $1 or are they making like hand over fist change here? Like, what's the scale of what you were able to determine? Justin?

Yeah, so sustainable to us means are they at a point or are they reaching that point where sort of the flywheel effect where the business has enough momentum, enough cash flow that they can survive. There's no longer this speculative arc of the question of, you know, first, can they do it, right? And then second, can they finance it? And then third is, will there be customers? And fourth, will those customers be willing to pay and are the costs sufficiently low

that you can build a business based off of it? And I think that's become pretty clear to us. in the last, you know, increasingly over the last couple of years, but in particular over the last six to nine months, that, yeah, this is a sustainable business. These guys will be with us for a while, long while. This isn't, you know, that, that ephemeral kind of flash in the pan situation. So with the level of cash flow they've got and with the amount of CAPEX that's required to sustain this network, they can keep doing what they're doing. There's no risk from that standpoint. What did you expect the existential risk to be? Like when you were digging into this project

the first time, did you have an inkling of like, I bet this is where, you know, this will be like the most crucial thing that they need to get over? I would say there's a lot of people, even today, that have grave doubts about two things. First, is the pricing that they're pricing their service at today artificially low? So bring on the customers with, uh, sort of low low ball pricing and then raise the pricing at some point in the future. And the answer is they don't need to. Now, whether they do or don't, I think that's a business question, right? I mean, they could look at supply and demand and elasticity of demand and figure out,

okay, what's the optimal pricing, but they don't have to is the bottom line. So then the second question that's been a recurring theme is, well, this is great. They're putting up this first generation of satellites, but these things don't last very long. And it's going to cost billions upon billions of dollars that are going to hit this brick wall once they need to begin to replenish the system and put up the next batch of replacement satellites. And so both of these things were, and we had hypothesis, but we didn't really have facts to prove that or demonstrate that. And so that's what we set out to do. And I think through that work, we've determined that actually the capital intensity, is not going to be that much more than a lot of telcoves for that matter once you get to a certain level of scale.

And they're getting close to that point right now. Right, right. Because even classical telecos, they've got to run wires and replace wires and put up poles and towers and all that kind of crap, right? So infrastructure is infrastructure, whether it's in space or not, right? Oh, for sure. And, you know, look at the wireless carriers, right? You know, nobody would claim that those are maybe the best business. model in history, but every few years, it's it's 3G, it's 4G, it's 5G, okay, 6G is around the

quarter, and each and every time it's, you know, more base stations, it's upgrades, it's all this other stuff. Buying spectrum from satellite companies, you know? How is that different than just putting up more satellites? It fundamentally isn't, but I think people have this view because of the legacy baggage of the satellite industry that satellites are expensive, launch is expensive, and, and therefore, you know, no way. You can't do that, right?

And with scale and with vertical integration, I think that that model has been proven. You can do that. It's not different fundamentally. Right, right, right. The other big one that everyone always says is like the terminal side of things, right? And I know, Caleb, over the years,

I've grilled you about different people's terminals and there was like that mythical tweet that I always refer to where Greg Weiler tweeted some picture of like a pizza box with the cool drawing on it. I was like, this is a super cheap terminal. It'll come out soon. Never really heard from that.

So, I mean, Jake bolted one to his roof. So they certainly exist. And it's just a matter of the argument of how much money did SpaceX spend so that Jake could bolt that thing to his roof. And does that have a big impact on the overall financials? Or is that similar to what you were saying of like there's the initial arc to get up to speed? But once they're there, things are kind of hit a cruising altitude and they're able to sustain that from that point. The terminals was a big question. Honestly, back in 2020, 2021, it was probably question number one for any Leo system, right? Because traditional VAT antennas are not expensive, especially a lower grade one. Whereas, you know, these electronically steered antennas and other types of terminals to support, you know, NGSO networks exorbitantly. expensive. So SpaceX, Kuiper, you know, they're showing that you can, you can manufacture a terminal

that that is not, you know, extraordinarily expensive. I wouldn't go so far as to say they're cheap, but they're no longer the fly in the ointment. You know, they're not subsidizing these to any kind of major degree, whereas a few years ago they were. And that was a huge, huge, cash cost for them a few years ago. I'm really interested to kind of find out what the arc is on like reliability on the terminals because I have this, maybe this is completely wrong assumption, but I kind of have this assumption that the early ones that are that are a little more expensive are probably like, you know, pretty beefy and designed to like be pretty robust.

