Starlink : Markets and Marketing

[TheRadicalModerate]

The conventional wisdom seems to be that there's so much pent-up demand for satellite internet services that other constellations can move in at any point.  If Starlink can't satisfy all of the demand, that seems like a reasonable conclusion.

But the way that tech companies usually freeze out competitors from market niches, even when there's almost limitless demand, is by figuring out how to leverage network effects, i.e., as the customer base grows and interacts with itself, the network becomes more and more valuable to them, so that switch costs become too high for any customer to move over to a newbie competitor.

As a rule, ISPs have a hard time making network effects work for them, except in their ability to lock up a particular locality.  That's not going to be something that Starlink can do, because anybody can plunk down a dish for a competitor's satellite system next to Starlink subscribers.

So here's the question:  Are there any network effects that Starlink could engineer into their architecture?

To some extent, I guess that eating all the bandwidth is sorta-kinda a strategy like that, but I doubt that the FCC will let that happen.  So is there some property that can be engineered into the Starlink system that makes its users less and less likely to switch as there are more and more of them?

Low-latency apps might fill the bill here, but they'd have to prevail against the network effects of applications designed to run on the vanilla Internet.  Could Starlink go vertically into the gaming or telerobotics markets?  Is there something that they could do with content distribution that would hike switch costs?

The obvious first competitor is Kuiper, but it's hard to see it scaling up without access to New Glenn, which appears to be, if not dead in the water, at least grounding itself on a sandbar and moaning softly.  If there truly is no way for Starlink to leverage network effects, Kuiper can probably afford to wait for its ride.  But SpaceX is awfully clever.  I'd kinda think that if Amazon waits too long, somebody at Starlink will think of a way to lock the market up.

[wannamoonbase]

So here's the question:  Are there any network effects that Starlink could engineer into their architecture?

To some extent, I guess that eating all the bandwidth is sorta-kinda a strategy like that, but I doubt that the FCC will let that happen.  So is there some property that can be engineered into the Starlink system that makes its users less and less likely to switch as there are more and more of them?

Low-latency apps might fill the bill here, but they'd have to prevail against the network effects of applications designed to run on the vanilla Internet.  Could Starlink go vertically into the gaming or telerobotics markets?  Is there something that they could do with content distribution that would hike switch costs?

The obvious first competitor is Kuiper, but it's hard to see it scaling up without access to New Glenn, which appears to be, if not dead in the water, at least grounding itself on a sandbar and moaning softly.  If there truly is no way for Starlink to leverage network effects, Kuiper can probably afford to wait for its ride.  But SpaceX is awfully clever.  I'd kinda think that if Amazon waits too long, somebody at Starlink will think of a way to lock the market up.

New Glenn will fly eventually, but by then Starlink will be global and flying birds with laser interlinks.  Kuiper is 2 years behind 2 and will be 4 years behind by the end of next year. 

Others will get in but SpaceX has big advantages with Starlink.  If they don't drop the ball they have a multi year lead that will be hard for others to catch up for a very long time.

I don't think many jurisdictions will want to give Starlink an advantage.  But scale and being operational will make the cost of adding each customer cheaper and cheaper.  That's their advantage, being first.

I think Starlink will be gigantic, larger than SpaceX.  Probably hundreds of billions in Market Cap with revenue and profit that can easily fund Mars vehicles.

The next couple of years will be really interesting to see how they take off with customers and how the constellation evolves once Starship is able to play birds on orbit.

[freddo411]

The conventional wisdom seems to be that there's so much pent-up demand for satellite internet services that other constellations can move in at any point.  If Starlink can't satisfy all of the demand, that seems like a reasonable conclusion.
...

So here's the question:  Are there any network effects that Starlink could engineer into their architecture?

...

Good question.   

Some possible lock in effects (some mild, some less so)

* Making the startup cost (buying the dish) a non-trivial expense, as opposed to amortizing this cost for the consumer
* Cancel fees (I don't expect SX to do this)
* Unique content deals (HBO and/or ESPN ) have kept cable going.   
* Unique hardware/SW features.   A bundled "land-line".   This could be copied by any low latency provider, but might be viewed as a "must keep" feature by the consumer.
* Bundle with a "RV starlink" device -- just a portable starlink dish.   Could by copied by other low latency sat providers

[oldAtlas_Eguy]

The conventional wisdom seems to be that there's so much pent-up demand for satellite internet services that other constellations can move in at any point.  If Starlink can't satisfy all of the demand, that seems like a reasonable conclusion.
...

