Starlink's threat to wired broadband 'minimal' – analyst
US cable operators and other wired broadband service providers likely won't have to fret too much about SpaceX's Starlink in the coming years. But the emerging satellite broadband service is poised to play a significant role in providing service in rural US areas alongside opportunities in the much larger global market, according to one analyst study.
"We have no trouble imagining Starlink as a potential game-changer for helping to bridge the digital divide in the US," Craig Moffett, analyst with MoffettNathanson, explained in a new report that dissects the potential for the satellite broadband service. "The opportunity outside the US, where broadband availability is generally lower, is even larger. The product appears to deliver what it promises. Globally, the rural opportunity is a very, very big one."
For now, it appears that most US cable operators and telcos need not worry about Starlink encroaching too heavily onto their wired broadband turf. Moffett says that's largely because capacity constraints will likely require Starlink to focus its resources on the best use for that capacity: in relatively uncompetitive, rural markets.
"For investors worried about Starlink's threat to terrestrial broadband, we think the threat is minimal," Moffett surmised. "Based on our analysis of the data available today … Starlink is much better suited for bringing broadband to unserved or underserved markets than it is for bringing competition to already-wired markets."
That assessment syncs up Elon Musk's own view of the service. Speaking at an event last March, he suggested that Starlink is "not some huge threat to telcos."
Sizing up Starlink's US and global opportunities
Though Starlink's global potential could be massive, Moffett's estimation on Starlink's total addressable market (TAM) for the US is expected to be relatively small and focused.
According to his analysis of available capacity and anticipated usage, Moffett estimates that Starlink's US TAM, at a full deployment of about 12,000 low-earth orbit (LEO) satellites (about 1,400 are in orbit today), is in the range of just 300,000 to 800,000 households, or less than 1% of the US market, Moffett notes.
Moffett uses mix of data to come to that range, including current Starlink's per satellite capacity estimate of between 17 Gbit/s to 23 Gbit/s along with actual usable capacity hovering at about 10% of the system capacity of a typical LEO satellite. Moffett also pairs that with recent broadband usage data, including recent info from OpenVault showing that the average North American broadband user consumed between 1 gigabyte to 1.2 GB per hour (our roughly 2.2 Mbit/s to 2.7 Mbit/s) during peak usage hours.
But Starlink's US reach could expand further when some aggressive assumptions on both the number of satellites in orbit and the amount of capacity per satellite are applied. Starlink's US TAM could expand to about 6 million customers (based on current broadband usage levels) through an expansion to 42,000 LEO satellites (based on SpaceX's request in 2019 for an additional 30,000 satellites), and the deployment of new, more capable satellites that could be capable of delivering up to 60 Gbit/s each, Moffett explained.
Moffett also makes some subscriber support calculations based on Starlink having 7 Tbit/s of capacity at full deployment. He suggests that system could serve at least 70,000 subscribers at full advertised speeds in the neighborhood of 100 Mbit/s. "Assuming an oversubscription rate of 10:1 (a common, if imprecise, rule of thumb for network capacity planning), then perhaps it can serve 700 thousand potential subscribers," he adds.
Moffett acknowledges that it's still difficult to apply a precise ratio, as the ranges could vary across markets and speed tiers. However, a focus on low population densities in rural markets means "the Starlink network will likely never have oversubscription rates anywhere close to those of a dense urban center network that fully benefits from efficient multiplexing," he added.
Even though the Starlink TAM looks relatively small on a US market basis, Moffett believes Starlink's potential to reach unserved or underserved parts of the globe – a measure of people in the billions – is too big to be called a "niche" opportunity.
"Extending our capacity analysis to Canada, Europe, Australia, and Japan, which we believe will be the core markets for Starlink, effectively quadruples the landmass and triples the addressable market," Moffett explained. "Europe will most definitely be a major market, and perhaps the biggest market, for Starlink."
The critical role of Starship
Moffett also believes that SpaceX's $10 billion cost to build Starlink is reasonable, assuming that launch costs drop 80% through the use of the emerging, fully reusable Starship launch vehicle, and the cost of the satellites themselves decline by nearly half.
Using Starship for Starlink satellite launches could be critical for SpaceX to meet the FCC's 2027 requirements. According to Moffett, the company would need to increase its launch cadence to 200 Starlink satellites per month, up from the 120 per month it's been sending up of late – via two Falcon 9 missions, each carrying about 60 satellites.
Though recent tests ending in crashes and explosions show that Starship has a ways to go, Moffett notes that it's believed that the next-gen launch vehicle will be capable of carrying 400 Starlink satellites at a time. If Starship becomes a proven and reliable launch vehicle, SpaceX will be in a position to vastly accelerate the cadence of Starlink satellite launches.
"Already at a record pace, SpaceX will need to accelerate further," Moffett explained.
Increasing role for inter-satellite laser links
Moffett's report also took a closer look at the importance of Starlink's proposed use of optical inter-satellite laser links (ISLs) to provide continuous coverage in large areas, including oceans, that can't support traditional ground stations, while also improving the capacity utilization of LEO satellites.
Calling it another critical piece of Starlink's puzzle, the use of ISLs would effectively turn the LEO constellation into a mesh network that allows individual satellites to share capacity from others. That activity remains in the early stages, as Starlink is using ISLs in just ten satellites to help provide coverage in the Earth's polar region.
"The importance of linking satellites together cannot be overstated," Moffett wrote. "Not only do interlinks enable sharing of capacity more efficiently by making use of otherwise wasted satellite capacity over regions without ground stations, they also enable the provision of service to areas where it is impossible to put a ground station (think the aeronautical or maritime markets, where signals can be 'hopped' in steps to a satellite over land and only then sent to a ground station."
But Moffett also points out that the use of ISLs is costly, as they require the larger solar arrays and batteries, as well as the laser terminals themselves. But Musk tweeted in January that ISLs will become a common piece of Starlink's repertoire in 2022.
All sats launched next year will have laser links. Only our polar sats have lasers this year & are v0.9.