But as you like refine the manufacturing process on them, you'll try and like, you know, you'll take those cuts where you can to try and get the cost down. And I wonder if they get if the reliability kind of goes with that. Because like these early ones are like this one on my roof has been rock solid. Like it's been exposed to tropical nature. Like I'm at 21 degrees latitude. It's a bajillion degrees outside this weekend. It just sits there in full sun all day for and it's I've had it there for a year and a half and

it's never ever given me one hiccup. So like that's like fantastic story from a reliability standpoint. But like this is also you know that this is like the gen two one. It's the first one that was square. right? So I think they have a new one now that's like a lot. It's, you know, it's got the corners cut on it per se. It's got a few different design changes to try and make it cheaper, cheaper plastic, I'm sure is one of the things in there. So I wonder like if that's going to be an issue, if they kind of have to get some sort of, I wonder if that like equilibrium point is going to be

good enough wherever the cost ends up, right? That's kind of a question that I have. Yeah, I think that if you are looking at where they are shedding costs, I don't know if they shipped to it. a cheaper plastic, maybe, maybe not. But the two things that caught my eye were removing the motor. I think the motor that would help the antenna steer. And then the other was, I think there's been a reduction in the number of antenna elements inside of it from the early versions.

So you think of the elements as kind of like cells almost. And so I don't recall how many they have, but you pick a number. Say they had 2,000 of these initially, if a later version of them has, you 1500 you shed a lot of cost historically for flat panel antennas this was the intractable cost element it was for element like a better word if you couldn't get the price of an element down lower again if you need and you need a lot of them so if you needed 2,000 and you can't get the price down below uh 10 dollars for each of these you got a 20 000 antenna if you get down to one dollar it's still a $2,000 antenna.

So one of the things that SpaceX has been able to do with Starlink's, they've got so many satellites up there that they don't need as many elements as some competing systems in order to see a satellite, to see multiple satellites, to gather as strong a signal, they can let the constellation do a lot more of the work and simplify the antenna. for space systems, for satellite systems, it's typically one of the other. If your antenna, if your antenna is cheap,

your spacecraft are going to be more complicated. If your spacecraft are simple in design, your ground network becomes much more complicated. Somewhere in the system, a lot of work has to be done. Space X shows for it to not be the user terminal. Right. I mean, so the other aspect, too, that's hard to get away from in the conversation

is the launch of all these satellites and how different their situation is when it comes to that than everyone else is in the world. So there's two parts, right? There's within your own analysis of it, how does the actual financial side of launch really shake out? Because that seems, we've had some indication over the years of like what an internal cost is on a Falcon 9, but Jake was trying to figure out like how any of that gets accounted for of like what is a Starlink cost and what is a non-Starling cost. Yeah. And then the other aspect is, does that capability to leverage their own launch system actually drive them more in the direction that you're talking about, Caleb, where they can be a little bit more flippant with a launch year or there or 10 or 20 or 100 a year that lets them actually make that call, whereas for other companies, that would be prohibitive because they'd have to buy so many very expensive launches from external sources?

I think because we only have a few more minutes with Justin, I'm going to defer to him. on this one and then I'll try to help answer. He doesn't want to get sorned. Yeah, I don't make them too angry. Yeah. So, I mean, the launch side is a critical part of it. And, you know, the fact that they are truly vertically integrated and the reusability of these Falcon 9s, I mean, this is a key part of why they're able to do what they do, right?

And we've done some work around, you know, the internal cost. of these Falcon 9s once you start to reuse them, you know, this number of times. So it gets pretty attractive in terms of, you know, the contribution to the cost of the system, you know, in very, very fast and loose terms. I mean, it's roughly half satellite manufacturing cost and half launch cost, if you think about the space segment. I mean, it's gotten to be that low. But, you know, Caleb, you want to fill in the blanks in terms of, you know, what they've been doing there.

Yeah. So, I mean, like you said, the more that they can reuse it, the more they can kind of amortize the cost over a long time. And, like, SpaceX through their reusability by having it vertically integrated and by having spacecraft that, I mean, now the V2 minis are bigger, but you saw how small they used to have 50 to 60 of these things per rocket. it just really enabled them to drive the launch costs way, way down, because not only were they reusing the same vehicle, but they were putting up more capacity than the average satellite operator than any other satellite operator. And so that kind of helped them a lot with keeping those costs down.

