There were good reasons to be skeptical of OneWeb, but "consumer satellite communication companies have gone bankrupt before" is not one of them.
What lots of people expect is kind of irrelevant... OneWeb going bankrupt suggests that profit margins will be low at best, negative at worst. Cheaper launch cost is not enough to change the business from to non-viable to hugely profitable.
Who will be the customers of Starlink? Internet is a commodity for the average person, you can't make money in that market. This leaves special cases such as abandoned areas, airlines, ships, but that market is much smaller and it's unclear whether it can feed the massive capital costs of Starlink: launches, satellites, ground stations.
Edit: Providing service to ships in the ocean could be a problem, because of the lack of ground stations.
Also I don’t think you truly grasp the cost efficiencies spacex has and will achieve. Elon originally said that they could lower launch costs under $1m if the only cost was propellant and minor refurbs, compared to hundreds of millions. Even today their confirmed cost is 6x less than average [1]. This is not a 20% reduction or something. It is incredibly significant. one web was launching on other companies’ rockets, so that’s an additional disadvantage since spacex is only incurring the cost of the launch and no margin since there will be recovered from the internet service.
[1]: https://www.google.com/amp/s/theconversation.com/amp/how-spa...
Exactly my situation. I have symmetric gigabit and due to a nature of my business requiring symmetric fat pipe I have to live in a city. The minute I could get that gigabit in a wilderness I am out.
It definitely was an impactful change for us at first. I had to change some work workflows to do heavy network stuff on a remote machine. It took some time to refine and make (nearly) as efficient as doing it locally. I also had to sandbox all the streaming gadgets in the house and limit their bandwidth allowance.
But, the key for me was to move to an area with a reputable, locally owned WISP. It took a couple years but I found it. My WISP is run by an ex-IBMer down the road from me. The network is rock solid and goes down less than Comcast. I’m 10ms away from the nearest Google PoP. Everything on the network is solar-first.
These WISPs are out there and they make the decision to move to the country at lot easier.
Keep an eye out! I’m out in Colorado- Email me if you want to know where to look around here for good internet in a rural area.
This assumes that razor thin margins are bad, which is usually a safe assumption but in the case of Starlink, I say it's not open minded enough.
To boot Starlink online requires a minimum cost for service area (sats), but once you've got that area booted, onboarding new customers is trivial (vs laying fiber, cable, digging holes).
The user terminal is for each individual - "Looks like a thin, flat, round UFO on a stick."[0]
If you were offered $40 per Gb/s would you switch, 30? 20?
Personally I would switch because it's simply not f'ing Comcast.
[0] https://www.businessinsider.sg/spacex-starlink-user-terminal...
You would switch because you've imagined this service is magically cheaper and better, but now I say it'll be expensive and worse are you still enthusiastic?
Providing network service is a utility, and so yes razor thin margins are typical and that's too bad. It does make very high capital investment (such as launching huge numbers of satellites every few years) difficult to stomach though.
These projects (whatever Musk says about them when hyping his company) deliver expensive network access to places where it's otherwise not practical to offer service at all. They aren't going head-to-head with Comcast, or even with AOL, they're competing with "Physically take a day's / week's work to the A site where they have network access".
Yeah, there are a lot of places in the US--some of which aren't even really all that rural--that just don't have good Internet options. Today, satellite basically falls into the slightly better than nothing category but it's neither price-competitive nor good if you have essentially any other option. While I know people in tech who make do with only a satellite connection, I'm sure many reading this wouldn't consider that viable and it will be increasingly less so over time.
5G will presumably help in some cases as it rolls out but it certainly won't reach everywhere. And, presumably, over time places that don't have a certain level of Internet access will be increasingly unsuitable at least for long-term residence.
I read a nice overview of the constellations a while back, but I realize now that it's actually pretty outdated (2018, and it talks about optical crosslinks between SpaceX's satellites, which has been punted for now): http://www.mit.edu/~portillo/files/Comparison-LEO-IAC-2018-s...
So I don't know if my comment about throughput is accurate anymore.
It will soon become totally overloaded. A city like NYC probably has home/office bandwidth use of 40gig/sec per a few city blocks, nevermind the whole metro area.
