September 24, 2005
Google the ISP?
(Note: I'm going to try to keep this as non-technical as possible, but it's not going to be easy! I'll explain some of the terms as I go along so hopefully it makes sense to everyone. The term explanations will be in italic blocks of text.)
So what the heck is Google on about?
Back in January of this year there were job postings on Google where they were looking to hire people who were experienced in negotiating "dark fiber" contracts. An original story about this development still exists over on c|net news.
The term Dark Fiber refers to fiber optiic lines that have already been laid, but are not in use. In years past a lot of companies laid a lot of fiber. Much of it has never been used, some was used but is not now being utilized. Usually because whoever owns it has either gone out of business or ended up in bankruptcy.
Word has it that Google has been quietly and happily snapping up these dark fiber contracts at bargain basement prices. Because of the current climate in the Telecom industry, as well as because of the current glut of bandwidth that's available out there, these networks are able to be purchased for far less than it would cost one to lay it themselves.
It was originally thought that Google was buying up the dark fiber to beef up their own network and to help reduce costs so that they wouldn't have to write that big check they've been sending to the likes of AT&T each month. The dark fiber purchases made sense when viewed in this light, so nobody from the outside looked any farther for explanation.
Slightly later in the year Google began letting RFP notices to relevant tech firms regarding building a DWDM fiber optics network. The RFP process was reportedly completed earlier this month and Google are now reviewing the bids.
DWDM stands for Dense Wavelength Division Multiplexing and can drastically increase the bandwidth one can send across a fiber optics network. All very techie, and you don't really need to grasp the minute details. Just know that we're talking Bleeding Edge technology here. Only a handful of the largest telecom providers have anything like DWDM available on any part of their network; And even those had to pay a lot more to build it out than Google will, because the costs have fallen quite quickly.
One source, IP Media Monitor suggests that Google could build an entire US network for under $100 million --not counting what they've already spent on the Dark Fiber to this point-- based upon information received from those in-the-know. The full story is here, though it does require a free subscription to read.
The increased bandwidth ability and dynamic possibilities of a DWDM network would be very useful if Google were going to start delivering high-bandwidth applications. Such as Video, Audio or VoIP (Voice over IP) services. The VoIP component again makes a lot of sense considering their recently released GoogleTalk software.
However such a network could also be valuable for other things. Some who have seen the actual RFP requirements have stated publicly that the details point to idea that something more than Video/Audio/VoIP is in the works.
A nationwide DWDM networkd could be very valuable if say --for the purposes of discussion-- Google wanted to jump into the ISP market to compete directly with the Telecoms and Cable broadband providers. They would most definitely have the ability to roll out a nationwide broadband network with limited financial exposure on their part. The only problem with this concept being that Google would still be constricted from doing so because of the whole Last Mile problem that every other provider-wannabee has stricken with.
The deal has always been that the big telecoms and cable networks, who happen to own that Last Mile network to every home or business, really don't want the competition. And since they own the wire/cable they can charge pretty much whatever they want to anyone who wants access to their local network. Thus severly limiting competition and keeping broadband prices artifically inflated.
But now comes along Google WiFi, something that first came to my attention a few weeks ago.
Currently it's a Beta program that they're running in the San Francisco area. In fact, --from here in the Midwest anyway-- wifi.google.com redirects to the main Google page. So I can't even give you a good URL that explains the details. However I can still reach their WiFi FAQ page and Download page, which at least proves that I wasn't seeing things before.
Google WiFi, as I understand the concept, is basically a Secure Access Wireless Network. Sort of similar to what many people have in their homes and offices where they can set up a router that will send and receive encrypted data, rather than run CAT5 cabling to each machine. This Secure concept is different from most WiFi networks, simply because of the fact that it's Secure from your computer to the Radio Station and back.
Google WiFi would more than take care of the Last Mile problem I mentioned above. Plus they have this Dark Fiber Network that they've already purchased. And they're reviewing RFP's to fill in the gaps, as well as bring the network up to specs that are better than anyone else can currently offer.
So is GoogleNet far behind? Is that what they're going to do? Compete head-to-head with the big telecoms, cable modem providers, MSN and Yahoo/SBC via Secure WiFi connnected to a vast nationwide fiber optics network?
Again, I must point out that this is all total speculation on my part. However the parts and pieces they've already assembled certainly do lead me to this (rather shaky) conclusion!
The thing about using WiFi, with the way it's been expanding its reach recently, is this: It's quite possible that a single Radio Station tower --each of which only runs about $30,000-- could easily service homes and businesses for a 20 mile radius. Forget about the Last Mile, it's not a concern with today's WiFi capabilities. In pretty much any metropolitan areas the number of users utilizing a single tower would become a more serious issue than the reach of the signal.
Plus, even a few WiFi Radio Towers in any given area would allow broadband service to rural customers who at present have no broadband options. We're talking similar to Cellular Towers here. If you can get a cell phone signal you should be able to use WiFi from the same tower, in theory.
Are these the markets GoogleNet will be going after? The rural component makes definite sense, though we're not talking huge numbers here. Though when you're only talking $30,000 per Radio Station tower, you wouldn't have to have huge numbers to make it a profit center after a few years.
The experts are saying that Google could certainly roll out the beginnings of a nationwide network in under a year if they chose to do so, assuming their San Fran beta test of Google Wifi goes well. And the estimates of the total cost are very attractive for a behemoth like Google.
They're talking in the $3-4 billion range for a network that has complete nationwide coverage. Maybe less if Google can strike deals with some of the Cellular companies that already have existing towers that have space available. Or to put it another way, just about the same amount that eBay recently paid for Skype, if you add up the cash plus stock options eBay handed over. Skype is just software, not some sort of network with tangible assets.
No matter how you look at it, it's all that much to become the major provider of Internet services in the US market. And it's been said that Google is currently sitting on a stockpile of some $7 billion in IPO cash while deciding their next move.
In a few years a person in the US may be able to be completely wireless. Wireless PC's at home, Laptops that you can use at home or anywhere else, Handhelds that will work from anywhere. And you can take any of the above to a completely different city and log onto the same service provider with no hassles!
It would be a different world, wouldn't it?
And what other advantages would such a notion give Google that they now struggle with?
The Personalization and Localization of Search.
Today they don't have a good way to tell where you are located or what your personal preferences are. Throw in the above however, where you have to install Google software on your computer and they know exactly which tower you're coming through... Suddenly both personalization and localization become trivial issues.
While I'm on a roll with my speculation, let's throw another monkey wrench into the mix.
Google already has all of these advertisers lined up for their Search component, right? And their Gmail delivers ads as well. What would happen if Google made the initial investment to build out this new Secure WiFi Broadband Network and then made it 100% free to use, in exchange for the advertising revenue potential?
Free Broadband? Everywhere? I'm not sure I can even fathom the branding power of such a concept! But if not free, certainly at a reduced cost from today's broadband pricing, given that they'll be able to easily monetize other things.
As a final warning, this is all complete speculation on my part at this point in time. But all of the pieces of the puzzle fit.
I for one wouldn't mind seeing the boys and girls at Google, with their unusual approach to everything, take on the telecom/cable industries and their crushing hold on broadband.