How does that get accounted for within either your analysis of Starlink or like their internal analysis, right? Like when they're buying a launch from themselves, does that, you know, whatever the marginal thing is now 10 to 20 to 30 million dollars depending on who you talk to, does that get kind of pulled into the Starlink side of the budget? Or I guess for you guys it had to, but for them, like, what do they do that? Launch gets on the Starlink P&L, right? Like what does the internal accounting have to look like?

That's an endlessly fascinating question to me. Yeah, I mean, there's lots of, and there's lots of parts to that, right? I mean, there's the tax side of it, you know, should they spin out and an IPO this company or if it's no longer a wholly owned subsidiary, I think things start to look different, right? So in terms of the way we're looking at it in our model, it's basically at, you know, the fully loaded internal cost, but there's no kind of margin built into it. So we're not, you know, we're not assuming that SpaceX, if you will, is earning a margin. If you, again, air quotes around the launch, right? But I think it would look different if you were to strip this away and sell this business or take this business public. Then you have to account for the fact at that point that SpaceX needs to, you know,

they're not doing this just out of altruism. They need to make money. Yeah, yeah. Yeah, even just thinking, like, it's pretty easy to do some of the obvious things like, okay, well, the fuel on a Falcon 9 should probably be like part of the cost calculations for a Starlink business, right? But then you get into some of the more like complicated fixed costs, like, you know, how much of the lease for the launch pad goes over to the Starlink P&L? How much of the, you know, the fees to pay for some sort of like the valve company. the valve company, yeah, safety stuff, you know, range stuff. Like, you know, they got to have a relationship with the Air Force or whatever

and trying to figure out how much of the cost that they incur just by being on the space coast or whatever. Like, how much of that does Starlink incur as a cost to them versus, you know, just a freebie they get by being vertically integrated, right? It's a tough question to answer. And I don't even, I imagine that SpaceX kind of doesn't really have a super fine grain, like, internally as to how they do that, right? That's kind of the point is like, yeah, we've got this, this is a huge asset that we can leverage to actually make this thing possible in a way that, like, we changed, we didn't like the maths, we just changed the math by way of becoming

the best launch company in the world and then using that. And everyone can be like, well, but we don't have our own launch. It's like, well, you could have made a whole launch company for 20 years and then try to be a satellite operator, but you started with the satellites. So that's your cross to bear carry at that point, you know? Yeah, yeah. It's quirky, but Justin, I know you got to run, so we appreciate you hanging out with us for a while here. And, yeah, I mean, we'll talk to Caleb a whole other chunk about this.

But is there anything in particular before you leave that you want to mention or point people to? Starling's making money. No. Jake wants to know if it's $1 or a billion. That's Jake's one question. Yeah, yeah. So even at a less cap X this year, 0.6 billion. All right.

That's good. I mean, it does have a B in it. So not a full B, but that does have a B in it. It rounds up. I'll take it. That's a lot, man. That's a serious amount. You can buy two Delta four heavies for that.

No problem. Anyway, Justin, thanks for hanging out. Thanks, for fun. I appreciate it, guys. Thanks, Justin. All right, now we get to bully Caleb around on, you know, some more of these questions. Jake, do you have more like low-line questions? Because I have about 8,000 questions on how they actually did this tactically.

Yeah, we need to get into methods. But maybe I'll ask one question first for we get into that. So just sort of like thinking of Starlink as an international company, you know, I'm curious to know what you learned about how important is. Okay, so the prefix here, the preface of my question is like there was a lot of talk about how Starlink is going to be this amazing thing that goes out into these developing countries, into these rural areas where there was no internet and they're going to bring light to the people, you know, like this very, this very altruistic narrative.

And I'm curious to know how much of the international segment, like, and let's maybe we split it up by like developing versus developed countries, like how much of that is coming from developing countries in the analysis? Or did you learn anything kind of interesting about how that ended up being set up? Yeah, so the vast majority of Starlink subscribers are coming from your high ARPU, high average revenue per user markets. It's not a whole lot of the emerging world. I remember right by our account, and they have occasionally, SpaceX is occasionally given some figures about like what percentage of their subscribers comes from the U.S. They had like 200,000 coming from Australia and something like that.

