Starlink : Markets and Marketing

[ChrisWilson68]

Is the US military going to be satisfied with running combat-related bandwidth on a civilian-owned network?  I can see using it for basic peacetime comms, but during a war I would think they would want their own dedicated constellation.

Currently, the US military takes bandwidth on GEO sats, even those that are not US-owned.  So it shouldn't be a problem for Starlink.

Yeah, and they also pay for bandwidth on Iridium, so they're already doing it for LEO constellations too, and have been doing it for decades.  The Pentagon interest in Iridium was what kept the system running after the original Iridium bankruptcy shortly after they launched their first-generation system.

[M.E.T.]

https://www.itnews.com.au/news/nbn-co-may-try-buying-into-a-spacex-like-service-for-future-sky-muster-537274

The article linked above discusses Australia’s National Broadband Service potentially getting out of owning their own satellites and becoming a Starlink customer instead. Interesting. I think the potential markets for Starlink are set to expand even more once it is operational.

[Eka]

On hardware issues, 5 years after fix rolled in, they would all be upgraded and replaced...

[PADave]

All military communications would be encrypted with military grade encryption. So it's not like access to a ground terminal or even a satellite would expose the content. Granted that doesn't alleviate all concerns but it makes Starlink (and the other mega-constellations) too good to pass up. I'd be shocked if the military doesn't end up using all of them to some degree.

[TheRadicalModerate]

All military communications would be encrypted with military grade encryption. So it's not like access to a ground terminal or even a satellite would expose the content. Granted that doesn't alleviate all concerns but it makes Starlink (and the other mega-constellations) too good to pass up. I'd be shocked if the military doesn't end up using all of them to some degree.

Security is a lot more complicated than just ensuring that all the traffic stays secret.  If somebody can mess up your network so that traffic doesn't get through or gets through with much lower bandwidth, that's almost as bad as the enemy being able to read your mail.

[PADave]

All military communications would be encrypted with military grade encryption. So it's not like access to a ground terminal or even a satellite would expose the content. Granted that doesn't alleviate all concerns but it makes Starlink (and the other mega-constellations) too good to pass up. I'd be shocked if the military doesn't end up using all of them to some degree.

Security is a lot more complicated than just ensuring that all the traffic stays secret.  If somebody can mess up your network so that traffic doesn't get through or gets through with much lower bandwidth, that's almost as bad as the enemy being able to read your mail.

I disagree, I worked in contingency planning and operations for the USAF, and the enemy reading your communication is several orders of magnitude worse than a denial of service (DOS) attack. Both are bad but DOS is a threat for all forms of communications and an argument for using all available friendly communications.

Again, I would be surprised if the US Military doesn't use Starlink to some extent. They will very likely get deeply involved with SpaceX and the security to eliminate as many security concerns as possible but in the end they will be using Starlink eventually provided SpaceX delivers a viable product.

[RedLineTrain]

I recall that the US military was even using unencrypted signal on the GEO birds for gunship video in Afghanistan.  So in a pinch, the military is sometimes less picky on some communications.

[thirtyone]

I recall that the US military was even using unencrypted signal on the GEO birds for gunship video in Afghanistan.  So in a pinch, the military is sometimes less picky on some communications.

Eh, let's hope they don't repeat that fiasco...

I think one of the earlier articles, the military said they were planning on testing other megaconstellations as well. A great distributed architecture would be to subscribe to all of the available constellations, with or without network priority, and abstract the actual data transport layer so that any critical data is routed through multiple networks to guarantee delivery. It's about time they tried some more modern architectures.

ISL availability on some constellations would probably make them a bit happier (probably difficult to have reliable ground stations in active war zones). Almost wonder if they're considering providing some funding to push that part ahead.

[FutureSpaceTourist]

https://twitter.com/thesheetztweetz/status/1224736266538930176

SpaceX has a foot in the door to offer Starlink in Australia, as the country's regulator amended its satellite operator list to include the company.

But SpaceX may face a fight with Australia's Foxtel – a venture owned by Rupert Murdoch's News Corp.

twitter.com/thesheetztweetz/status/1224737083996110849

Australia added Starlink on Jan. 24, as well as satellite operators Kepler Communications and Swarm Technologies.
cnb.cx/2vHsp7t

https://twitter.com/thesheetztweetz/status/1224749458274938881

Matt Botwin, SpaceX director of global satellite government affairs, also outlined Starlink's target market:

Everyone, everywhere.

"Services for residential, commercial, institutional, governmental, and professional users throughout Australia." cnb.cx/2vHsp7t

[FutureSpaceTourist]

twitter.com/9newsaus/status/1224908480626200581

Elon Musk could soon fix all of Australia's internet woes. #9News

https://twitter.com/elonmusk/status/1225192950956744704

Starlink will be great for any sparsely populated areas with expensive or little to no connectivity!

