Just 5 years after its first launch, the Starlink constellation is profitable

Starlink's estimated free cash flow this year is about $600 million.

See full article...

 

[James Shewey]

I'm no expert, but it's hard for me to imagine how--even if reusable--rockets and satellites are cheaper than running & maintaining lines, deploying MMWave, etc.

If that's really the core reason I'd love to see the analysis.

It doesn't necessarily have to be cheaper in some cases. There are a series of Microwave towers that run Line-Of-Site repeaters between the NYSE and Chicago stock exchanges. I think they might have been installed by MCI back in the day. Nonetheless, light travels at the speed of light in a vacuum. So when light travels through a glass fiber? Guess what: It's slower than "light speed". As it turns out, our atmosphere is a less dense medium than glass fiber. So you wouldn't think it, but Microwaves are actually faster than fiber optic cable.

Enter Goldman Sachs (and their ilk) and High Frequency Trading. Those microseconds matter. In the NYSE's data center, every cage gets the same length of cable. That's to make sure each cage has the same amount of latency. Because in High Frequency Trading, if I get my trade in before my competitor I make (or lose less) money.

You know what's faster than microwave through the atmosphere? Waves in the vacuum of space. It's over 750 miles from NY to Chicago. It's only 62 miles to the Von Karman Line where you hit outer space. So that reduces time in the atmosphere to only about 124 miles of travel time. Some of the SpaceX satellites are at an altitude of 211 miles. So if, even on egress those wave's don't suddenly accelerate, that's still only 273 miles in the "slow lane" - meaning that SpaceX is faster than microwave or fiber. Long story short, Goldman Sachs and other wall street firms will pay buckets of cash for that kind of speed - even if it is more expensive. Because they stand to make more money from the high frequency trading than they would lose by not moving to the faster system in order to keep up with their competitors on Wall Street. The do this kind of trading to the tune of millions - if not billions of dollars.

SpaceX disrupts other things too: In Flight wifi (previously IntelliSat/ViaSat/Panasonic's cash cows) and that kind of communication and communication with ships at sea (those have all been historically very expensive endeavors). Cell phone companies (TMobile, AT&T, Verizon; no dead zones anymore, and data that rivals 5G), and if those pay for the cost of getting the satellites into orbit, then you even have the option to go head-to-head with terrestrial internet providers - and you can pick up the rural markets they overlooked to boot.

Even if SpaceX satellites are more expensive, they also stand to be able to make more money than traditional ISPs.

 

[Mandella]

First off, nice to see a Starlink article on Ars that is more than a cut and paste hit piece from dubious sources. I was starting to despair.

More on topic, glad they are already making money. I've been on Starlink for about two years now, and the overused term game changer totally applies. Plus the service, which even initially was vastly superior to Viasat, has been consistently improving. Obstructions and network outages (never very lengthy to begin with) are much more rare now, speed has been constantly edging up, rain fade less of a thing (although still an occasional issue, but total cutouts require a really scary towering cumulonimbus overhead).

Plus as a follow on effect, my daughter is now considering moving her family out of the city and back closer to here in the country. She has told me flatly before that needing fast internet was a major factor in where she could live, and being able to get Starlink could make the difference.

And I should point out that we're not exactly in the middle of the wilderness here, but only a few miles from town and fiber.

 

[Snark218]

As I've been reading through the comments, and thinking about this some more, I've become more convinced that Eric is telling us to read between the lines here. There is no one in space journalism with more/better SpaceX insider access than Eric. No one. One of the things that means is that he has all sorts of off-the-record info that he cannot report (yet). This is true not just of SpaceX, but of many other areas of the industry. Pay attention when Eric does some of his long-form retrospective stories - he often calls out having "heard" about various things years before. Whenever you see that kind of comment, it's shorthand for "yeah, I knew about it, but I couldn't say anything before now".

As I mentioned in a reply to DistinctivelyCanuck, Eric's manner of reporting this story (i.e. unironically, without a ton of point-counterpoint expansion) is pretty instructive.

