This page is an archive from our previous website. Please check out our new website where you can read new COMMENTARY eNewsletters, TELL IT LIKE IT IS blog posts or Press Releases.
 
One last point here, what if Cyren Call, Morgan O'Brien's company that is now advising the Public Safety Spectrum Trust (PSST), ended up bidding for the D block?
 
Blog

700 MHz: Business as Usual?

Monday, December 17, 2007
 

In almost every industry publication you pick up, every wireless blog written lately and even in the business publications, you are likely to see articles on the implications of the 700-MHz auctions. The December 3, 2007 deadline for filing intent to bid has passed and little by little, the names of the potential bidders are finding their way into the press (the complete list is due from the FCC shortly) and speculation is rampant that this auction will result in significant changes for the wireless industry.

 

In reality, we are already seeing the beginnings of some of these changes. There is more discussion about open access networks where any approved device will be able to download anything from the Web, new pricing models are evolving, next-generation technologies are being chosen by incumbent operators and new players are looking to make a dent in the wireless industry come auction time.

 

So far, we know Clearwire, Sprint Nextel, T-Mobile, Time-Warner, Microsoft, Apple and DirecTV are not planning to bid for the spectrum. We also know the bidders will include AT&T, Verizon Wireless, Google, Frontline, Metro PCS, Leap Wireless (Denali), Cox Cable, Qualcomm EchoStar and CableVison. However, with the exception of Google's stated interest in the C band, and the assumption that Frontline is only interested in the D block (shared first responder) spectrum, we don't know which blocks these companies may be going after, nor do we know the entire list of the bidders.

 

Even so, we can make some educated guesses. For example, Qualcomm is probably interested in the E block, which is 6 MHz of unpaired spectrum next to its existing 6 MHz of MediaFLO spectrum. But could Qualcomm also be interested in the shared public safety spectrum? We can guess that Cox Cable probably won't bid on a nationwide footprint of spectrum, but perhaps it will pick and choose from the multiple licenses in the A and B blocks within its present cable coverage areas. It is also a pretty safe bet that both AT&T and Verizon Wireless will be going after portions of the A and B blocks to fill in cities where they are light on 850-MHz, 1900-MHz and AWS spectrum, and that both companies will be going head-to-head with Google for the C block. We could also speculate that both companies will be bidding on the D block of shared first responder spectrum, but that is not a bet I would make in Las Vegas at the moment.

 

Leap and Metro PCS do not appear to have the financial resources to try for a nationwide footprint and their model of serving specific metro areas has been working fairly well, so we might surmise that these two will be going after A and B block licenses for cities they don't presently serve and perhaps one or two of the twelve licenses in the C block. EchoStar and CableVison could be going for the C block, but the more bidders there are for the C block, the more it will cost the winner and I am not sure either of these companies has the stomach for spending $6 billion for 22 MHz of spectrum for a nationwide service.

 

As other bidders become known, I, along with a lot of other people, will be trying to figure out what blocks will be bid on by whom. Once the official list is out―and I assume there will be some surprises on that list―filling in a spreadsheet by company and blocks available will be a fun project. Once this spreadsheet is finished, the only other thing that could be guessed at is who has the most staying power and who needs the spectrum the most. In the recent AWS spectrum auction, T-Mobile was the neediest bidder by far because it did not have enough spectrum to build out a 3G network. It went after AWS spectrum with gusto and came away with what it wanted, which is the ability to construct its own 3G network on an almost-nationwide basis. The cable companies came away with an average of 20 MHz per market, but so far there has been little word on their plans for the use of this spectrum.

 

The real question in everyone's mind is this: When the dust settles and the bids are tallied, will we end up with business as usual because the bulk of the spectrum has gone to incumbent networks, or will we really be on the verge of a new era in wireless with new players with different types of plans and business models for the spectrum? In the weeks ahead, there will be plenty written about this question and I am sure everyone who writes about it will have a different answer. My view is that it could go either way, but I believe that to begin a new era in wireless it will take a company winning all of the C block (22 MHz) as well as the D block (10 MHz), paired with the 12 MHz from the public safety side. This would give the winner 44 MHz of spectrum to work with MOST of the time.