So here's the question:  Are there any network effects that Starlink could engineer into their architecture?

...

Good question.   

Some possible lock in effects (some mild, some less so)

* Making the startup cost (buying the dish) a non-trivial expense, as opposed to amortizing this cost for the consumer
* Cancel fees (I don't expect SX to do this)
* Unique content deals (HBO and/or ESPN ) have kept cable going.   
* Unique hardware/SW features.   A bundled "land-line".   This could be copied by any low latency provider, but might be viewed as a "must keep" feature by the consumer.
* Bundle with a "RV starlink" device -- just a portable starlink dish.   Could by copied by other low latency sat providers

An item discussed extensively is data caching within the Starlink network for content providers like for video and music. Here for subscriber leverage over other sat networks it would be an exclusive contract for "local" data caching to improve the content subscribers experience when using Starlink. Such as the reduction of the occurrences of those annoying buffering pauses. With the explosion of content streaming providers. Subscribers are going to be picking  the ISP that offers the best average service for content delivery of the streaming content they watch/listen to.

[OTV Booster]

The conventional wisdom seems to be that there's so much pent-up demand for satellite internet services that other constellations can move in at any point.  If Starlink can't satisfy all of the demand, that seems like a reasonable conclusion.

But the way that tech companies usually freeze out competitors from market niches, even when there's almost limitless demand, is by figuring out how to leverage network effects, i.e., as the customer base grows and interacts with itself, the network becomes more and more valuable to them, so that switch costs become too high for any customer to move over to a newbie competitor.

As a rule, ISPs have a hard time making network effects work for them, except in their ability to lock up a particular locality.  That's not going to be something that Starlink can do, because anybody can plunk down a dish for a competitor's satellite system next to Starlink subscribers.

So here's the question:  Are there any network effects that Starlink could engineer into their architecture?

To some extent, I guess that eating all the bandwidth is sorta-kinda a strategy like that, but I doubt that the FCC will let that happen.  So is there some property that can be engineered into the Starlink system that makes its users less and less likely to switch as there are more and more of them?

Low-latency apps might fill the bill here, but they'd have to prevail against the network effects of applications designed to run on the vanilla Internet.  Could Starlink go vertically into the gaming or telerobotics markets?  Is there something that they could do with content distribution that would hike switch costs?

The obvious first competitor is Kuiper, but it's hard to see it scaling up without access to New Glenn, which appears to be, if not dead in the water, at least grounding itself on a sandbar and moaning softly.  If there truly is no way for Starlink to leverage network effects, Kuiper can probably afford to wait for its ride.  But SpaceX is awfully clever.  I'd kinda think that if Amazon waits too long, somebody at Starlink will think of a way to lock the market up.

There's a technique rarely used in the telecom market. Give superior service at a reasonable cost. All the customers will do a self lock-in.

[Tomness]

I always felt Google's investment of Starlink could play out with Gateway directly with Data Center to Starlink cluster with Laser Link to Starlink to User Terminal. Having litterly ~10-20sec ping maybe less for Google Services like, Movies, Music,  Games, & IOT etc. Stream 4k - 8k directly with out hugging up other Teir I, Teir II ISPs bandwidth. Maybe Google Stadia was released to early or to late.

[DreamyPickle]

No other ISP has managed to build a significant "network effect" and rules like net neutrality make several strategies outright illegal (can't privilege some services over others). Starlink's competitive advantage right now is mass-on-orbit and time-to-market.

The mass-on-orbit advantage might go away eventually but competitors to Falcon 9 have their first launches several years into the future.

The time-to-market aspect is also significant. Right now they're offering much better service than a traditional operator. If somebody else comes along in 3 years with 20% cheaper service many people just won't bother to switch.

[Nomadd]

 Once laser links are going, point to point, avoiding gateways and the internet altogether, would be a valuable feature.

[TheRadicalModerate]

An item discussed extensively is data caching within the Starlink network for content providers like for video and music. Here for subscriber leverage over other sat networks it would be an exclusive contract for "local" data caching to improve the content subscribers experience when using Starlink. Such as the reduction of the occurrences of those annoying buffering pauses. With the explosion of content streaming providers. Subscribers are going to be picking  the ISP that offers the best average service for content delivery of the streaming content they watch/listen to.

