When comparing costs of building (per home passed/connected), it is also important to see if those quoted costs include the regulatory costs of dealing with cities. If a municipal project won't suffer costs of negotiating for diggging/building permits, already has the land to build the "CO", and won't be delayed by stalled paperwork, this could represent a significant difference compared to an incumbent that constant hits a brick wall of bureaucracy which cost money and delays the project. (In Canada, a delay of a couple of months can cause a delay of 1 year due to winter). In the case of Google, with cities begging on their knees to get Google's project, I suspect they didn't have many problems getting the city's cooperation. Yet, there have been stories of delays due to bureaucracy. In the case of Verizon, I suspect those bureaucracy costs are much higher. The other aspect which you need to compare is existing infrastructure. If there are already conduits between CO and neighbourhood, and there is room inside, you can just blow your new fibre through them which costs a lot less than having to dig and install new conduits. (and "space available" is one of the issues that lead telcos to go GPON instead of wanting 1:1 strand to home ratios since that requires much more conduit capacity the closer you get to your point of aggregation. So when comparing both Verizon and Google projects, you need to factor exactly how much needed to be built for them, and how much will be needed for you. If you have 0 existing conduits and need to build 100% of your FTTH plant, while Verizon had x% of conduits already built, then your cost per home may be higher. BTW, out of curiosity, how many spare copper pairs were traditionally provisionsed on a cable run that passed 100 homes ? And in a fibre system, do you keep the same ratio of extra strands for each home passed ?