I'd have agreed with you right up to the doers vs. checkers op-ed, which really badly shredded my trust and confidence in his willingness to ask SpaceX tough questions or honestly and fully account for its failings, real or potential.

And I don't actually agree, in any case, that a seemingly uncritical bit of reporting should be assumed to be winking and nudging at secret knowledge. If there's a range of perspectives on whether Starlink makes money, that's relevant context to acknowledge even if the author knows something off-record.

 

[henryhbk]

To me the key advantage of SpaceX over all the others has been manufacturin, as in cranking rocket motors out like ford cranks out engines, but if you cranked out engines like ford but had need for launching at the speed say f1 race cars are run, you are going to have a giant pile of motors somewhere. And just because SpaceX could offer launches significantly cheaper to the world, unlikely customers were going to crank satellites on a moment’s notice so there would be zero demand for rapid manufacturing allowing iterative development. Starlink happens to not only accomplish all that, but also is a useful service on its own. I realize others have had this view before, but since this is somewhat devolving into a referendum on musk’s mental state (poor) and Eric’s journalistic integrity (high) seems reasonable to being it back to the actual topic.

 

[@inemesitaffia]

Isn't this Brodkins beat?

Relying on analysts speculations? Hmm.

We have concrete numbers for launch ($15 to $28MM) and Terminals(under $750) but not sats or ground stations which is a real issue. Actual Internet is cheap once you buy in bulk. The ground operator (if it ain't you ) will give you a wave, tunnel or dark fiber to a POP nearby with your monthly opex

For me it's not Mars or whatever Elon's dreams are, it's that the company is taking in more money than out so the Internet keeps working. SpaceX declared a 2023 profit. Free Cash Flow. Not EBITDA

OneWeb (Wyler) made claims about Sats and Terminals he couldn't deliver. OneWeb Sats are nearly 2MM each and I've never seen a Ku band terminal that claims to be under $10k. I've seen 25k in reality. Something that actually surprises me since even 5G MmWave capable client systems can cost between $100 to $300 for all the radios with far fewer elements than what SpaceX delivered

Gateways might cost as much as 5 million each or at least that over 5 years (estimate via number of dishes in community gateway) E band is ready

Forget Quilty and Payload and wait for the WSJ to give you the numbers from insiders around this time next year.

These Quilty projections are lower than what Bloomberg or WSJ claimed (10Bish) but are more realistic to me. They are also behind 2015 projections even if you adjust the start date to when service started(understandable because terminal subsidies were high).

But in the end this is the fastest ever growing sat provider and that's impressive especially with COVID and the electronics parts shortage. There's things they should be doing to attract enterprise clients they haven't. Some things I can point out off hand are official or sanctioned DC inputs, third party antenna and smaller first party aircraft antenna for smaller planes. Phone Sales even if it's premium rate numbers like Amazon had, Fixed IP addresses , L2 and MPLS without extra hardware, proper invoicing and payments other than cards (I know resellers can handle this), L1/ L2 dish teaming to minimize outages like other providers do & allow business customers to have one dish tied to one constellation and one to the other. (Suspect this happens in Aviation)

For customers in Africa they can offer night plans for 30—50% less. 11 pm to 7 am? but it's already cheaper than alternatives and suggesting lower ARPU in areas where it's lowest is kinda <<<

I'm tired and done with my rambling. I wish them success

 

[phoenix_rizzen]

You're confused. In BC we have fibre all over the place in very rural areas. I live in a community of 8000 people and we have GB service. I've had fibre for about ten years now. Your providers in the US are all socialist waiting for the government to pay to have their services built out to your MAGA hicks.

It's crazy the locations that have fibre Internet in BC, and even crazier the ones that don't.

The lower mainland has multiple fibre ISPs and several fibre wholesalers that lease access to ISPs. Gigabit links are pretty much the norm from Abbotsford to Vancouver. Even the Island has fibre.

