Elon Musk has high hopes for SpaceX’s Starlink satellite internet service, hyping its billion-dollar revenue potential and its ability to provide high-speed internet to all corners of the globe. But SpaceX has launched just 2,000 satellites out of the tens of thousands on deck for deployment, and people are paying high costs to tap Starlink internet today — $599 just for installation, and an additional $110 per month for the basic plan.
But satellite internet nerds, Elon Musk stans and people who live in rural areas with low broadband coverage have signed up in droves for the chance to test out Starlink since it first became open from the beta program to the public in February 2021.
A year later, Musk tweeted that Starlink hit 250,000 customers. Though it's unclear what the company’s current waitlist looks like, last May, the company had 10,000 active customers and a waitlist of 500,000. In October, SpaceX pushed its wait times for Starlink to “late 2022, early 2023” after originally slating orders to ship between late 2021 to mid-2022. In a Reddit thread where Starlink customers share when they first ordered service and when they actually received it, posters are reporting wait times as long as 14 months: Some placed orders in February 2021 and just received their equipment this month. One user even reported having their preorder canceled after missing an email with their seven-day window to claim service.
“I'm no longer an excited beta user on an exclusive wait list. I'm now a sour, bitter, hater that may never forgive the company for their treatment,” the user wrote in a post. “I understand there may be more demand than supply, but to COMPLETELY cancel my order and send me to the back of the line? Wow.”
Irate early adopters aside, the important question remains: How does Starlink stack up to traditional broadband internet? In the U.S., it’s still lagging behind, according to research firm Ookla.
An Ookla report from March found that SpaceX’s internet service offers median download speeds of 104.97 megabytes per second, and upload speeds of 12.04 Mbps, lagging behind broadband providers’ median download speeds of 131.3 Mbps and upload speeds of 19.5 Mbps.
Starlink does, however, easily outperform rival satellite internet providers. HughesNet only gets download speeds of 20.92 Mbps and upload speeds of 2.54 Mbps, according to Ookla, while Viasat offers downloads at 21.81 Mbps and uploads at 2.88 Mbps.
However, Starlink’s service depends on the country. Starlink beats out broadband with massive margins in several countries, including Australia, Mexico, France, Italy, Ireland, the U.K. and Germany. Starlink also has slightly faster download speeds in Canada, with speeds of 106.6 Mbps compared to fixed broadband download speeds of 96.4 Mbps.
Though it beats out broadband in some countries, the price tag is much higher than what your average broadband ISP charges. A Starlink terminal, which includes a satellite dish and a router (which users have to install themselves), costs $599 to set up, then $110 per month for the service — and that’s just for the basic plan. The premium plan is $2,500 for setup and $500 per month for the service, promising speeds up to 500 Mbps.
Compare that to Verizon, which claims to offer download speeds of 300 Mbps for around $40 a month with a setup fee of $99 for its basic plan, or Spectrum, which offers a basic plan of 200 Mbps for around $50 per month (setup is an extra $50 for professional installation and $20 for self-installation). Of course, real-world download speeds are often far slower than the ones ISPs promise, but for the average internet user, particularly those that live in suburbs or cities, Starlink simply may not be worth it.
Starlink is designed primarily for remote and rural areas, as its satellite dishes work best with an unobstructed view of the sky. It promises download speeds between 100 and 200 Mbps, though it only reached a median download speed of more than 100 Mbps in the fourth quarter of 2021.
Starlink itself claims to deliver “high-speed, broadband internet, even to places where access has been unreliable, too expensive, or completely unavailable.”
But the service has a ways to go. For one, SpaceX needs to shoot more satellites into the sky. The company is currently working on building out Starlink’s satellite constellation, having launched more than 2,000 satellites into orbit over dozens of launches since it started at the end of 2019. SpaceX is aiming to deploy a total of 42,000 satellites, making its megaconstellation even more mega. Though speeds will likely improve once Starlink builds out its constellation, the process is costing SpaceX a pretty penny. The Starlink constellation could cost the company a total of between $20 billion and $30 billion to finish, Musk said in June.
For Starlink and other satellite internet operators to compete with fixed broadband, the cost of everything — from launching a satellite into orbit to the materials needed to make them — is going to need to go down, said Brooke Stokes, associate partner at research firm McKinsey & Company.
“Then I think you've got the question of how patient will investor capital be relative to how fast can these megaconstellation operators make markets and unlock demand,” Stokes said. This question puts Starlink, which operates the biggest satellite constellation in the sky, in a shaky position.
That doesn’t seem to deter Musk, however, whose steep investment is fueling big ambitions. Starlink could have colossal returns, bringing in as much as $30 billion annually by 2025, Musk predicted last year, making around 10 times the revenue that its launch services business would bring in. But for Starlink to actually bring in that much revenue, it would need to become more than just an internet service provider. It would need to be the internet service provider, serving tens of millions of subscribers annually. (Years ago, SpaceX predicted that Starlink would attract more than 40 million global customers by 2025. For context, Spectrum has 32 million internet customers in the U.S., and Verizon has around 6.74 million.)
Needless to say, Musk’s ambitions are, well, ambitious (though when are they not?). Blair Levin, a nonresident senior fellow at the Brookings Institution and policy analyst at New Street Research, said the total addressable market of people who live in rural and remote areas and would actually be willing to pay the amount of money per month that Musk asks for Starlink is simply not high enough. And the infrastructure bill passed in November dedicated tens of billions of dollars toward grants for states to build out rural broadband, Levin said, a factor which also might take a bite out of Musk’s market in the U.S.
“That's a significant investment into the higher-functioning network that will have its [capital expenditures] paid for, and for whom the operating expenses are going to be lower,” Levin said. “So that's going to hurt the potential of [Musk’s] business.”
One possible saving grace in the bill is a $30 per month subsidy, totalling in $14 billion, allocated to rural families through the Affordable Connectivity Program to help cover the cost of internet, Levin said, bringing down the monthly cost of Starlink from $110 to $80 per person.
But Starlink may have more luck focusing on enterprise contracts, said Stokes, such as major contracts with cruise ships and airlines. Though the consumer internet market is the largest by the number of potential markets, it’s very fragmented, Stokes said, with many users not able or willing to pay the high cost for Starlink.
Starlink may have realized the importance of enterprise contracts, too: SpaceX inked its first in-flight Wi-Fi deal with JSX earlier this month to cover up to 100 planes, and is currently in talks with Delta Airlines to bring its services on board (though this partnership would likely take a few years to be realized, given that Delta would have to outfit its planes with Starlink equipment).
The verdict? From speeds to filling orders to expanding its customer base, Starlink has a long way to go. But Musk seems committed to the cause — at least unless he gets too busy taking over Twitter.