So you can pretty confidently say that like at least two-thirds and probably closer to like 80% or more of their subscribers are from geographies that are used to paying, paying a higher amount for their subscription to the U.S., in Europe, Australia, Japan, New Zealand, these kinds of places. some that have surprised me, I think they do have a pretty decent subscriber footprint across Latin America. I get a Jake where you are. And something that I'd actually attribute that to, you can almost see that this was a market that was kind of softened by two things. First was actually SpaceX as competitors.

If you look at where like Hughes and Biasat have expanded beyond the U.S., they went to Mexico, they went to Brazil, they went to Brazil, it went to Central America, is that these are markets that we think that some level of satellite broadband will be adopted, either on a per household basis or on a community Wi-Fi basis. And so there was some familiarity there. The second is that you can count, I think, a pretty high number of universal service obligation programs across Latin America that have relied on satellite connectivity. This was in Brazil with the SGDC satellite or Mexico with the MECSAT satellite program. and kind of go down the list. So there's no stranger, there are no strangers to the use of satellites

for reaching rural and underserved populations. And that's why I think you're seeing that there. Now, we have noticed that the percentage of international subscribers is going up. It's ticking up. And I would not be surprised if, I don't know I'm going to throw a number out there, but call it 18 months.

that they have more subscribers outside the U.S. than inside. That would be really interesting to see. But still, their core, their core revenue generating customers are from the U.S., Europe and Australia. Yeah, because like you said, the quote you call that high revenue per user or whatever. So, I mean, because when I bought this service, so I'm at a year, almost two years now, and it was, my original price was like $2,300. hundred pesos a month, which is like, call it 140 US, 130, 140 US, somewhere in that range, right?

I don't remember what the exchange it was at that time. But then they did a thing almost immediately after I bought it where they lowered the price and like by half. So I pay today like a thousand forty two pesos or something. So like it dropped significantly. And I was always trying to figure out why that was. Was it like, you know, was it sort of a recognition of like this price was out of line with what the market was and like people just wouldn't pay that and that's just like a reality

or was it like they were trying to drive adoption and that price is going to go back up later almost two years later it hasn't gone back up so maybe that's not what it is but you know that's kind of an interesting thing to see about so even if like subscriber counts start to go up like that average per user is makes it a significant difference right because most of the u.s subscribers are paying twice as much as i am for this kind of thing right so it's kind of interesting to see how that plays out i don't know Yeah, certainly. And there's also, you know, SpaceX is now increasingly moved into non-consumer markets. You know, they've got terminals that are being used by the U.S. military.

They've got terminals on all manner of boats now, cruise ships, merchant vessels. Yeah, yeah. And in the aviation space, airplanes, trains, they're starting to see more adoption. And those are even higher, you know, average revenue per user customers. customers. Some good ARPU on those things. Yeah, good ARPU. Yeah, exactly. Those are the only ones that could afford those $20,000 flat panel.

Yeah, right. So now if you come in with an even cheaper Starlink one, you understand why they're having such a disruptive effect on the market. Yeah. Yeah, I always thought that like a country like Mexico or Brazil would be in this category two where it's like it's a developing country, but it's like also like an advanced country, like that kind of like middle ground sort of place.

because like, you know, I don't live in a place that's like completely undeveloped. I'm not like, you know, we have plenty of modern things, but also you can see all these signs of like this is a country that doesn't have it together yet. Like, you know, we're halfway down the road. I kind of like to describe it as here. Because, you know, because I used to live in the city in Merida here, which is a big city, you know, a million people. And I had great internet access.

I was, you know, I was paying a good price. It was gigabit, like, fiber rate to my house. It was rock solid, awesome, like exactly the kind of stuff I would have expected back home. And I moved a half hour north outside the city limit. And literally, I had no option. It's like Starlink was not my choice. It was my only option. There was not another internet I could get.

And I'm just not that far away from where I used to live. You know, it was like 20 kilometers or something. So you can kind of see there's like, there's this middle ground where like, you know, if you get in the right spots, there are these like beacons of development in a country like this. And there's people who are ready to pay, you know, first world prices for whatever you want to get. Like, you know, $100 internet is not not unreasonable for a lot of people that live here. But then you also come out to this place like this. And, you know, I have the best internet on my block.