Interior of BC, there's gigabit fibre available in tiny little places like Lee Creek and Scotch Creek. The Kootenays have municipal fibre networks in multiple towns.

But, for every little town with fibre, there's probably two or three without. Our little village of 3500 has 2 schools with fibre Internet; Telus (as mandated by the Ministry of Education) ran fibre to the 2 schools, but didn't expand it to any homes or businesses. We have a cable ISP that offers gigabit download, but the upload speeds are horrendous (25 Mbps max, only available on 500+ Mbps plans) and the infrastructure around town is horrible. Neighbourhoods with buried utilities have good service; neighbourhoods with overhead utilities have frequent cable outages.

When we moved out here, we put up Dishy in our backyard and had amazing Internet (200 Mbps down, 25-35 Mbps up). Two years later, every third house or so has a Dishy on the roof, and throughput has tanked (25-45 Mbps down, 5-15 Mbps up). We moved to the cable ISP in January, as it was almost impossible to watch Netflix in the evenings on Starlink (with 2 kids also streaming or playing games), and video calls on Zoom/Teams during the day were getting choppy when working from home.

Would be nice if Telus continued expanding their fibre network into the rest of the towns around BC.

 

[blackhawk887]

Yes, this is not journalism. It smells like SpaceX marketing, right on time when more questions are starting to get asked about LEO pollution.

(Also the way Berger bigfooted coverage of the Starliner test by posting a “Feature” article hours earlier headlined to cause confusion about its “loss” was pretty egregious. The actual Starliner article was assigned to the other space reporter, published closer to the launch, got preempted and relegated downpage.)

Starliner was a loss. Boeing lost the race, and they lost well over a billion dollars. You're just misreading the headline.

 

[blackhawk887]

I'd have agreed with you right up to the doers vs. checkers op-ed, which really badly shredded my trust and confidence in his willingness to ask SpaceX tough questions or honestly and fully account for its failings, real or potential.

And I don't actually agree, in any case, that a seemingly uncritical bit of reporting should be assumed to be winking and nudging at secret knowledge. If there's a range of perspectives on whether Starlink makes money, that's relevant context to acknowledge even if the author knows something off-record.

There is absolutely no shortage of critical coverage of SpaceX, most of it wildly overblown or severely deficient in facts or context.

There's no need or reason for Eric to be like that. He pretty much tells it like it is.

 

[Iceroadman]

Do you have a point?

Yeah, the cry babies whinging about X should give it up. If they don't like X then they don't have to use it. X is paying it's top contributors with dividends so there's that.

 

[Iceroadman]

It may not be cheaper, but the traditional companies aren't doing it. I'm rural - 5 miles from a traffic light, poor cell reception (on 4th gen equipment) and my local DSL lines are oversubscribed. DSL is not available for my house. My local WISP said I might get coverage from them, but I'd have to pay to put their receiver in a tree for $500 to test it out.

Yes, CLink could run fiber down the road, but it would cost millions they'd never recover. My neighbors who are grandfathered in to DSL are OK paying $50 a month for service that can't reliably stream Netflix, and are unwilling to drop $120 for StarLink . There's no way they'd pay $80 for fiber. The installation would cost at least as much as SL because most of the old phone lines are direct buried and there's no conduit to slide fiber through (a short driveway here is 100', and 500' is common).

In short, the population density for real internet isn't high enough to ever pay out. They just get federal subsidies and keep the money.

SpaceX & SL may pay more for the system, but once it's up it can reach everywhere. Mongolian yak herder? Internet. Guy on a sailboat? Internet. Millions of rural folks like me? Internet.

When fully scaled, it's probably cheaper per household in low density areas than a wire. The last mile is expensive, but the last few hundred feet (road to house) is really expensive.

They run fibre for thousands of kilometers here in BC without having conduit in the ground just like they did decades ago with coax.

 

[James Shewey]

whilst I'm not disagreeing with you, you're neglecting to mention maintenance. How long will the masts and hardware last? How much does replacement cost? How does that compare to Starlink.