 

But we will have to wait and see. One last point here, what if Cyren Call, Morgan O'Brien's company that is now advising the Public Safety Spectrum Trust (PSST), ended up bidding for the D block? That would certainly prove to be an interesting situation to say the least!

COMMENTS: This is an archived post. Commenting is no longer available.

Scott Goldman - 12/17/2007 19:59:31

Interesting analysis, as usual, Andy. I agree with your assessment of the various interests of Qualcomm and Metro/Leap. Google's pretty much a slam-dunk in my mind for one of the national blocks, but other than the Big G my guess is that it's going to be business as usual with AT&T and VZW taking home the spectrum prize.

There are probably people much smarter than me trying to figure this out and coming up with different answers; my analysis is based on anecdotal observation over the course of my years in wireless. Simply put, while these other ersatz wireless operators might be able to sense the potential in these airwaves, the incumbent carriers understand the _value_ of the spectrum - and therein lies the difference.

So, in the final analysis I think we're going to have a little of the "new era" via Google's near-certain entry. The incumbents will be forced to respond to Google's innovation, "open" approach and more-nimble decision making, yielding an increasingly vibrant market despite the fact that they are of the "business as usual" school.

As a side note I do think that there is a large potential contract at stake for one of the carriers (Sprint/Nextel? - it certainly won't be AT&T or VZW if they win other blocks) to build out the system for Google once they win the auction. They don't have the expertise to do that and hiring the personnel to do it would be too lengthy a process for them as they're accustomed to moving at software, not hardware, speed.

As usual, it's just my 2 cents. I'm anxious to hear what others think.

Scott

David Boettger - 12/17/2007 21:58:58

"AT&T and Verizon Wireless will be going after portions of the A and B blocks to fill in cities where they are light on 850-MHz, 1900-MHz and AWS spectrum."

I have no doubt that this statement is true, but just think about the implications: How could *anyone* be "light" on spectrum if they're holding pieces of 850, 1900, and AWS?? The auction hasn't even started yet and it's as if we've already thrown up our hands and resigned ourselves to the fact that AT&T and Verizon Wireless will tighten their strangle-hold on the market. Am I the only one who laments this? Am I the only one who thinks this is a failure in federal spectrum allocation policy?

I will admit, I don't know what the solution is. Changing allocation rules to favor (or disadvantage) this or that company is risky in terms of its outcome and would represent government manipulation of markets, which isn't fair. But it would seem to me that when deep-pocketed contenders can buy spectrum that they don't even need and sit on it until the day (possibly years in the future -- or maybe never) when they do need it, there's something wrong with the system. (I've personally seen the big dogs do this: lighting up one carrier in otherwise empty spectrum just to comply with the letter of the FCC law.)

I have often wondered why radio spectrum -- a public asset like any other public asset -- isn't treated more like logging or oil exploration rights, with limited-term leases and revenue sharing with the government. I can't imaging AT&T and VZW putting up with that, though.

Andrew Seybold - 12/18/2007 11:06:57

Scott--thanks for your comments--your 2 cents is always worth reading--
David--I here what you are saying but I guess that you must come from the land of the Internet and not the land of wireless--when Google needs more bandwidth they merely purchase another FIOS link to the Internet, when AT&T needs more bandwidth to serve their customers they have to either deploy additional spectrum--if they have it, or build new cells sites closer together--in any event the bandwidth is shared--and you can only get so many bits per Hz down a wireless link--so you need more spectrum in order to accomplish that--this is true world-wide, not just in the United States--if Google or someone else were to win the C block, that is 22 MHz of spectrum (11X11) and build a nationwide network they would run out of bandwidth a lot sooner than AT&T or Verizon who hold more spectrum--and some of their customers would be unserved, or would suffer dropped calls or poor data speeds--compared to the rest the world our network operators have LESS spectrum per company than in every European Country and elsewhere.
Best regards

Andy

Scott Goldman - 12/18/2007 18:22:44

@David:

Just one more point in regards to your comment about limited-term leases... the rights (licenses) to provide these services are, in fact, limited terms. If I recall correctly (Andy, please correct me on this) they are 15-year licenses. Now, that said, there's never - at least as far as I know - been a case of a carrier losing a license at the end of the term.