Like no one else, I'm the only internet on my block in some cases, right? So it feels like the right kind of market for exactly what Starlink is because you can kind of sneak in and grab all those customers that would happily buy the landline, but it doesn't exist, right? It's also a topography situation when you get like south of the glacial plains of North America where like that gets rugged and there's crazy jungles and a lot of mountains and it's hard to put like regular old earth infrastructure in. So yeah. We're having this thing. Big driver. We're having this great conversation in the discord today about kind of like business cases for this, you know, for satellite internet. Like where does it make sense right?

And like the common thing that you'd say is like it doesn't make sense in an urban population. like Starlink's never going to get popular in New York City because you can run a really dense cable like for a pretty low price and serve a lot of people for you know cost per infrastructure is low right but like in Mexico that's not necessarily true and you can kind of see an example of what happened with this with cell phones because for a long time Mexico didn't have great phone service because running plain old telephone system pots lines to everywhere Mexico was just a cost the government could not incur. So like people just didn't have phones. And then cell phones came and you realized you could just put one tower and cover everybody. And like they just jumped on it. Right. And so like Mexico kind of skipped landlines for phones. And you could almost imagine that happening now for for internet.

Like maybe we'll just skip all the fiber runs out to the rural communities and just get satellite. And it's like perfect. Right. So it's kind of interesting to see how that could play out, you know. Your ARPO has got to come up a little bit though. A little bit. Yeah. Those are rookie numbers down there.

Hey man. 130 million people here, though. You can keep the Arpoo down and still make a good chunk of change, right? Kim, how did you draw out the subscriber numbers for the analysis, right? Because occasionally SpaceX will throw a tweet out, like, hey, we've got our 100,000 this or that or whatever, but are you drawing it from that plus like the negative space of other competitors that they lost 300,000 subscribers and SpaceX seems to have gotten that amount?

Like, how does that worksheet actually function? Yeah, so I would say it's only so granular at this point in time. There are some countries where we were able to obtain solid data points. The U.S. is one. Australia I mentioned is another. You know, it's not all 75 or however any countries Starlink is in today, but we're able to use that and get a pretty good sense offhand. I mean, again, I think the U.S., the figure that's coming to,

mind is at some point they told the FCC they had 1.3 million. I think they had 1.3 million American subscribers when they had 2.2 million total. I'm trying to I'm drawn from memory, so I could be wrong. But I think that's what they said there. And so yeah, you add that up, then you add Australia, and then you get a figure for Europe and you start to see, you can piece it together and like, all right, I'm not going to have every single, you know, I don't know how customers they have in the Maldives, but you can get a good sense of where their revenue is coming from. Yeah, yeah.

Hmm, there you go. Look at that. Yeah. This is the most 2023 chart of all time in which it's the, the groups are consumer, government slash defense, enterprise slash IOT, mobility, and Ukraine. That's like the most could not be more 2023 on its own. Yeah. Well, SpaceX has said that they had 70,000 terminals in Ukraine.

And that's an interesting case there because if you look, and I haven't looked at it this year, so these numbers may have changed. But like around the time that the conflict began, $500 was a really high bar. I want to say that that was like a monthly, like the average monthly salary in Ukraine, and maybe monthly discretionary. 500 was like a figure of importance in terms of like measuring like the Ukrainian economy, like the average person's well-being, you know, fully. That's not easily attainable by everybody.

You know, it's a significant purchase. So for so many of them to be there tells you about just the sheer volume that are being used for military applications. And it's not at all a surprise given the ongoing conflict. but that gave us a sense of, okay, what is Starlink's military business, their DoD business? They have contracts that have also been placed that are in the public domain with DOD, and you can follow those and see, all right, here's where terminals are going. Yeah. Plug all those things into a model.

That initial price is like, again, I'll be going to give you all the Mexico perspective, but that's like a big deal here too for adoption, right? because, you know, like, obviously Mexico is a poorer country than something like the United States, but that's not just like imagine the average American wealth and then just like make it less. It manifests differently, right? And that upfront cost is like much more significant to people down here. There's less capital flowing around here, right? So like the monthly rate may be fine for them, but that upfront cost can be just absolutely

prohibitive because they just can't, you can't assemble that much money in one single point of time for a lot of people, right? And loans and stuff are just like nowhere near. Now, the financial instruments available to people here is just significantly different. And so, yeah, that's something that I feel, you know, taking back, coming back to our discussion about the terminal costs and stuff, like that is clutch for getting it out into the rest of the world. It's getting that thing down below 500 bucks for sure, for sure.