Well, Voyager 1 and 2 are still phoning home after 46 years, so if well designed they ought to amortize pretty nicely. If well designed they ought to last long enough that replacement costs are justified as upgrades because the technology on them is growing obsolete. They may not be well built, but clearly we have that capability. Honestly, their orbits may degrade or they may run out of gas before the hardware fails.

 

[Iceroadman]

Yes to all that and the nature of leo sat coverage is that in order to cover that KY resort, you end up covering a swath of the entire planet. The cost to add entire countries is basically regulatory effort.

I'm not surprised if starlink is already profitable but even if it isn't it's certainly going to be. If I had an opportunity to invest in this business I'd jump at the chance. That opportunity is more important than whether this analyst is precisely right about his revenue and expenses guesses, however well informed.

DXYZ has about 50 million invested in SpaceX amongst other ventures.

 

[@inemesitaffia]

Not unless Musk takes his usual sledgehammer from Twitter and Tesla to it.


TBF that all depends on the reusable rocket business which none of them could have known/predicted.

Spangelo and Longmier run a tight ship and are very aligned. The guy who wasn't runs Kuiper

 

[Echohead2]

Come on think about it for a few seconds. How much is a launch? SpaceX owns the rockets! Indeed, the real breakthrough will be totally reusable Starship Cargo variant. When that baby finally takes off and lands (without blowing up); civilizational level game changer. It'll probably slash launch costs in half or more. Anyway, SpaceX is still a private company, you can be sure it's investors have all the stats you want.

I don't think most people really grasp what kind of a game changer Starship will be. 50k vs 200k lbs (60k to Geo). The cost per kg is already much lower than it used to be. It used to be around $7k/kg. Now falcon 9 it is around $3k/kg, Falcon 9 heavy...$1.5k/kg. Starship...$0.2k/kg. Or about $100/lb It will enable all kinds of things.

 

[James Shewey]

You're confused. In BC we have fibre all over the place in very rural areas. I live in a community of 8000 people and we have GB service.

No, that's not the kind of rural we are talking about. Think more rural. As in "I live in a community of 5 people. They all live in my house" rural. Farmhouses and ranches with dozens to hundreds of acres between them. It's more common than you think.

My folks live in a neighborhood of maybe 30 houses on acre parcels (where I grew up). They never got cable. Never got DSL. (too far from the node for that). There's a fiber line maybe 200 yards from their house, but no traditional internet. Beyond that neighborhood they are in are farms and ranches on dozens to hundreds of acres for a few miles. They are only maybe 5 miles from a town in either direction, but they have to use fixed point wireless for internet and so do their neighbors. They were stuck with Dial-Up well past the mid oughts. Far longer than any human should have had to suffer that fate.

But your neighborhood of 8000? Of course you get internet. You're big enough.

 

[GodFather]

No, that's not the kind of rural we are talking about. Think more rural. As in "I live in a community of 5 people. They all live in my house" rural. Farmhouses and ranches with dozens to hundreds of acres between them. It's more common than you think.

My folks live in a neighborhood of maybe 30 houses on acre parcels (where I grew up). They never got cable. Never got DSL. (too far from the node for that). There's a fiber line maybe 200 yards from their house, but no traditional internet. Beyond that neighborhood they are in are farms and ranches on dozens to hundreds of acres for a few miles. They are only maybe 5 miles from a town in either direction, but they have to use fixed point wireless for internet and so do their neighbors. They were stuck with Dial-Up well past the mid oughts. Far longer than any human should have had to suffer that fate.

But your neighborhood of 8000? Of course you get internet. You're big enough.

Correct, I'm 2 miles up a private country road that has a total of 40 houses on it, most of them actually content without home internet. I'm 7 miles from the nearest store. There is literally zero chance I'm getting any sort of wired internet access out here. This is rural.

Everyone talking about how Canada managed to pull it off seem to be forgetting that 90% of the Canadian population live within 150 miles of the US Border. That makes a significantly smaller surface area that needs to be covered with fiber than the entirety of the continental US.

