Google Fiber: Not So Revolutionary

Yesterday was the big day that Google announced its Kansas City Google Fiber launch. I wasn’t available for the live broadcast, and I haven’t found the time yet to watch the replay (see below), but from what I’ve had the time to read the announcement is rather underwhelming. I will however reserve full judgment until I’ve had time to look into it more thoroughly.

There are three aspects where Google was expected to strongly disrupt things: services, costs and replicability.

On services, Google distinguishes itself by offering gigabit fiber at very competitive rates, and offering a “free” broadband service for customers willing to just pay the $300 connection fee. All in all, not much in there to get people extra-excited, especially considering that the attached TV offering is in no way disruptive compared to what a cable operator provides (and I suspect, although I haven’t investigated that, that it’s rather a sub-par TV offering in terms of the content it carries…) Sure, a gigabit for $70 is much lower than any other gigabit offering in the US and likely to get the geeks excited, but it’s still $70, ie. higher than most broadband solutions available to KC customers today. So the arbitration for these customers will remain. Plus the $300 connection fee can be a serious deterrant (although they’re waving it off for now, it wasn’t clear to me until when). As for the “free” low level broadband service, it does require paying the $300 connection fee. We’ve seen similar models for “free” service as a pathway to paying service in social housing in Europe, and it’s never worked (the transformation rates from free to paying customer always stay extremely low). All in all, I don’t think the service offering as it currently stands is anywhere disruptive enough to get Google Fiber beyond the core geeks and early adopters. They seem to have adopted an Iliad like approach in that they announce they will keep adding services over time. Let’s see if they actually do and if these services have more appeal.

On cost, Google also took a note from Iliad’s book and decided to design their own hardware. It’s not something new for Google since their search server farms were home designed virtually from day one. I’m not quite sure it’s that much of a cost saver in this arena (economies of scale do matter in the hardware business and they won’t be shipping that many ONTs and STBs), but I very much doubt Google will share numbers on that anyway. Other avenues to save costs are highlighted in Stacey Higginbottom’s article The Economics of Google Fiber and What it Means for US Broadband. But again, there’s little there that we haven’t seen elsewhere: demand aggregation is nothing new in FTTP, many European projects have successfully implemented it, and as for the QR code thing, it’s hard to see how that could represent significant cost savings. In fact, a number of hardware manufacturers in Europe like Genexis are offering plug & play FTTP hardware which is well ahead of what Google Fiber is planning to do. Also, I’d be interested in understanding the cost savings associated with Google’s deep backhaul network which, I suspect, allows them to offer connectivity at a much cheaper rate than anyone else could.

And that leads us to replicability. Google Fiber when it was originally announced was supposed to be not just a landmark project but a blueprint for municipals in the US who want high-speed broadband where established telcos and cablecos are unwilling to provide it. It seems that whole aspect of the project has now dissapeared. Google certainly isn’t sharing enough on what they’re doing to enable anyone to replicate it, and furthermore most of the savings that they highlight are Google specific and as such not replicable by anyone else. Finally, the decision to not open their network to competing ISPs puts them firmly in the camp of vertically integrated telcos/cablecos.

Based on these elements, my views at this stage is that Google Fiber is an interesting commercial project not because of what it offers or how (there’s nothing very different there from your standard cableco) but because of who is offering it. As such, I doubt it will have a massive impact (or indeed any significant impact) on the US broadband market unless Google was suddenly to decide that they want to do 10 of those projects. Which seems unlikely as I read things right now.