What you're looking for is "increasing returns on scale".  The content distro market doesn't really have that property.  (I also think that putting content on the birds instead of in the gateways is silly, but we thoroughly litigated that up-thread a while back.)

I've been thinking more about tele-robotics or, more accurately, tele-operation.  It's super-sensitive to low latency, and it (mostly) requires that both ends have the same QoS, which in turn strongly pushes users toward using the same ISP end-to-end.

This seems like it has exactly the increasing returns on scale that you need:  The more Starlink-enabled tele-operator-enabled applications you have, the more demand you have for tele-operators in the same (Starlink) network.  The more tele-operators you have, the higher the switch costs.  The larger the pool of tele-operators gets, the more applications appear, and so on.  The virtuous cycle (or vicious cycle, if you're a would-be competitor) is complete.

I suspect that if you do this right, the list of applications is very wide and deep.  Off the top of my head, it includes:

1) Remote driving (for cases where autonomous driving won't work).
2) Surgical and other medical hands-on applications.
3) Clearing robotic machinery glitches or corner cases.
4) About a bazillion military applications.
5) Field research.
6) Construction.

The question then becomes whether the regulatory framework forces Starlink to play nice with the other constellations.  Arguing that seems analogous to the arguments that were made for and against network neutrality, but if Starlink is enabling a green-field market, it's a lot harder to argue that they're trying to threaten the existence of the internet as we know it.

[TheRadicalModerate]

There's a technique rarely used in the telecom market. Give superior service at a reasonable cost. All the customers will do a self lock-in.

That's because it doesn't work.  People don't switch because their entire lives are entangled in their phone service, and both the prices and customer service are just barely good enough to keep the customers "surly but not rebellious".  These people aren't stupid; they know what they're doing, and why they're doing it.  It's a lot like the airline business:  everybody complains about the service, but they really only care about the cost.

[TheRadicalModerate]

No other ISP has managed to build a significant "network effect" and rules like net neutrality make several strategies outright illegal (can't privilege some services over others). Starlink's competitive advantage right now is mass-on-orbit and time-to-market.

Completely agree, and it's why this is an interesting question.

The mass-on-orbit advantage might go away eventually but competitors to Falcon 9 have their first launches several years into the future.

The time-to-market aspect is also significant. Right now they're offering much better service than a traditional operator. If somebody else comes along in 3 years with 20% cheaper service many people just won't bother to switch.

That's not what happens.  If a competitor can get a toehold, eventually they can pry the incumbent's fingers off of the market by offering better service, or filling emerging niches faster than the incumbent can.

On the other hand, that doesn't happen if you can build network effects.  That's one of the reasons why the pro-net neutrality people were so aggressive (although I disagree that non-neutral networks would have created the network effects that its advocates thought it would).  But this has to be more than just lukewarm loyalty to the brand.  There has to be something that makes it really painful to switch, and that's usually associated with having to go to much lower scale.

Just to reiterate:  If the constellation internet market will eventually saturate, there's a pretty good chance that Starlink has such a lead on everybody else that they'll be in a dominant position they get regulated out of it.  But if the market is nowhere near saturation for the next decade or so, competitors will get a foothold and the market will become as balkanized as any other ISP market--unless there are network effects.

[vsatman]

remember that the US internet market is unique. In the US, Hughes and ViaSat now have nearly 1.6 million subscribers for satellite broadband. There is nothing like this anywhere in the world. For example, in Europe, the KASAT satellite, analogue of ViaSat -1, has been operating since 2011, so the largest operator has 50,000 subscribers on it.

//Bigblu Broadband is the largest distributor of satellite broadband packages in Europe with a proven track record, as evidenced by its success as the main Gold member of Euro Broadband Infrastructure’s Preferred Partnership Programme since 2019. Bigblu Broadband has developed a well-established platform for satellite broadband, relying on a unique network of installers and resellers. The activities to be acquired by Eutelsat (BBB Europe) currently count around 50,000 subscribers across an expanding pan-European footprint which includes operations in the UK, Ireland, France, Germany, Italy, Spain, Portugal, Poland, Hungary and Greece.
https://bigblu.co.uk/eutelsat-to-acquire-european-satellite-broadband-activities-of-bigblu-broadband/

[oldAtlas_Eguy]

The Starlink sats handling of routing and resource allocation has to be handled in a netrality position without regard of the sender but can be managed based on the basic type such as file transfers, video streams, etc. This is so the more bandwidth intensive data type does not drown out the less demanding data bandwidth types. The Starlink sat to sat network is international in nature and actually has to operate without any regulation due to that it handles data for multiple country regulated (ISP like) situations simultaneously on a single sat. A countries actual network regulations basically mostly affect the gateways and UTs that are actually  located within that country and does not not  affect how Gateways and UTs in other countries are regulated. US network regulations is a overlay on the completely open basic sat to sat network in regards to the network routes to and from a UT in the US and a Gateway in the US. From a practical standpoint the US regulations actually only have primary impact on how the Gateways (US example) like how the Gateways located in the US handle and route network traffic.