[cebri]

If a Starlink sat has no subscribers underneath it, it generates $0. it's a wasted resource. it would make sense to reduce the price in those countries so that the sat makes something.

That's not accurate, SpaceX will make money as long as the price paid by a suscriber is higher than the marginal cost of giving him/her access to the constellation. SpaceX cannot lower the prices beyond a certain point, no matter where the customer lives.

[envy887]

If a Starlink sat has no subscribers underneath it, it generates $0. it's a wasted resource. it would make sense to reduce the price in those countries so that the sat makes something.

That's not accurate, SpaceX will make money as long as the price paid by a suscriber is higher than the marginal cost of giving him/her access to the constellation. SpaceX cannot lower the prices beyond a certain point, no matter where the customer lives.

Sure, but that marginal cost is very low, since it's just the cost of operating a ground station plus data over that station's fiber link. And one ground station can serve tens or hundreds of thousands of subscribers.

[tbellman]

Sure, but that marginal cost is very low, since it's just the cost of operating a ground station plus data over that station's fiber link.

Plus the cost for that data from/to the rest of Internet...  SpaceX will need to buy transit from other network providers (like NTT, Telia Sonera, Cogent, et.c).  They charge money based on how much bandwidth you use.

And even if you peer with some content providers (e.g. Google, Akamai, Netflix, et.c) with no money changing hands for that peering (they make money from the end users instead), upgrading the network equipment for that peering to support the higher bandwidth, is not free.

[envy887]

Sure, but that marginal cost is very low, since it's just the cost of operating a ground station plus data over that station's fiber link.

Plus the cost for that data from/to the rest of Internet...  SpaceX will need to buy transit from other network providers (like NTT, Telia Sonera, Cogent, et.c).  They charge money based on how much bandwidth you use.

And even if you peer with some content providers (e.g. Google, Akamai, Netflix, et.c) with no money changing hands for that peering (they make money from the end users instead), upgrading the network equipment for that peering to support the higher bandwidth, is not free.

That's what "plus data over that station's fiber link" means. That's not very expensive.

[Lar]

Here we go. Astronomers doing economic analysis of Starlink assuming it will be sold worldwide for $60 USD.

The satellites will be over Africa by orbital necessity, at no additional cost, since SpaceX needs to serve the populated, affluent areas in the northern hemisphere.  Therefore every dollar they can get from Africa is straight marginal profit.  Prices will be set by supply and demand.  The supply of bandwidth is the same everywhere on Earth (minus the poles), so for most of the southern hemisphere where the demand is low the prices will be low - I suspect much less than $60/month.

There is some price point they can't go below, presumably, since there is a cost to service a subscriber, and there is a base station cost, but I suspect these total a LOT less than 60/month...  so yeah, they will price service at the market clearing price.

[ZachF]

Here we go. Astronomers doing economic analysis of Starlink assuming it will be sold worldwide for $60 USD.

The satellites will be over Africa by orbital necessity, at no additional cost, since SpaceX needs to serve the populated, affluent areas in the northern hemisphere.  Therefore every dollar they can get from Africa is straight marginal profit.  Prices will be set by supply and demand.  The supply of bandwidth is the same everywhere on Earth (minus the poles), so for most of the southern hemisphere where the demand is low the prices will be low - I suspect much less than $60/month.

There is some price point they can't go below, presumably, since there is a cost to service a subscriber, and there is a base station cost, but I suspect these total a LOT less than 60/month...  so yeah, they will price service at the market clearing price.

Also much lower than it's competitor, OneWeb, who is probably paying 10x what SpaceX is terms of installed $/GB.

[RocketGoBoom]

All military communications would be encrypted with military grade encryption. So it's not like access to a ground terminal or even a satellite would expose the content. Granted that doesn't alleviate all concerns but it makes Starlink (and the other mega-constellations) too good to pass up. I'd be shocked if the military doesn't end up using all of them to some degree.

If SpaceX is planning 30,000 satellites for Starlink but can achieve global coverage with under 1,000 satellites, then I am sure it is feasible to carve out 1,000 to 2,000 satellites purely for a US Military only constellation network. If the US Military thinks that their data cannot ride in the same satellites as commercial data, then the US Military will likely pay for exclusive use of their own smaller constellation.

[ChrisWilson68]

All military communications would be encrypted with military grade encryption. So it's not like access to a ground terminal or even a satellite would expose the content. Granted that doesn't alleviate all concerns but it makes Starlink (and the other mega-constellations) too good to pass up. I'd be shocked if the military doesn't end up using all of them to some degree.

