3 ms·

$80 million for 20,000 households is about $4,000 per household. Let's call it $10,000 per household like the other guy, assuming most people won't be intereste

$80 million for 20,000 households is about $4,000 per household. Let's call it $10,000 per household like the other guy, assuming most people won't be interested. Let's estimate that about $60 million of the cost is the actual build out of fiber, and $20 million is other expenses. If we're going to upgrade our backend equipment every 3 years over a 20 year life span on the fiber, we're talking about another $100 million on top of our $80 million investment. So we could have constantly upgrading backroom fiber laid down on a 20 year infrastructure cycle for $180 million, or about $22,500 per household (over 20 years). So the cost per household to lay down fiber, even if only 40% of the households are paying, is about $100/mo per household for 20 years. Given that we need to pay bond interest rates and cut some money in for the company managing your ISP account on the fiber, we're talking more like $130/mo, maybe $150/mo. So in short, if $130/mo-$150/mo for 1gig internet with open ISP competition instead of getting dicked by Comcast all the time sounds good, talk to your local government.

rayiner 11y ago
40% is an incredibly high uptake ratio. Google is not hitting that in Kansas City with gigabit at $70. And the higher the price, the lower the uptake. At $130 to $150 per month, uptake will be a lot less than 40%.
- ObviousScience 11y ago
  The water and electric grids in my city have near 100% uptake, especially in the denser parts of the city. While in most cases I would agree with the uptake analysis (eg, if we were talking about coffee shops), I'm not sure that it applies to the particular case of talking about rolling out a government monopoly in place of a natural monopoly using city bonds to cover the upfront cost (and time to smother people not using the city network). Further, I was actually just interested in plausibility: I didn't account for growth rates, the attractiveness of a city wide network, competition between ISPs, etc. That the answer was ~100 and not ~10 or ~1,000 is interesting to me, because it means that even if the technology isn't there, that the details work out a lot closer than I was expecting -- expensive, but not impossible. tl;dr: Deciding if we should roll out government infrastructure isn't quite the same analysis as seeing if a business works, and I only did the latter as a proxy for plausibility.
- ensignavenger 11y ago
  Do you have any data on this? Does that include their $300 installation for 10 years of service?
fiberrun 11y ago
Replacing all equipment every three years is way too aggressive. No network operator spends $1000 per subscriber for new equipment every three years. A more realistic and conservative replacement schedule would be seven years.
sre_ops 11y ago
Oh, so transit is free?

rayiner 11y ago

40% is an incredibly high uptake ratio. Google is not hitting that in Kansas City with gigabit at $70. And the higher the price, the lower the uptake. At $130 to $150 per month, uptake will be a lot less than 40%.

ObviousScience 11y ago
The water and electric grids in my city have near 100% uptake, especially in the denser parts of the city. While in most cases I would agree with the uptake analysis (eg, if we were talking about coffee shops), I'm not sure that it applies to the particular case of talking about rolling out a government monopoly in place of a natural monopoly using city bonds to cover the upfront cost (and time to smother people not using the city network). Further, I was actually just interested in plausibility: I didn't account for growth rates, the attractiveness of a city wide network, competition between ISPs, etc. That the answer was ~100 and not ~10 or ~1,000 is interesting to me, because it means that even if the technology isn't there, that the details work out a lot closer than I was expecting -- expensive, but not impossible. tl;dr: Deciding if we should roll out government infrastructure isn't quite the same analysis as seeing if a business works, and I only did the latter as a proxy for plausibility.
ensignavenger 11y ago
Do you have any data on this? Does that include their $300 installation for 10 years of service?

fiberrun 11y ago

Replacing all equipment every three years is way too aggressive. No network operator spends $1000 per subscriber for new equipment every three years. A more realistic and conservative replacement schedule would be seven years.

sre_ops 11y ago

Oh, so transit is free?