So the FCC Net Neutrality regulations only affect mainly the gateways located in the US. Plus the "modem" UT located in the US also has some minor effects on what actions it can take based on FCC Net Neutrality regulations. Starlink with ISL creates a completely new wrinkle to the Net Neutrality debate since a country cannot control another country's network traffic. So Starlink ISL routing and management must operate in a neutral method. Even much more open than any US FCC regulations. This is so that no country can accuse SpaceX and Starlink of playing favorites or tampering of routing/throttling/etc in favour of one country with another countries data traffic.

SpaceX/Starlink is going to need a large staff of International Law lawyers. When a UT connects to and routes data through a Gateway in the same country everything is fine. But when a UT in one country connects to a Gateway or even a private connection between UTs in another country the headaches start.

[freddo411]

There's a technique rarely used in the telecom market. Give superior service at a reasonable cost. All the customers will do a self lock-in.

That's because it doesn't work.  People don't switch because their entire lives are entangled in their phone service, and both the prices and customer service are just barely good enough to keep the customers "surly but not rebellious".  These people aren't stupid; they know what they're doing, and why they're doing it.  It's a lot like the airline business:  everybody complains about the service, but they really only care about the cost.

Not, it's not really like the airline business ;    I can actually never fly frontier, because they are awful, and I have 5 or more other airlines to choose from.

Most of these telecom folks have (had, actually) cushy monopolies, either by gov't interference, or by lack of a LEO comsat constellation.   They are about to be given a really hard punch as SL comes online for a decent chunk of their customers.    I believe that a large segment of semi-rural customers will switch if SpaceX allows them to sign up.   

The only thing cable and GEO sat companies can do is to lower their prices and improve their service (or try political games to defend their monopoly).

[oldAtlas_Eguy]

There's a technique rarely used in the telecom market. Give superior service at a reasonable cost. All the customers will do a self lock-in.

That's because it doesn't work.  People don't switch because their entire lives are entangled in their phone service, and both the prices and customer service are just barely good enough to keep the customers "surly but not rebellious".  These people aren't stupid; they know what they're doing, and why they're doing it.  It's a lot like the airline business:  everybody complains about the service, but they really only care about the cost.

Not, it's not really like the airline business ;    I can actually never fly frontier, because they are awful, and I have 5 or more other airlines to choose from.

Most of these telecom folks have (had, actually) cushy monopolies, either by gov't interference, or by lack of a LEO comsat constellation.   They are about to be given a really hard punch as SL comes online for a decent chunk of their customers.    I believe that a large segment of semi-rural customers will switch if SpaceX allows them to sign up.   

The only thing cable and GEO sat companies can do is to lower their prices and improve their service (or try political games to defend their monopoly).

At $29 to $49 /month for cable is the primary item for Starlink to beat. That is if you can even get cable internet which many cannot. So the initial $99/month is a good starting point but they will have to lower their prices soon as in a couple of years. Like when most if not all sats are launched on Starship and all contain ISL as well as 2 or even 4X the capability for literally not much more if any additional mass or cost in sat manufacter /sat/ costs. Such that the costs drop significantly as much if not more than to 25% of the current. This would make it possible to drop prices to $49/month and offer higher bandwidth such as close to 1Gbps to its subscribers in direct competition to what would be available for most cable providers.

The price drop to ~$49/month is necessitated by competition and the need to expand the subscriber base to even create higher economies of scale that would then generate even more profits. Which also allows to maintain the position of low price leader as well as become a direct competitor to all other ISP options. This is likely to be the case in 2023 to 2024 while Starlink is still in the middle of rapid expansion of its numbers of sats on-orbit.

Eventually Starlink may offer several data rate options. A very low price option for a 100 to 300Mbps at $29/month and a 1Gbps at $49/month. As well as a possible business option at 10Gbps at a possible $500/month.