 

[@inemesitaffia]

Having worked in multiple startups over the years: and having seen more than one Marketing VP talk about variations of "if we need a market report to justify what we're doing , we can get one" or "hey, I can make sure we're an award winner"
(I worked at a startup that was poorly funded and trying to compete in a segment that Cisco already owned: and we got some "industry analyst" report as "most promising" Total BS)

As such: forgive my cynical take on this:

It is pretty easy to dismiss an 'industry analyst' and their "well, I really have no visibility into the black box that is SpaceX financials: but I can make some guesses and try to make it look real."

Has anyone outside of SpaceX seen a BOM for any of the satellites?
So: how the f*** can you say you know what the cost of the thing is?

What do you know? Ok, the approximate launch cost.
Do you know how many terminals have actually shipped? No. (SpaceX being private does not have to publish quarterly reports, right? so do we know, precisely, how many terminals have been sold?)
We've seen some teardowns on the terminals: so we can come up with a +/- 20% BOM on those.
Do we know how many ground stations that Starlink has? with precision? The cost of those ground stations? CapEx? OpEx?

Honestly: 99.9% of the time i love Eric's reporting: but I really struggle with seeing that this 'analyst report' has any credibility.

The only thing we don't know reasonably well is the cost of Ground Stations and Sats. Everything else is something we have a good ballpark figure for.

We know who makes the GS antenna/radome (minus transmitter and modem) and it's power consumption(actually rating); we know the minimum sized transformer recommended.

We know the ion thrusters were replaced with a much cheaper gas.

We know SpaceX declared a profit.

Most objections to the tale of starlink profitability are that of antenna subsidies, those are much lower than before and near zero in the US. Or that SpaceX hasn't yet paid off the initial investment for starlink which isn't in question. And isn't an issue since it's equity rather than debt.

With the steep subsidies for equipment gone as of last year (they should stop allowing Ukranians to buy the equipment at a loss, they've done enough), is SpaceX going to make more money in 2024 than they spend on starlink capital and ops? I say at least $1 more

 

[Echohead2]

So it strikes me as very weird that this industry analyst has come to precisely the opposite conclusion. I don't know which one is wrong, but the timing is enough to raise my hackles, and it's not beneath SpaceX to pull an Adrian Dittman. At the vey least, there are analyses that reach the opposite conclusion, and they merit mention in an article about analysis of Starlink and SpaceX's financials - at least, one that isn't just uncritically showcasing good news about SpaceX. But I guess that's just me being a tedious checker, as usual.

I don't know which one is correct...but what I do know is if they aren't profitable now, they will be. It might be this year like the guy in this report, or next year. But they have a product that absolutely has a market and in reality, they could charge a lot more since it is a captive audience. It is a technology that is game changing. Take cruise ships...their wifi is notoriously horrible, the few that have upgrade to starlink is getting reports of pretty darn usable wifi. Plus remote people, companies, etc.

They announced 2.7mil subscribers as of April. 2.7mil x 12 x $120 is $3.88B. And they are still growing (so I definitely think their estimate of $6.6b this year is high. It has gone from 1mil to 2.7 in about 16 months. I think 5mil by end of 2025 should be relatively likely.

 

[Echohead2]

I remember broadband over power lines hype from Bush jr days. That was exciting..
Until it never happened.

I remember that too. It was going to be the future....then nothing. You can buy devices and do it around your house, but that's about it AFAIK

 

[veldrin]

Having worked in multiple startups over the years: and having seen more than one Marketing VP talk about variations of "if we need a market report to justify what we're doing , we can get one" or "hey, I can make sure we're an award winner"
(I worked at a startup that was poorly funded and trying to compete in a segment that Cisco already owned: and we got some "industry analyst" report as "most promising" Total BS)

As such: forgive my cynical take on this:

It is pretty easy to dismiss an 'industry analyst' and their "well, I really have no visibility into the black box that is SpaceX financials: but I can make some guesses and try to make it look real."