And I do think that the progress that Starlink has made and is making, it is expanding the market. I mean, they doubled, like they grew so fast compared to their competitors, like Hughes and Biasat, who had name dropped a couple of times because those were, that was the barometer. Like, that was how you understood the satellite broadband market for a time. It was you had these two players. I guess you had K-A-Sat in Europe for a little bit, which was kind of a smaller. I think Yul-Sat at the time launched one satellite, and then they never built a successor,

so it ran out of capacity, and then the numbers just kind of dwindled from there on out. but we were at a point where Fiasat hasn't disclosed their numbers in some time but they plateaued around 600,000 subs and Hughes plateaued around 1.2, 1.3 million and so and that's like that's as good as they got. Right, ever, over years and decades and period. Yeah.

And then Starlink came in and had more than both of them combined and is still growing. And we expect that they're going to reach somewhere between 3.7 and 3.8 million consumer subscribers by the end of this year. Plus, I think the total, if you count all terminals, would be like 3.9 to 4 million, including all of your enterprise and mobility and defense. So it's a staggering growth rate. And it comes back to those lower cost terminals. their large supply of bandwidth and the fact that Starlink,

even though they have run into capacity constraints from time time because of how rapid customers were adopting it, they've continued to scale that network. There's a recurring pattern for all three of those players with their satellites. You launch a new satellite, you grow for a little bit, and then you max out, and then they flatline, and then they start focusing on that term I keep using the ARPU.

They'd say, oh, we're going to focus on the higher ARPU customers, and we're at a capacity, but maybe we can give premium services to the guys that'll pay the most. Then you start seeing those numbers. You see attrition. SpaceX is not playing that game. Every time they get low on capacity, they launch more satellites. And so people have this thing to look forward to. They can say, all right, maybe it's not great today, but three to six months down the line,

it's probably going to be better. And then not only are they launching more satellites, but they're launching another generation of satellites. So it's not waiting three, four, five, six, even upwards of seven years for a new geosatellite, which has the unfortunate risk of a single point failure, which we saw with Viasat 3. There's that safety in numbers, there's that redundancy, there's the assurance that this is a product that is iteratively improving in a way that, again, can mirror the terrestrial telco experience of, okay, It was SD now, it's HD, now it's 4K. You can count how many megabits you're getting and see it continue to improve and say, okay, this is nice.

Yeah, yeah. Should you talk methods? Anthony, I know you have a lot of questions about. Yeah, because, like, well, we talked a little bit about this when you were hanging out up here last week, but, you know, you had to get down into details to make the model work, like satellite costs and low-level stuff that there's almost no insight into.

So when you're approaching those kind of problems, like how do you even start assessing which of those you need to fill out and then going about that? Yeah, we assume SpaceX didn't give you access to their accounting and stuff, right? Yeah, yeah. They reached out and been like, uh, they just email you a bunch of PDFs and be like, can you make sense of this and make the report? We've outsourced our business intelligence department, so. That would be nice. No, I don't know how much of the secret sauce I can give away. Yes, tell us how to do your job so that we can take your job.

Clearly, Jake and I will be the director of research somewhere someday. There you go. You can start the off-nominal research company. Yeah, we're just doing vibes. We're doing the vibe check company. Don't worry. We're not going to get in on your actual data. Yeah, we're not doing actual data.

Well, I mean, I'll say the plug for subscribers, you know, Fultiespace is the like are kept on by people who buy the research. The model is there and that you can see, you know, the numbers are color-coded based off of which ones are assumptions and which ones are hard numbers. And people who buy the research can tinker with it. Like I've already had several conversations with people where they say, okay, I agree with you on this number, but maybe not on this one. Or how did you reach this price point for the satellites? Like our, the V-1 satellites, we estimate we're around $200,000 apiece, which is incredibly love. How do we reach those conclusions?

We did just a ton of research. We spoke to a bunch of people. We looked at cost models for other spacecraft. We looked at what SpaceX is doing and recognize the value of things like how vertically integrated they are, how little outsourcing they do. What a high volume they're building at compared to every other space program. You know, like for example, a lot of, a lot of costs for satellites can kind of be predicted commensurate with the mass. You can look at how big the satellite is and say, okay, the price kind of scales along this way.