[TheRadicalModerate]

The Starlink sats handling of routing and resource allocation has to be handled in a netrality position without regard of the sender but can be managed based on the basic type such as file transfers, video streams, etc. This is so the more bandwidth intensive data type does not drown out the less demanding data bandwidth types.

I can't speak for other countries' regulations, but this is not how the Open Internet Order was written when it was in effect.  (Note that this was repealed under Trump's FCC, and is currently waiting for a fifth commissioner before anything happens.  The mostly likely outcome is that the OIO will go back in place.)  It was not only source-neutral but also application-neutral, so ISPs weren't allowed to distinguish between consumer web traffic, file transfers, etc.  However, it only applied to "BIAS" networks (Broadband Internet Access Services), which were defined as mass-market consumer-facing networks, and explicitly didn't cover, "enterprise services, virtual private network services, hosting, or data storage services."  (Note that a data storage service is not a file transfer in this case.)

So if Starlink wanted to offer a premium low-latency QoS package, they'd have to go through some contortions.  ISTM that the easiest thing they could do would be to toss them into a VPN, but there are also exceptions for some "specialized services", whose definition doesn't quite fit into what the FCC considers specialized services.  (The poster child for these would be a black phone VoIP service provided by a cable MSO.)

[TheRadicalModerate]

Most of these telecom folks have (had, actually) cushy monopolies, either by gov't interference, or by lack of a LEO comsat constellation.   They are about to be given a really hard punch as SL comes online for a decent chunk of their customers.    I believe that a large segment of semi-rural customers will switch if SpaceX allows them to sign up.   

The only thing cable and GEO sat companies can do is to lower their prices and improve their service (or try political games to defend their monopoly).

This was partially true when most telecom was land lines, but it's not in the case of urban or suburban mobile networks, where competition is widespread.  It's still true in the case of a lot of rural mobile nets, but people in big cities hate their providers just as much as anybody else, and none of the providers has managed to dominate the market with improved customer service or price reductions.

I agree that ISPs have very little competitive power against one another.  But that will also be true of Starlink if Kuiper or One Web or one of the Chinese offerings goes on line--unless they can think of some way to generate network effects.  Tying something big (much bigger than financial) to low-latency service could do that.

[Robotbeat]

Most of these telecom folks have (had, actually) cushy monopolies, either by gov't interference, or by lack of a LEO comsat constellation.   They are about to be given a really hard punch as SL comes online for a decent chunk of their customers.    I believe that a large segment of semi-rural customers will switch if SpaceX allows them to sign up.   

The only thing cable and GEO sat companies can do is to lower their prices and improve their service (or try political games to defend their monopoly).

This was partially true when most telecom was land lines, but it's not in the case of urban or suburban mobile networks, where competition is widespread.  It's still true in the case of a lot of rural mobile nets, but people in big cities hate their providers just as much as anybody else, and none of the providers has managed to dominate the market with improved customer service or price reductions.

I agree that ISPs have very little competitive power against one another.  But that will also be true of Starlink if Kuiper or One Web or one of the Chinese offerings goes on line--unless they can think of some way to generate network effects.  Tying something big (much bigger than financial) to low-latency service could do that.

Do you live in the US? There's tons of regional monopoly or duopoly situations, often better than the alternative of no broadband in rural areas. There actually ISN'T that much competition in much of the US.

[su27k]

SpaceX takes aim at satellite mobility operators with Starlink expansion

SpaceX is seeking regulatory permission to connect moving vehicles to its rapidly expanding Starlink constellation, branching the broadband network out of fixed homes and offices.

The possibility of diversifying into more markets will raise eyebrows in the satellite mobility industry, as Elon Musk’s launch company continues to raise sizable funding rounds to back its capital-intensive plans.

Satellite operators Inmarsat, SES and Intelsat are among those likely to feel the most disruption in the land-mobile market, Northern Sky Research principal analyst Brad Grady said.

Depending on the exact form factors and use cases, Grady said mobile-satellites service (MSS) businesses at Inmarsat and Iridium Communications might see disruption akin to when maritime products transitioned to very-small-aperture terminals (VSATs).

However, antenna maker Kymeta and other players in the communications on the pause (COTP) market segment are likely to feel more of an impact from early Starlink adoption.

[FutureSpaceTourist]

https://twitter.com/pbdes/status/1369184193826222080

Not connecting Tesla cars to Starlink, as our terminal is much too big. This is for aircraft, ships, large trucks & RVs.