Has anyone outside of SpaceX seen a BOM for any of the satellites?
So: how the f*** can you say you know what the cost of the thing is?

What do you know? Ok, the approximate launch cost.
Do you know how many terminals have actually shipped? No. (SpaceX being private does not have to publish quarterly reports, right? so do we know, precisely, how many terminals have been sold?)
We've seen some teardowns on the terminals: so we can come up with a +/- 20% BOM on those.
Do we know how many ground stations that Starlink has? with precision? The cost of those ground stations? CapEx? OpEx?

Honestly: 99.9% of the time i love Eric's reporting: but I really struggle with seeing that this 'analyst report' has any credibility.

I think the real takeaway is that it's likely that Starlink is already at the point where it isn't hemorrhaging money. It's questionable whether there's really $600 million a year in free cash flow, but even reaching approximate breakeven on the project is a pretty big deal.

 

[@inemesitaffia]

Isn't this Brodkins beat?

Relying on analysts speculations? Hmm.

We have concrete numbers for launch ($15 to $28MM) and Terminals(under $750) but not sats or ground stations which is a real issue. Actual Internet is cheap once you buy in bulk. The ground operator (if it ain't you ) will give you a wave, tunnel or dark fiber to a POP nearby with your monthly opex

For me it's not Mars or whatever Elon's dreams are, it's that the company is taking in more money than out so the Internet keeps working. SpaceX declared a 2023 profit. Free Cash Flow. Not EBITDA

OneWeb (Wyler) made claims about Sats and Terminals he couldn't deliver. OneWeb Sats are nearly 2MM each and I've never seen a Ku band terminal that claims to be under $10k. I've seen 25k in reality. Something that actually surprises me since even 5G MmWave capable client systems can cost between $100 to $300 for all the radios with far fewer elements than what SpaceX delivered

Gateways might cost as much as 5 million each or at least that over 5 years (estimate via number of dishes in community gateway) E band is ready

Forget Quilty and Payload and wait for the WSJ to give you the numbers from insiders around this time next year.

These Quilty projections are lower than what Bloomberg or WSJ claimed (10Bish) but are more realistic to me. They are also behind 2015 projections even if you adjust the start date to when service started(understandable because terminal subsidies were high).

But in the end this is the fastest ever growing sat provider and that's impressive especially with COVID and the electronics parts shortage. There's things they should be doing to attract enterprise clients they haven't. Some things I can point out off hand are official or sanctioned DC inputs, third party antenna and smaller first party aircraft antenna for smaller planes. Phone Sales even if it's premium rate numbers like Amazon had, Fixed IP addresses , L2 and MPLS without extra hardware, proper invoicing and payments other than cards (I know resellers can handle this), L1/ L2 dish teaming to minimize outages like other providers do & allow business customers to have one dish tied to one constellation and one to the other. (Suspect this happens in Aviation)

For customers in Africa they can offer night plans for 30—50% less. 11 pm to 7 am? but it's already cheaper than alternatives and suggesting lower ARPU in areas where it's lowest is kinda <<<

I'm tired and done with my rambling. I wish them success

A lot of plane providers can't deliver sat Internet until you're in the sky. SpaceX can sell ground only Internet for resale via already existing antenna to the IFC provider if the third party antenna are capable

 

[SirOmega]

Oh man, this is so much harder than you imagine. We've been pumping literally billions into the US rural internet industry and what do we have to show for it? A bunch of PtMP/PtP wireless providers with limited coverage? I just stayed in a resort in Eastern Kentucky this week (so, Appalachia, for those unfamiliar) that had ViaSat because, while there's a local wireless provider in the region, they couldn't justify the cost to run 5 miles into the mountains where the resort is. I assume that's because it would have required repeaters every .5-1 mile, and that would mean access to power, which means a meter, which means costs, plus all the construction... and all for one measly subscriber. There are a lot more of these situations than you realize.

You also should research mmWave coverage... it's a solution exclusively for very dense, urban settings, and literally cannot cover long distances.