But that doesn't work perfectly for Starlink because they get a volume discount that few others in the industry can say they ever have. And so how do you balance these these trends of like satellites gotten a little bigger, but you're producing thousands of them, you know, 2000, a thousand, 2000 a year, whatever it is. So we look at all of those things and came together. And then we would debate them internally. And then we would run those numbers against subject matter experts in their various fields, whether that was building satellites or building antennas or building gateways and even forward looking questions.

Like how low do you think the cost of this particular subsystem can go? and then use that as part of our kind of forward-looking. We didn't look that far in the future. Our projections are just to the end of 2024. But we put all of that together to say, okay, here's what we feel confident about. Here's where we think they are on how much it costs to build a satellite, how much it costs to launch a satellite,

how much it costs to make an antenna, all of these things. And then you take it and you kind of get their costs together and you look at how much they're making and you do your eBed-a-9-Q-X and you figure out if they're cash positive. Easy. This is math. Bing, bang, boom.

Done. Revenue in, costs out. What's left? That's it. That's literally the entire thing. I mean, the model is probably like two dozen pages of Excel charts. Yikes. And it's a monster.

But yeah, that's why we put something together that we feel confident about. And of course, again, SpaceX is not private. They didn't give us, or excuse me, SpaceX is private. They didn't give us their numbers. But we did a tremendous amount of homework to come up with what we believe are robust and defensible figures to understand their business. And lots of satellite operators and other companies, investors are trying to do the same thing to figure out if this makes sense. we figured as specialists in the field of space. We've already got models for a dozen publicly traded companies

that we are updating every three months right now. We know how these businesses run, take our shared knowledge and apply it to Starlink, and this is what we came up with. In all the other areas of spaceflight that we talk about with SpaceX, one of our core tenants is that, SpaceX is just an absolutely extreme outlier, like in every way, right? They are so significantly different than the entire rest of the industry that it doesn't

really even make sense when you tell other people about how this stuff works. And then there's a lot of companies in the industry that go, like, well, we're going to be just, we're going to do the SpaceX thing. We're going to do it the way that SpaceX does it. And it just doesn't work because of something, right? That thing is, whatever it is SpaceX has is makes them an outlier. Does that track over to the Starlink side as well? Or is it the fact that this is the first successful version of the new generation?

Like what's the, you know what I mean? Like in comparison to one web or the other ones that are kind of their peers, are they also an outlier or were they just the first successful version of this generation? They are still an outlier. To give you an example. So optical cross-length, though, the laser links, that they have, those are a really expensive component. And through the space development agency,

the constellation that they're building, there's some public pricing information about how much those laser links typically cost. It's hundreds of thousands of dollars for those things. And when we started doing our research, we heard multiple times. We were already pricing them lower for Starlink. And then people were like, you would be

terrified of how low the cost is for the laser lengths that they've built. I don't know if they're just like taking like pieces off of telescopes and doing that. Astronomers will use some telescopes have like laser pointers that they use to calibrate, remove atmospheric distortion. There's other fields of optics that I think the satellite industry could learn from. I don't know if that's what SpaceX did. There's certainly a recurring theme I think for SpaceX was looking at fields. outside of the space industry?

How does the space industry do this? Do we have to do it the same way? Is the story from the old days apocryphal where they were buying solar panels from Home Depot? I feel like there was some crap about that in the early V1 and I was like, I get the vibe of that story, but I don't know if that's true. I don't know if that's true,

but I do believe they are using lower quality silicon for their solar arrays, And that's why they're so big compared to like the triple junction, you know, 25 to 30% efficiency, radiation hard-end, the solar arrays that you see on other spacecraft. Like, SpaceX said, no, we're going to use the cheaper ones. And like, we don't care if we make them bigger because we own the rocket. Yeah, right. We've got this whole rocket.

Might as well jam it full of solar panels. Wait until Starship. The thing should be like an R.O satellite up there. It's going to be ridiculous. But you look at what they did for key parts of the system. solar arrays, they bucked the industry standard, optical cross lengths, bucked the industry standard, trying to think of that. I hear rumors about other systems where it's like they did something radically different. They talked publicly about taking a very different approach to propulsion

using Krypton and I think now Argon versus the industry standard of Xenon. They really work through a first principle's approach to everything here. And you combine that with the, you the SpaceX factor and they were able to produce this radically lower costs and it also helps them i would say compared to other constellations that they've got a low a low altitude that shields them from some radiation that you would encounter at higher orbits so if you look at where a lot of other leo constellations are starting to concentrate tell us at space development agency one web we're all in that 1,000 to 1,200 kilometer range, and your radiation exposure starts to creep up around 1,000 kilometers.

