Triangle preens for Google:

"It's almost incomprehensible to me," said Lee Bolin, a Raleigh resident and IBM software engineer who contacted city officials to urge them to apply. "It could mean a real shift in how we look at the Internet and how we deliver content."

Comments

[Roch101]

From the article: "Most homes now get speeds of 50 megabits per second."

Is that right? In the Triangle, maybe? Time Warner Cable's Road Runner with Power Boost advertises speeds up to 10 Mb/s and mine are consistently around 5 Mb/s, about a tenth of what the article cites (and that's downloads, uploads are a measly 0.36 Mb/s)

[justcorbly]

Roch, I'm on TW in the Triangle. TW's speed test has me around 8500 kb/s. Every other test I run has it in the 2000-3500 kb/s range.

[Roch101]

Thanks, JC. I suspect they added a zero.

[Steve]

I'm also on TW (in Durham) and just got 22M via Speakeasy's Speed Test. Nowhere close to 50. Upload is the same puny 0.36.

[John Tasker]

All of the communities I've looked at are trying to say who they are that makes them deserving of Google's plan. I don't see any saying how they can help Google make money or increase market share. Ultimately, I would think that would be Google's goal. Perhaps we can all cut our neckties in half and wander around town feeling like we therefore "belong".

[Ed Cone]

JT, I agree that explaining why we're good for Google, and not just vice versa, is important.

But Google's core business is not and most likely will not be infrastructure -- the company is trying to spur broadband construction and adoption, which will be good for its information businesses.

It's kind of like a car-maker paying to build a roads in a town to prove how awesome cars can be. So for GOOG, it's not about market share or revenue in the broadband business.

To some degree, the question of what GOOG broadband does for a place is relevant -- the better a test market you are in terms of finding utility for broadband, the more convincing the test.

That said, I've tried to push the idea that GSO is a good test market because of what we bring to Google -- a real slice of American life that will tell the company something about broadband in flyover country.

[liv]

Maybe we could put them out of the running by using Mafia tactics?

[Ross Myers]

Ross - can you point to Google asking a city to pay for the network themselves?

No I can't Jim...that is why statements like ...all indications are that Google is going to pay to build the network. raise questions. I can't see how Google could get a black eye asking for us to help foot the bill if they have never indicated that they won't and past history demonstrates that they might.

The scale of Google's investment should not be a debated point now.

I agree...from the link Google using the 'hundreds of millions' phrase

Feb. 26 (Bloomberg) -- Google Inc. said it may spend as much as “hundreds of millions” of dollars on an experimental broadband service that offers Internet speeds 100 times faster than current networks.
The cost of the test project, announced this month, isn’t known and will depend on demographics, the lay of the land and the number of households that use it, said Richard Whitt, Google’s Washington counsel on telecommunications and media issues. The company hasn’t determined the location or size of the network, which could serve 50,000 to 500,000 customers

There are great opportunities in Google's offer for our communities to consider...without getting off track on stretching words like...may...isn't known and could serve.

As to a GSO/W-S competition.... I'm of the opinion that we should pursue the possibilities of making a regional bid.

[Jim Caserta]

Ross, I think you were responding to one of my comments from another post. Here is the logic, echoed by the Rhino article Ed linked to a couple weeks back.

After the network is completed, Google, or whatever provider is running it, will need people to shell out in the ballpark of $50-$200/month for a service.
If residents & consumers feel put out by the up-front money they need to put up for the network, that will make them less likely to sign up for service.
If the sign-up rate is too low, setting up the service will be unjustified.
Google has a ton of cash, why piddle over $1M on a $500M project.

This is different than the Lenoir situation because the residents of Lenoir aren't paying Google for anything on a monthly basis.
combine
Ed - Google is already in the infrastructure business, given all the dark fiber they've bought over the past 5 years. Google spent $6.5B on cap-ex in combined 06/7/8.

Our [Google's] business depends on increasing use of the internet by users searching for information, advertisers marketing products and services and web sites seeking to earn revenue to support their web content. If the internet infrastructure does not grow and is not maintained to support these activities, our business will be harmed.
Our success will depend on the continued growth and maintenance of the internet infrastructure. This includes maintenance of a reliable network backbone with the necessary speed, data capacity and security for providing reliable internet services. Internet infrastructure may be unable to support the demands placed on it if the number of internet users continues to increase, or if existing or future internet users access the internet more often or increase their bandwidth requirements. In addition, viruses, worms and similar programs may harm the performance of the internet. The internet has experienced a variety of outages and other delays as a result of damage to portions of its infrastructure, and could face outages and delays in the future. These outages and delays could reduce the level of internet usage as well as our ability to provide our solutions.

[From Google's 10-k, filed 2/2009, accessed 3/2010 via Edgar]
That says to me that Google is pretty interested in infrastructure. I would not argue that Google's long term plan might not be an ISP type business, but long-term is a relative term, and short-term, they definitely are going to offer ISP-like services. Isn't Time a magazine, and Warner a brother who used to make movies? What a company looks like today does not completely dictate what they will look like tomorrow.