You basically have this completely backwards in your head. Satellite is fantastic for covering difficult to reach areas (say, the last 8% of population) and way, way cheaper than running cables to each. It's not the best technical solution (I used to manage a university network and cannot tell you how many problems are solved by using a wired connection instead of WiFi), but it's a reasonable compromise with enough benefit to be worthwhile.

About 15 years ago, Australia put out their national broadband plan and they got a lot of flack for using satellite to reach that last few percentage of rural and very rural populations. At the time it got pushback. Now, it seems like anything else is a dumb option. Starlink isn't fiber-fast, but it's better than 3mbps DSL over ancient copper lines, or 25mbps geostationary satellite links.

 

[@inemesitaffia]

It's all speculation and "mind-blowing" puffery, really - the breathlessness reads like Quilty has some link to Musk and/or SpaceX.

I think you should watch the recent interview about this very analysis.

Quilty and TMFAssociates and just different Telecom, Satcom, and SpaceX industry people have been very skeptical about SpaceX projections in general and were predicting a Globalstar/Iridium. And that RDOF money would be the thing to save SpaceX if it didn't go into restructuring.

Now they believe they'll be profitable but not as much as projected i.e 6 vs 10 billion in revenue for 2024.

This article people consider praise is a downer for the growth rate and may mean less shifts at their terminal factory

 

[LetterRip]

Cool story bro.
You missed one small but important point though. Youtube was launched three years AFTER Musk started SpaceX.

Yes I must have misremebered the detail of the video source - presumably a blog post or whatever and my brain merged in 'youtube'.

 

[just another rmohns]

ProdigySim said:
I love how the broadband industry has failed to provide reasonable internet to rural areas for decades, despite countless federal grants being thrown their way.

And Starlink shows up in just a handful of years providing better, more consistent service. And now they're profitable on top of that.

TBF that all depends on the reusable rocket business which none of them could have known/predicted.

TBF that is the very definition of disruptive innovation.

Some new, immature technology comes along, the incumbents scoff at it. The new technology is good enough to serve a need, starts making money, improves itself, and scales up – and the incumbents, who are invested in the old way of doing/selling/whatever, are now the dinosaurs.

In this case, the incumbents didn't even try to solve the problem. They just took grants, did just enough to look like they were doing something, and ignored it.

(Verizon was probably the worst: They took money meant to build out wireless internet to underserved rural areas and used it to build their urban FIOS network and upgrade the core cellular network. They never did expand service, and never had to give back any of that Federal tax money.)

 

[@inemesitaffia]

Also, it was exactly one month ago that many outlets ran this story:

https://www.theverge.com/2024/4/11/24127509/is-starlink-actually-profitable
https://www.inc.com/chloe-aiello/ho...rlink-to-achieve-profitability-at-spacex.html
So it strikes me as very weird that this industry analyst has come to precisely the opposite conclusion. I don't know which one is wrong, but the timing is enough to raise my hackles, and it's not beneath SpaceX to pull an Adrian Dittman. At the vey least, there are analyses that reach the opposite conclusion, and they merit mention in an article about analysis of Starlink and SpaceX's financials - at least, one that isn't just uncritically showcasing good news about SpaceX. But I guess that's just me being a tedious checker, as usual.

Read the articles. Not the headlines.

Terminal subsidies - gone.

Startup capital was equity, not debt so it doesn't matter yet if it's not earned back

30 million users projection? They were late to launch and slow walked customer acquisition to limit losses due to subsidies. You'll have to move the starting line to figure out if they are growing fast enough. I don't believe they'll get it but it's not required for profitablity.

This is x company is losing x money making cars when the loss is the sunk cost of factory construction and each car itself and the operations are profitable.

 

[blackhawk887]

He pretty much tells it like you want to hear it, you mean? Because there's lots of perfectly valid criticism of SpaceX, and plenty of analysis that is fair and nuanced but not optimistic that the line goes up forever. And sorry, after the November article, I do not agree that Eric is going to tell it like it is, unvarnished. I think he's going to be as optimistic as he can be without actively misleading his readers, and I don't think he's gong to be as skeptical about what he hears and sees as I think anything connected to Musk merits.