And that means you need to have more radiation hardened parts. Otherwise, you're going to have spacecraft failures. So Space 6 is a little bit more shielded from that at the altitude they are. And Kuiper presumably will be too, because Kuiper is right above them. Yeah. That's your next project, I assume. I wouldn't be surprised. They are not currently profitable.

I'll save you the money. They're not currently. Yeah, yeah. And I should add, even on going back to the profitability thing, this is the first year that we show Starlink making a profit. It was losses and it was steeper losses before that. So this was a very capital-intensive business to get going. And even though it's, even though we believe it's sustainable now,

it's still not cheap to sustain. Yeah, yeah. I also don't know. Does it, do you factor in like the, I don't know how necessarily it would, but I don't know how you do any of this stuff. So maybe this is a dumb question anyway, but like R&D for the upcoming or now rolling out direct-to-sell versions, right? Like presumably they're working on V-whatever's of all their different technologies. But that one, I assume, was a pretty big effort for them over the last couple years to get to the point now where they're starting to roll them out. see how that market does.

But is that something that you're going to get to on the next version, or is that also in this? That's the next version. Yeah. That's going to be, yeah. So we shouldn't ask about this three minutes left in the show about the direct-to-sell market and what it's going to do to the satellite industry? It's one that we follow as well. I mean, we publish research on AST.

We publish research on Eridium. Both of those are companies that are involved in that feel. we pay very close attention to it because Jake you were talking about that high barrier of finding money to pay for a terminal several hundred dollars if satellite companies can go straight to phones it obviates that this is no longer concerned suddenly every smartphone is the satellite user terminal and your market size just explodes when that's the case it actually makes it really hard as an analyst to to deal with that because any, like, any change you make, like, moves the needle drastically. Like, suddenly you start talking about huge portions of the world population. And the numbers just get really hard. Yeah.

You've added a zero. It's all the math. It's, like, kind of ridiculous. Yeah. I can't wait for that market to actually exist because there's a, this is for everyone on this, on this little chat here and everyone listening, there are a shocking number of the non-space nerds in your life that think that your cell phone today uses stuff. in space that's not just like GPS signals that are coming down to Earth.

And I can't wait until that's actually true so that we can stop having this conversation. Like, no, you're actually talking to a cell phone tower. You are getting GPS, but that's not what you're thinking that this does. So one day that will be true, and then we won't have this conversation. That'll be great. Yeah. Yeah. And I can stop paying for my little Garmin orridium thing.

Yeah, one day. But I am one of those users right now. So I'm low Arpoo. I'm low Arpoo on that, though. It's like 15 bucks, I think. Little Arpoo said. I'm not a valued subscriber, I don't think. A low Arpoo said.

You should get Matt Dash on here. I bet he'd be fun. Can you hook us up? That'd be great. I'll shoot up an email. Okay, thank you. And with that now there, I got you one added guest for this show. I'll try to get you another one for a future episode.

We'll hook you up with the Jeff Bezos contact info, and then you can use that for your Kuiper. math that you do. I almost think we need to have you just like immediately back to continue because like we didn't touch anything to do with Kuiper. We didn't touch anything to do with cell phone. It would be convenient timing, Jake. It would be convenient timing, yes. We got a surprise all on the schedule.

Yeah, yeah. All right. We'll talk about this later. Yes. Caleb, thanks so much for hanging out. Where should people go if they don't know about? Quilty things. Yeah, and plug the subscriber thing.

I want to know more about that. Sure. You can find us. Thanks for pulling up the website, cooltyspace.com. You can follow me at, I think it's Chenery Space on Twitter now. We provide research on both thematic things in the industry,

industry trends, and specific research on companies, as exemplified here. So follow us if you've got space industry questions. Love it. And get the ARPU up on the Quil space subscriber count. Let's get our ARPU up.

Yeah, we've got to boost that ARPU. All right, guys. All right, Caleb. Thanks so much for hanging out. Everyone else, we're figuring out next week, as you may have heard us. So we'll let you know what's going on. But, yeah, yeah. There was some changes.

So we got to have a scrub that matters. There was a scrub that matters. See you, everybody. Bye, everybody. One, two, three, four, five, four, three, two, one, into death.