These are the obvious long-term customers. But I don’t think they are bankrolling the venture. High-income rural homes are a good beachhead since a partial constellation can deliver 100% uptime to some locations.
Median incomes in counties that are mostly rural skew lower [1] than those in urban counties [2]. But the difference, about 4%, is less than I had thought [3].
Moreover, “median incomes in rural areas of the Northeast and Midwest ($62,291 and $55,704 respectively) were higher than in the regions’ urban areas ($60,655 and $51,266)... Median household income was higher for rural households [versus urban ones] in 32 states. In Rhode Island, for example, median rural household income was $85,278, about 57 percent higher than the median for urban households ($54,324)” [4].
And since these are largely Census data, they ignore vacation homes.
[1] https://www.census.gov/library/visualizations/2018/comm/acs-...
[2] https://www.census.gov/library/visualizations/2018/comm/acs-...
[3] https://www.census.gov/newsroom/blogs/random-samplings/2016/...
[4] https://www.census.gov/library/stories/2017/11/income-povert...
That being said, your question sparked my interest, so here is some back of the napkin math.
Currently spaceX launches 60 satellites per launch. Each satellite is estimated to be roughly $250k (Exact cost not known), An estimated rocket cost of $50m (for used) So roughly $65mil per 60 sats, or 1.08mil per sat cost. Each sat has an estimated lifespan of 5 years
So 65mil / 60 / 5 = ~$217,000/y per sat, not including personal costs for support, ground base stations, etc.
Currently there are 900 sats in orbit, I believe they are currently approved for 12,000 sats, but want to go up to 30,000 sats.
So for the current constellation they need $195mil/y or just under 165,000 customers at the $99/m (165,000 * 99 * 12)
For the 12,000 cluster, they would need to make 2.6bil per year, or 2.2mil customers at the $99/m (2,200,000 * 99 * 12)
For the 30,000 cluster, they would need to make 6.5bil per year, or 5.5mil customers at the $99/m (5,500,000 * 99 * 12)
For comparison, comcast has roughly ~28mil customers for internet. Rural america is ~60mil people in total (census.gov) (not counting vacation or secondary houses from non rural americans) so starlink's break even on rural subscribers only would be just under 10% of rural America. (lets call the cush government contracts a wash for operational costs for people and hardware on the ground?) Note that ONLY 5.5mil Americans ( <2% of Americans) would fund all of the hardware cost and launch costs. Every other person in every other country around the world would theoretically be profit?
55M$ is the price to buy a lunch, Spacex internal price will be much lower, likely below 40m. Some stats are launch with rideshare, those can make cost better.
More importantly this is quite a temporary cost, soon they will start launching with Starship, witch should make the cost far, far lower. To get anywhere close to 30 or even 12k sats Starship will certainty be used.
Once they hit production rates making standing constellation of 30k sats possible the price of the sats should come down much further.
SpaceX really has put together a beautiful puzzle here. Literally nobody can replicate this anytime soon. Rural costumers are only one small branch, transportation, military, cloud computing, ISP backbone, backup net connection.
I suppose it also matters a great deal how geographically distributed those rural customers are.
Now there's expensive houses everywhere. A lot of the older, more humble houses are still there, but I get the impression they're gradually getting replaced and the whole area is gentrifying. I don't know why; we're far enough away from big cities to make commuting inconvenient, and there isn't any big industry, just farms and vineyards. Maybe those are more profitable than I thought. Or maybe people are just moving out of tech hubs and working remote.
Two things though.
Thing #1.
Definitions of 'poor' vary by countries, and my understanding of 'poor' in the US is:
* no safety net
* low, or no minimum wage often
* no/low quality health care
Meanwhile, in Canada (for example) with minimum wage, paid for health care, and a strong safety net, poor is a different thing.
I don't want to get into a big health care debate, but essentially taxes pay for health care here, and the less you make, the less tax you pay. And when you live in a rural area, you need significantly less pay to live to the same standard as a city dweller. Everything is more expensive in the city.
This means that a rural person pays less income tax, for the same standard of living. And therefore, essentially, less for health care.
When combined with housing costs which are immensely lower, the ability to reduce other costs (heating with wood, growing your own food, the ability and room to repair your own car, on and on...), there is more room for extras than one might expect.
An example; waiters/waitresses are paid, by law, minimum wage. And in rural areas, this is much more advantageous than in the city.
I guess what my point #1 is saying, is that rural poor are often better off than urban poor. And do have some discretionary spending, at least in some countries out there.
Point #2....
Definitions of rural.
There is a zone around smaller cities which is not 'suburbia', but is indeed rural. And these areas are often well within easy reach of cities. For example, the city I live close to has a tech businesses concentrated on its western side.
This means that when buying a rural home, one need to be concerned with reaching the entire city quickly. Only the western side. In as such, drive time from rural -> city, as the region is ~ 1.5M, is not that bad. Less than 30 minutes even.
In this rural zone, housing is far enough apart that on side/non-main roads, there is no high-speed internet access.
To speak to that, I live on the very last side-road, before high-speed internet access becomes non-existent. This is no mistake, when I looked to buy, I found the zone where high-speed ended, then searched in that area for housing.
To give you an idea? My house is worth ~ $200k. 10 minutes away? The same house? ~$400k, and I have a 1 acre lot, versus a 1/4 acre city lot.
Anyhow, the above is a bit of rambling, but the point is ; there is 'very close to the city rural, with no high speed' and 'farther out rural'.
Close-to rural enjoys city pay scales for higher end jobs, even without remote work.
And if starlink comes? People might move further out.
There are definitely applications for this in all those sectors, the onus should be on Starlink to confirm that they don't or assume they are – particularly in light of Elon's recent announcement about SpaceX weapon drops.
And AFAIK SpaceX's funding for Starlink is a combination of their own revenue from launch services, and private VC. So it's not correct to say that the military is "bankrolling" Starlink, except to the extent that they buy unrelated launch services from SpaceX.