In some ways, Google Fiber’s experience with gigabit internet access resembles Verizon’s earlier experience with Fios. Both firms announced aggressive fiber-to-home builds, and both halted those builds before completion of the announced or intended footprints.
As Google Fiber maintains its construction and marketing of existing gigabit networks, while pausing extension, so did Verizon halt its Fios build before all of its major metro areas were wired.
The business model seems to have been the issue for both firms. Google Fiber has been disappointed with take rates and revenue, while Verizon seems to have discovered the business model did not work as well as expected, either.
Likewise, both firms now believe fixed wireless could substantially improve the business model.
The dynamics of competitive access markets are among the key reasons both firms now think differently about fiber to the home deployment. In the monopoly era, the supplier of access services (voice, for telcos; video entertainment for cable TV companies) could safely assume that, once a network was built, adoption would range from 80 percent to 95 percent.
Other Major Private Company**
Total Top Cable
Total Top Phone Companies
That meant the “cost per customer” and “cost per location” were nearly identical. In simple terms, “build past a home or business, get that location as a customer” was a good rule of thumb.
All that changes in competitive markets. In internet access markets, one firm--typically the cable operator--can expect to get as much as 60 percent take rates, while the telco can expect to get 40 percent take rates.
Video, in fact, is the more competitive market, as there are at least three suppliers in every market, limiting cable operators to about 50 percent adoption, while satellite gets 20 percent share and most telcos get about 20 percent. The salient exception is AT&T, which owns both DirecTV satellite service and also has sold U-verse video services. In some instances (in its largest fixed network areas) AT&T could have as much as 25 percent total video account share.
The obvious implication is that network “cost per potential customer” (cost per passing) and “cost per customer” diverge quite a lot. Cost per passing represents significant stranded investment, when take rates range from less than 20 percent to 50 percent.
Simply, at 20-percent take rates, cost per customer is five times the cost per passing. Even at robust 50-percent take rates, “cost per customer” is double that of “cost per passing.”
Even Verizon, able to sell three anchor services, with 40 percent adoption of internet access and perhaps about the same voice adoption (as a percentage of homes passed), has found the business case for Fios a difficult proposition. That is why its new optical fiber deployments are based on enterprise and mobile backhaul requirements first, to create a relatively dense optical footprint, and then to extend from optical endpoints to potential consumer neighborhoods.
It is doubtful Google Fiber has managed to consistently get 20 percent adoption in most of its markets.
That is why both Verizon and Google Fiber have paused their original fiber-to-home footprints, and why both are seriously moving towards fixed wireless access to provide gigabit internet access.
Other major private company*
Total Top Cable
Satellite TV Companies (DBS)
Total Top Phone
Total Top Pay-TV Providers