[Ed Cone]

JC - GOOG is definitely interested in, and invested in, infrastructure -- and as we're discussing here, interested in furthering that to a non-trivial degree -- but as I said, it's not GOOG's core business.

That distinction matters in relation to JT's question, insofar as the purpose of this fiber initiative is not necessarily to generate immediate ROI on the local network(s) they build.

[Jim Caserta]

Maybe I'm being too technical, but the project would have 3 ROI inputs: (1) up front capital expenditure, laying the fiber, (2) monthly cost of maintenance/upkeep of network (3) fees paid by users of said network. (3) - (2) needs to be positive. I would estimate a rough (3) - (2) of $50/mo, $600/yr. Rough estimate of (1) is $2000/home. That's 30% ROI or very good. Details will probably lower, but a positive ROI will be needed. The question is how much ROI can be generated, and Google's case can spur more fiber investment.

[Ed Cone]

And I may be too casual in my use of ROI here -- I'm really saying GOOG is looking for a showcase project to boost high-speed penetration more than it's looking for a new profit center in the fiber build-out business.

[Ross Myers]

You are correct my response was directed at your thoughts shared in this thread:

Chicken or Egg?

After the network is completed,...or ....If the sign-up rate is too low, setting up the service will be unjustified.

Are you assuming Google's plan would be to come in and pre-wire every home in the City? How would you see that working? Would Google need to knock on every door and request rights of way to install service that a homeowner may or may not chose to use or does the City have the ability to provide a universal right of way across personal property for a service that benefits just that homeowner?

Google has a ton of cash, why piddle over $1M on a $500M project.

They might not piddle over $1M but what if that number was $250M?

Would it really cost us $500M to wire our City to be prepared to offer service to everyone?

I think we need to get a handle on what the estimated costs of offering High Speed Internet to our Community would be(not necessarily Google's costs).

If Google only picks a few places...and we are not one of them...we should be prepared as one of a couple hundred cities and towns in the Country with a "shovel ready" HSI plan to go after some of those Recovery Act $$$'s with.

[Jim Caserta]

In my mind, showcase project is defined by profitable project. If they come in and it costs $5k to fiber-up each user's home, and only net $25/month/user, you are at a ROI of 6% - good but not great.

Google will not run fiber to every home in the test community. Take a street - they will run the fiber along the main path, as you sign up, they split the fiber signal and route to your home. I think the main path rights-of-way are not opposable by a home, from my limited knowledge. So a street of 10 homes with 1 user will have a higher per-user setup cost than a street of 10 users with all 10 signing up. The $2k is a rough number which is taking total area investment/# of area users. Because you are running higher quality fiber longer distances on the more main trunk, possibly above ground, you get better cost structures with higher penetration percentage.

Google's purpose is to show that fiber access is profitable for the provider and pleasurable for the consumer. Google is taking a chance because it can, and it benefits its other benefits. Verizon built FiOS because it had to in order to make up for lost traditional POTS users and stay competitive with CATV systems. Google's cost-per-home average, and fees-per-home-per-month numbers are critical. Say FiberMagic wants to IPO, and their goal is to build fiber networks - how would google's profit margins in this test case impact their IPO? That is why the test-case profitability matters, and also why Google might prefer a less tech-centric area to show that the idea has more broad appeal than just in traditional tech centers.

If an area was going to foot 1/2 of the network creation bill, they might as well do it themselves as Google wouldn't be bringing much. If google set something like that up, they'd need to 1/2 share the fees, and again at that point you'd need to weigh pure public options.

[Ross Myers]

Thanks Jim...so taking the scenario....

Google will not run fiber to every home in the test community. Take a street - they will run the fiber along the main path, as you sign up, they split the fiber signal and route to your home. I think the main path rights-of-way are not opposable by a home, from my limited knowledge. So a street of 10 homes with 1 user will have a higher per-user setup cost than a street of 10 users with all 10 signing up. The $2k is a rough number which is taking total area investment/# of area users. Because you are running higher quality fiber longer distances on the more main trunk, possibly above ground, you get better cost structures with higher penetration percentage.

What would be your best guesstimate of what the start-up cost would be to set up the Infrastructure to take that approach in a City like Greensboro?

[Jim Caserta]

My guess for start-up costs, as I've written here a couple times, is $2k/home. That is what I remember second or third hand mentioned about Verizon's costs. They were many places using above ground cabling, where I would guess a lot of Greensboro would be underground. This is an advantage of older neighborhoods, as I believe above-ground is cheaper than underground. That estimate is worth what you're paying for it, but there are other estimates available online and it is not unreasonable. Again, physical issues of a locale would impact it, as would adoptions rate as I mentioned above with the two different street examples. Google is going to take the physical issues, gov't and regulatory issues, and projected adoption rates and figure a test location they can make adequate profit. I think Greensboro's (and Winston-Salem's) adoption rates would be inline with other cities of similar demographics, but I would imagine off-campus UNCG & NCA&T students to have higher adoption rates than non-students.

I am nowhere near the person to consult wrt costs & free-cash flow estimates for this project, but it is obviously something that deserves consideration.