Only speaking with knowledge of their first market (Kansas City), I would guess it is a combination of three things:
* Buildout costs
Google was using the same contracts as everyone else for the fiber infrastructure install. The guys running the trenchers and directional drills where the same guys that do it for AT&T, Time Warner, Comcast, whomever. Google built at such a pace they certainly were not getting any deals, and very likely significantly increased the costs due to the massive amount of work that was being generated vs. available contractors. Capex cost is everything and they were just blindly writing checks to other people to get it done.
Google was innovative in their "fiberhood" signup process where they heavily marketed to their targeted expansion areas to encourage a certain take rate before build out. I suspect this saved them some money (or rather allowed their CapEx to be spread over more subs), but after that initial signup, I'm not sure there were any efforts to promote further subs. (Actually I'm not even sure people could subscribe after the initial buildout).
There was no Google magic about the fiber installs. It is stringing fiber along poles or burying conduit and running fiber through it. If there was some breakthrough that was able to be made here or technology which gave them an advantage on the CapEx required, it would have been huge boon. There wasn't.
* Access or cost of access to AT&T/KCPL utility poles
Unsurprisingly those that owned some of the utility poles here (AT&T and KCPL) were more than happy to smile while screwing Google's fiber effort. The rates and access rules for these things are set at the Federal level and lobbying by telcos/utilities is fierce. Google's blank regulatory check from the cities was worthless on this front (except for parts of KCK where the local government owned some poles). The cost and regulatory burden for utility pole access is drafted to simply make use impossible or uneconomic.
* A gradual mindset transition to a "Wireless is easier than all this ditch-digging" mindset.
Google has been doing a lot of work with licensed spectrum. Maybe they ran the numbers and figure it is cheaper to buy some spectrum and see if they can work some wireless magic (either technology or special FCC regulation perks) and make it their special sauce. I don't think I'm betting on that particular horse.
It seems that a lot big players are betting fixed wireless for the urban market(instead of the smaller rural market) . It's probably the main purpose of 5G. And there's some nice technical innovation(cohere promises to solve fading, tarana talks about working in urban non-line-of-sight) . And even LTE-A offers nice capacities in that context.
Very interesting regarding CapEx and Google blindly writing checks. So it would appear that in addition to the regulatory hurdles Google also made some poor decisions in the rollout and signup process. Would Google have been better served targeting smaller cities that might have been more receptive to Fiber than the ones they ended up building in? Or is there not enough money to recoup costs in smaller cities?
Regarding utility pole access being handled at the Federal level, is there any lobbying campaign at the national level to rewrite the rules governing access to make the situation less skewed towards incumbent telcos? In my area here in LA county, it seems most utility poles are on public right-of-ways, I would think it would make a lot of sense to have those poles available for any company that wants to compete to offer services. Is there any meaningful reason this isn't done aside from cronyism on the part of the telcos?
I'm suspect of the CapEx arguments. The whole reason for the "fiber rally" procedure, and for offering free 5 MB/s Internet, was to get every house on the block to sign up. The marginal cost of a contractor going next door and wiring their house up is significantly less once they're already out there than if the contractor has to make a trip out to each home; it wouldn't surprise me if Google's CapEx costs were almost an order of magnitude less than Verizon or ComCast. (How much time does the cable guy actually spend installing cable, vs. driving to the place, making sure he got the right house, and context-switching between tasks?)
> In my area here in LA county, it seems most utility poles are on public right-of-ways, I would think it would make a lot of sense to have those poles available for any company that wants to compete to offer services.
> A utility shall provide a cable television system or any telecommunications carrier with nondiscriminatory access to any pole, duct, conduit, or right-of-way owned or controlled by it.
There was some dispute as to whether Google Fiber counted as a "telecommunications carrier" for purposes of Section 224, but that is probably mooted now that the FCC has determined broadband providers are all "telecommunications carriers" subject to Title II.
* Buildout costs
Google was using the same contracts as everyone else for the fiber infrastructure install. The guys running the trenchers and directional drills where the same guys that do it for AT&T, Time Warner, Comcast, whomever. Google built at such a pace they certainly were not getting any deals, and very likely significantly increased the costs due to the massive amount of work that was being generated vs. available contractors. Capex cost is everything and they were just blindly writing checks to other people to get it done.
Google was innovative in their "fiberhood" signup process where they heavily marketed to their targeted expansion areas to encourage a certain take rate before build out. I suspect this saved them some money (or rather allowed their CapEx to be spread over more subs), but after that initial signup, I'm not sure there were any efforts to promote further subs. (Actually I'm not even sure people could subscribe after the initial buildout).
There was no Google magic about the fiber installs. It is stringing fiber along poles or burying conduit and running fiber through it. If there was some breakthrough that was able to be made here or technology which gave them an advantage on the CapEx required, it would have been huge boon. There wasn't.
* Access or cost of access to AT&T/KCPL utility poles
Unsurprisingly those that owned some of the utility poles here (AT&T and KCPL) were more than happy to smile while screwing Google's fiber effort. The rates and access rules for these things are set at the Federal level and lobbying by telcos/utilities is fierce. Google's blank regulatory check from the cities was worthless on this front (except for parts of KCK where the local government owned some poles). The cost and regulatory burden for utility pole access is drafted to simply make use impossible or uneconomic.
* A gradual mindset transition to a "Wireless is easier than all this ditch-digging" mindset.
Google has been doing a lot of work with licensed spectrum. Maybe they ran the numbers and figure it is cheaper to buy some spectrum and see if they can work some wireless magic (either technology or special FCC regulation perks) and make it their special sauce. I don't think I'm betting on that particular horse.