If SpaceX is planning 30,000 satellites for Starlink but can achieve global coverage with under 1,000 satellites, then I am sure it is feasible to carve out 1,000 to 2,000 satellites purely for a US Military only constellation network. If the US Military thinks that their data cannot ride in the same satellites as commercial data, then the US Military will likely pay for exclusive use of their own smaller constellation.

If they did carve out a subset of the satellites for the US military but still had the satellites under the same control system and using the same ground stations, the only real advantage would be that they wouldn't be vulnerable to being crowded out by other users, either legitimate traffic or the fake traffic of an attack.  But you could get the same advantage by sharing satellites with other users but using software to give Quality of Service (QoS) guarantees to the US military.  This is a well-known technique that's been used for decades.

Having military-only ground stations could provide additional security, but this wouldn't require dedicated satellites.  With inter-satellite links, at any given time some of the satellites could be talking to the US military ground stations while others talked to other ground stations.  It could just be done dynamically the same as with any other ground stations, just with military traffic routed to those particular stations.

I think the only way carving out a subset of the satellites for the US military would really provide a benefit you couldn't get another way would be if you put those satellites under the direct control of the US military.  Then they could potentially have better security protecting the control of those satellites than those of the civilian side under SpaceX (at least theoretically better security, or the illusion of better security -- it's an open question whether that security would end up being better).

[RocketGoBoom]

Which competitors in the satellite data business are most vulnerable to Starlink competition?

In my opinion, it is companies like Viasat and HughesNet for the USA. They have GEO satellites, lousy service with bandwidth caps and high latency. Their customers hate them, but have no other options.

Viasat (stock symbol: VSAT) and HughesNet (Echostar owner: SATS).

Viasat looks most vulnerable.

1) They have only 4 satellites, some of them over 10+ years old.
2) Huge debt ($1.9 billion) relative to their revenue, junk rated bonds 5 notches below investment grade.
3) Tiny profits / very thin margins, now even before Starlink starts offering service.
4) Horrible service speeds, high latency.

Frankly if a high debt company with high fixed expenses loses 20% to 50% of their market, most of the time that company is going to end up in bankruptcy.

I have no idea if Viasat or others are going to file for bankruptcy in the future because of the coming competition due to Starlink, OneWeb or even Kuiper if Amazon catches up. But it seems to me like a huge chunk of Viasat, HughesNet, EutelSat and SES revenue is at risk of being taken by the new LEO competitors.

Just my opinion. I am shorting some of those legacy satellite operators with the expectation that this is going to crush their profit margins and capture a piece of their existing revenue.

Looking at the 1 year chart for VSAT (attached), it looks like the high happened at $97 in May 2019 when Starlink launched the first 60 satellites. Then it has been all downhill since then.

I suspect when Starlink/SpaceX launched 60 satellites, the big money woke up to the fact that competition is coming, it is funded by Google, Elon Musk, Fidelity and a few other deep pockets, and this train is going to run over the legacy companies in this market. It would not surprise me to see this trend continue and VSAT stock fall to under $10 per share.

Full Disclosure: I am short VSAT stock.

[Lar]

Somebody needs to ask EM what expected Starlink data rates are. I doubt there will be an answer as that could help competitors.

Bigger question is how much data you can use before being deprioritized. Current consumer satellite internet plans are about 50 GB every month before deprioritization.

Viasat user here, on the "unlimited" tier at 100 GB for about $155.00 a month. And unlimited is in quotes because "after that threshold your traffic may be prioritized below other users" translates to "sharp throttling starting at 100.01 GB."


Starlink needs to beat or meet that for me to sign up, but I'm thinking that won't be too hard...

HughesNet user here. I do not believe it will take much to have all existing satellite internet users to switch to Starlink. I do not have any experience with Viasat but looking at their offerings it is very similar to HN. The day Starlink becomes available to me is the day I make the switch. HN is very costly, with low speeds, low usage limits, and high latency.

What I am seeing is a possible progression.

Start with a small base of customers who are chomping at the bit to switch because they are on sat. These customers can be the "beta" set that are used to refine process.  And build up installer networks

Next up convert the high value comcast users of the developed world. They will possibly need more handholding, possibly less. This is a larger base.

Finally, do the hotspot in a box usage model for the lower revenue customers in the third world. But if selling to enterpreneurs who resell, the price can still be fairly high, since it will get sliced up 10 or 100 ways. There is more fixed cost capital investment there, since power (panels and batteries and etc) might need to be provided. But the power can be expanded with addon kits so over time our enterpreneur is also a power provider for his/her neighborhood.

World peace through world trade. - Thomas J Watson Sr.