Why Google Broadband Finally Makes Sense
Yesterday Google declared it was getting in to the broadband market. As Chris Dixon pointed out, this was about the fourth time. What’s different this time and are they really going to do it?
I don’t think the timing of their announcement is coincidental. Recently there have been significant developments with respect to the technology that Google would use to deploy 1 Gbps home connections (~200-1000x faster than what’s typical today).
Here’s why it finally makes economic sense for GOOG to pursue broadband access:
1) Fiber to the Home technology is ready: FTTx (the ‘x’ stands for ‘home’ or ‘curb’) relies on a technology called Ethernet Passive Optical Network (EPON). In January there was major movement in this space, when the king of networking chipsets, Broadcom, acquired a company and entered the EPON market. FTTx has been hyped as the next big thing for over 10 years. But only recently has it really started to take off (this is in large part due to the dominance of Ethernet). Marvell Semiconductor also announced two weeks ago they are entering the EPON market.
Coincidence? No. Broadcom and Marvell are basically the two most dominant Ethernet/networking companies in the world. Their timing is less prescient of Google’s announcement, and more of an indication that the industry is ready for mass FTTx deployments. Broadcom is famous for waiting to pounce on markets until they are rising from the trough of disillusionment, and are ready for mainstream. There are now multiple vendors competing in this space, and costs for premise equipment will fall rapidly from today’s levels. Much of this has to do with China’s acceleration of its own EPON deployments.
2) Broadband wireless is also ready: Google will almost certainly not get into the business of wiring all the way to a home. This would make the endeavor inordinately more expensive because they would need to dig up streets and lay fiber to homes and have truck rolls to administer the premise equipment.
Instead, Google will aim to provide the last portion of broadband access via wireless technology. Curb to home delivery could very likely be through a 4G technology called WiMAX, which just entered primetime in 2009. Intel, a huge proponent, is ramping and costs will drop precipitously on this technology. Recall that Google was one of the big proponents of WiMAX and an actual investor in Clearwire (a WiMAX focused service provider affiliated with Sprint) – they are obviously very close to the technology. A Google broadband rollout would also prod the FCC to accelerate use-case analysis for converting TV white space to wireless broadband, something Google really wants. Though WiMAX would not support 1 Gbps connections, other short range wireless technologies are evolving which can help get there (WiFi is also a possible alternative).
I believe Google had financial models ready which incorporated these costs, or at the least are aware of the market buzz (no pun intended) around EPON, and are now finally ready to pull the trigger. They didn’t want this “experiment” to cost more than $1B or so. GigaOM did some analysis of the costs here. I believe these are too high, since commoditization happens at a fervent pace under high volume deployments (chip & equipment vendors would never imply this to analysts or reporters).
So this time could certainly be different – Google may follow through on building a broadband network. As wall street better understands these economics, it could be viewed as a value-add project for Google since fast broadband would allow GOOG to deliver a ton more video content and ads. But Google will not become a service provider. Their deployments will do just enough to spur Comcast, AT&T and others to accelerate their own broadband / EPON deployments. This would be a definite win for consumers.