Got some examples of that fair and nuanced coverage? What I've seen outside the dedicated space beats is rarely worth the paper it's printed on.

 

[@inemesitaffia]

I'd just be happy with its government clients threatening breach of conflict due to Elon's drug use unless he's shitcanned entirely from the company.

those terms don't mean what you think. SpaceX is compliant with whatever the government requires and the rest is waived

 

[xoe]

Gotta hand it to Musk. He took his $180 million cashout from Paypal. Use some in SpaceX, used Some in Tesla, Used a little leftover in more far-out ventures like Solarcity, The Boring Company, xAI..... some have valuations in the tens of billions, others are loss leaders. SpaceX is proving a gold mine though, after Tesla (facing lots of headwinds).

He sure does make the rockets launch on time!

 

[Echohead2]

We'll see. I don't know that the audience is as captive as you seem to think and I don't agree that cruise ships and remote workers with no broadband represent big chunks of a potential Starlink customer demographic mix. But you may be right, or it may be profitable on a longer but still reasonable timeline. I don't want it to not succeed. I just want a more critical and multifaceted look at the question than this.

Well...rural customers are pretty captive. Meaning people without any other real option (like multiple posts on here). There are MILLIONS of people like that. They already have 2.7mil subscribers. There's probably 10mil in the US alone, I think that is a very conservative low-ball estimate. That alone would be $14B/year in revenue.

 

[bryanlarsen]

I'm no expert, but it's hard for me to imagine how--even if reusable--rockets and satellites are cheaper than running & maintaining lines, deploying MMWave, etc.

If that's really the core reason I'd love to see the analysis.

It looks like they're spending about $1000 per customer (+ the terminal cost, but the customer pays for that). Running broadband costs about $20,000 per mile from what I heard. And for me the definition of rural means multiple miles per customer.

I have no idea how MMWave compares in terms of cost.

 

[eyesoars]

I'm no expert, but it's hard for me to imagine how--even if reusable--rockets and satellites are cheaper than running & maintaining lines, deploying MMWave, etc.

If that's really the core reason I'd love to see the analysis.

It's all scale. And it covers the entire Earth with the (large-ish) number of satellites, where comparable ground facilities can't even exist (how do you put up a microwave tower in the middle of the Indian Ocean?). So it's not great for high-population-density centers (underserved), but everywhere else it costs essentially nothing to add service. There are lots of people and industries that can and will pay large sums to have that kind of service (marine, aviation, rural and remote areas).

 

[The Dark]

You're confused. In BC we have fibre all over the place in very rural areas. I live in a community of 8000 people and we have GB service. I've had fibre for about ten years now. Your providers in the US are all socialist waiting for the government to pay to have their services built out to your MAGA hicks.

By the US Census Bureau's definition, you'd be urban. An urban area is any census tract with at least 2,000 residences or 5,000 residents. Rural is any tract that's not urban. That's actually a significantly stricter criterion than the 2010 Census, where an urban census tract was any tract with at least 2,500 residents.

 

[eyesoars]

Come on think about it for a few seconds. How much is a launch? SpaceX owns the rockets! Indeed, the real breakthrough will be totally reusable Starship Cargo variant. When that baby finally takes off and lands (without blowin g up); civilizational level game changer. It'll probably slash launch costs in half or more. Anyway, SpaceX is still a private company, you can be sure it's investors have all the stats you want.

Estimates vary a great deal, but probably on the order of $20M for a Falcon launch. An upper stage is deorbited & discarded, plus booster recovery and refurb. That's probably 10-20% of any other launch provider's cost, which itself is a huge game changer.

When Starship/SuperHeavy arrives, the cost should be comparable fairly quickly if not immediately on a per-kg basis, and rapidly improve from there towards parity on a per-launch basis as second stage re-use develops.