It's not going to work as a home internet replacement if you already have fibre or cable service. Furthermore, fibre and cable can be upgraded easily by putting new equipment on each side (even cable internet has went from 40mbit/sec to 10gig/sec - fibre has even more capacity).
Elon himself has admitted that it won't work in cities for this reason.
My point wasn't that it doesn't make sense whatsoever, it's just that it is going to be a niche solution. Considering the insane capex of getting the sats in space, if the market is too niche, will it be able to pay back the capex costs?
You can't know until you do extensive market research.
>a lot of rural areas have semi-workable 4G coverage
Yes, but also extremely restrictive data caps and expensive overage fees per kilobyte.
>small wireless p2p ISPs
Yes, that would be the biggest competition to LEO satellite broadband. However, PTMP (point to multipoint) WISP requires line of sight and expensive CPEs to reach speeds over 20 Mbps for more than 1-2 km (0.6-1.2 miles). There is still shared capacity, and RF noise problems (as with satellites).
>it is going to be a niche solution
Unfortunately, yes. And if it ever gets anywhere near to being profitable and threatening the traditional fixed access telcos, they'll just spend some pennies (compared to revenue), lay fiber to most of the residences, and do a mix of FTTH and FTTC (VDSL2 with profile 35b for up to 50 Mbps at 1km (0.6 miles) or 350 Mbps at 100 m (0.06 miles)) to save costs for users who have good-enough-quality twisted pair telephone lines but are too far from the DSLAM to get good speeds.
VDSL2 is quite servicable, it is unclear to me why is it so underutilized in the USA. DSL is a dedicated service, since the telephone line isn't shared with multiple people nor multiplexed in any way, so it provides a very stable connection, if the speed is conservatively set per the line's maximum length.
A satellite covers a much wider area. We'll see how it plays out, but as a person living out here, I'm excited.
He’s about 7 years in and now serves 70% of our area. It ain’t cheap, but he’s been profitable enough now to be able to buy additional bandwidth from his upstream. We’re about to get 50/25 in the coming months.
I offer my services to him for free just to help contribute and make sure he doesn’t go out of business. It feels like a community effort! I set up the online billing portal and recurring payments, another guy goes on people’s roof to install the ubiquiti receiver, etc. we all chip in.
Initially I had setup a PTP link at 3 km with MikroTik gear and got 10-20 Mbps, varying wildly. I actually subscribed to an Internet connection over DSL in a building where I don't live and got permission to put an antenna and modem there to send Internet to my house. Later I subscribed to an actual Internet company, doubting my own skills in setting up the link. They don't offer anything over 10 Mbps, but I asked for them to let me have an unlimited link for an hour, and it did the same 10-20 Mbps with Ubiquiti gear.
Some WISPs use Cambium gear and apparently do offer 50 Mbps, but I'm not quite sure how well does that work, and how much better is Cambium, or i.e. Mimosa, compared to UBNT/MTik.
I think the clear line of sight and little radio interference were key to making this work. We don’t have cell or tv signal here. From what I understand the sheriff’s radio coverage is spotty as well.
Unfortunately, shocking as it is, most WISPs in developing nations are stuck on 802.11n equipment, especially using MikroTik's CPE and AP hardware.
If I'm ever redoing this link (I probably will, I've torn it down because I lost permission) I'll get Ubiquiti AC Gen 2 PTP gear, the best I can reasonably afford. And some aiming tools.
For reference, people are talking about wanting 25 to 50mbps, while I am getting gigabit for $45/mo (Cincinnati, Ohio). Hence the cat and mouse game.
It suggests that investors didn't think that OneWeb would have sufficient profits. The example I gave suggest that investors think that Starlink and the new Iridium satellites will be profitable.
The market for Starlink are the millions of people in rural and sparce suburban areas for whom internet connections are slow and expensive.
Your comment here is pure speculation.
From everything we know, the problem for them was huge capital cost and massive required cashflow to buy launches. All that when your competition has the exact opposite situation totally vertically integrated from the rocket launch to the manufacturing of the whole satellite.
Starlink has so far always gotten tons of money when they asked for it, suggesting the exact opposite of what you stated. OneWeb was just unlikely to get there.
It's too early to speculate about profit margins before the company has finished launching its constellation and started providing a commercial service.
A single contract from the US government that displaces their contracts with various providers could be profitable to spacex.
Interesting times.