SpaceX’s Starlink satellite broadband service has the potential to be a game changer for rural broadband, according to an analysis by PCMag of Starlink speeds. The analysis is based on beta tester data exclusively provided to it by Ookla Speedtest.
The site looked at data from rural, suburban and urban areas. Among its more than 10,000 users in its semi-public beta were “a perplexing” number in urban and suburban areas where a variety of high-speed options already are available. The story cites Chicago, Seattle and Minneapolis as places where there were testers, despite readily available alternatives.
The site compared download speeds against other fixed service providers in 30 counties with at least 30 samples in any month from December 30 to February 24. The counties in which the fixed providers had the biggest speed advantage over Spacelink were urban or suburban: Los Angeles and Santa Clara counties, CA; Cook County, IL; King County, WA and Washington County, MN.
It is in rural areas that Starlink shines, according to the research. The five counties in which Starlink had the biggest download speed advantage over the fixed group were rural: Vilas County, WI; Ravali County, MT; Waldo County, ME; Okanogan County, WA and Lamoile County, VT.
The number of counties in which Starlink beat the fixed providers and those in which the fixed providers beat Starlink appeared to be about equal, as was the speed differential.
“Our own analysis shows that Starlink will make the biggest difference in rural, low-density, low-population counties with few options other than lower-quality satellite services,” wrote Sascha Segan, author of the PCMag article about Startlink rural speeds.
There is some skepticism about Starlink and its ability to serve rural broadband at scale, especially considering it has committed to serve 642K locations through the FCC RDOF program. Detractors have argued the service will struggle to provide adequate broadband speeds to that many rural customers.
At this point, Starlink is geographically constrained. The story says that reports put its current constellation most effectively covering areas ” between either 44 degrees or 45 degrees north, and either 52 degrees or 53 degrees north.” This region is in the northern third of the country and extends into Canada. A distribution map shows most beta testers in the northwest, with some in the upper Midwest and a smattering in the northeast and central and southern California.
Beta users report download speeds of as much as 170 Mbps with no data caps.
Starlink may be getting a speed boost. Last week, Space X CEO Elon Musk tweeted that he expects download speeds to hit 300 Mbps later this year. He added that latency will be 20 milliseconds.
In beta test mode, Starlink currently supports about 10,000 customers around the globe. Once commercial service launches, it will be interesting to witness how customer satisfaction levels perform over time.
The experience of previous satellite platforms (HNS, WildBlue –now ViaSat) was similar during the early stages of their satellite deployments. Early adopters were delighted with throughput speeds of a network operating at 10% capacity. However, as customers continued to sign up and load the satellites, network contention increased dramatically. Customer experience deteriorated and churn accelerated.
This is unfortunately a function of satellite network architecture, whether it is a GEO, MEO or LEO platform. Although improved latency of LEO vs GEO will help, it will not compensate for oversubscription and bandwidth contention. In today’s world, most services (streaming, etc.) are not latency sensitive in any event.
Deteriorating speed will be exacerbated by greater numbers of people moving to rural areas and attempting to work remotely. And as LTE & 5G expand beyond urban corridors, churn will rise. The only way to maintain speeds would be to continue adding more satellites (network nodes). That will be uneconomical. The high upfront and recurring monthly cost of service represents an additional challenge, particularly in environments will limited disposable income.
StarLink’s current marketing tagline “Better than nothing!” captures the marketing challenge pretty well. I would expand it by saying “…but more expensive!”
Let’s see what happens in a year or two after commercial service becomes widely available.
Thoughtful comments; thank you. I am in a very rural State House District (54 towns and settlements) in a very rural state (Maine) and the eastern interior of Maine is desperate for internet. Some still have dial-up modems here, though little-used, while most simply have awful service from Consolidated Communication or a hot spot. Several folks have purchased the StarLink system and are very pleased. However — and it’s a big “however” — the pricing is approx $600 for start-up equipment and $110/month for data. These are prohibitive numbers for at least 2/3 of our population. I hope the new infrastructure bill allows the internet to become our 21st century equivalent of the 1936 Rural Electrification Act — the internet is now as necessary as electricity, and it needs to be as affordable to ALL citizens.
Interesting comment, thanks. Hopefully, the Starlink term went to school on previous experiences that you allude to. I agree with your very last line–let’s see what happens…
While I agree with you David on some of these issues, the price for this service will also increase. Plus again we read reports of issues in bad weather again and again. So the question is, can you allow this to happen and still get government funding?
The latency is not an issue as far as I’m concern. We run Internet via GEO on C Band and 500ms is ok with me. We run in bad weather on a 1.4m antenna.