To take it a step further (just to play the devil's advocate here) what logging or oil exploration company do you know of that has paid _billions_ to the government for the right to those limited-term leases? None. Have any of them ever lost their lucrative leases? None. Finally, think about what the levels of profit are that come from the ground vs. those that come from the air: ExxonMobile: $10.25 BILLION in profit in the last quarter of 2006. Verizon (the entire company, not just the wireless division) $1 billion in profit during the same period.

As I've frequently pointed out to others I'm not a defender of carriers - you'll hear me complaining loudly and vigorously when they screw up (which is often) or when they receive even more largess from the government (even more frequently). That said, wireless is a tough business to make a buck in - it costs the average company $800 to sign up a new subscriber and somewhere between $.5-$1.0M to build a new cell site - presuming they can get it approved and past the local zoning boards.

The resource isn't easy to expand. Spectrum (admittedly, like oil) is finite. Extracting the most usage from it can be costly, time-consuming and aggravating. The government's policy may not be the best, but compared to the prices in Europe and Asia for similar services I think we're doing pretty well.

David Boettger - 12/18/2007 20:43:40

Andy -- my entire 15-year career has been in digital cellular (the big, old telecom iron). I'm not sure where your point about Google came from as I didn't mention that company -- perhaps you are referring to a previous post of mine. But regardless, as you know, there are three ways to increase the bandwidth available to users at any location: You can reduce the path loss between the sites and users (put the sites closer together), you can use more spectrum at each site, or you can use a more clever radio interface. It's a tri-variate equation, so I don't know how you can make the blanket claim that Google would run out of bandwidth any sooner than anyone else: Google could always build more sites to accommodate the traffic, just as companies like T-Mobile USA and MetroPCS do now. Even if this were true, are you really positing that AT&T and Verizon *deserve* the bandwidth because failure to allow them to get some would result in their subscribers being under-served? You're not under retainer from these folks by any chance, are you? (kidding)

Scott -- You make some interesting points. I would point out that oil companies pay royalties over time (I believe that, in Alaska, the tax rate on extracted oil is something like 50%) rather than up-front: It's a gift that keeps on giving. And Verizon derived substantially all of its profits from the USA, where substantially all of its spectrum assets are located, but which represents 5% of the world's population. In contrast, ExxonMobile's oil assets and customers are world-wide. So it's hardly a valid comparison. Besides, I'm not lamenting how much money Verizon makes: Good on them if they can make a few cubic megabucks. My concern is with public assets being permanently transferred to entrenched players whose vested interest is in preventing competition and maintaining their operational status quo.

I know with some certainty that your $800 subscriber acquisition figure is way high, and your site construction figures are certainly on the high end of things (you must be assuming that you start with raw land with no road access). But this is beside the point. The carriers' choice in operating models should not be a driver of the spectrum allocation policy. Can't make a buck in cellular? Boo-hoo. Fix your operating model or do something else. Just because spectrum allocation policy in Europe and Asia comes right out of the Stone Age doesn't mean US policy is optimal -- it's just better than anything else out there at the moment.

Scott Goldman - 12/19/2007 10:19:42

@David:

It seems that we're in agreement about one thing - the U.S. spectrum allocation policy is similar to democracy, if you'll remember Churchill's famous quote: "It has been said that democracy is the worst form of government except all the others that have been tried."

Recall, if you will, the other processes that have been attempted: Comparative hearings and lotteries. Having given the first testimony in a comparative hearing before the FCC back in the early 1980s I can tell you that we'd still be waiting for licenses to be granted in L.A. if we continued along those lines. And the lotteries? Sure, great idea if you want the local hairdresser or plumber running the system in your area (or cashing out, as they did, to the "majors" and moving on, thus turning it into an auction of sorts).

I do not believe that my $800 cost is too high when you fully factor in the complete cost of acquisition which, by definition, would necessarily include subsidies for phones, additional internal resources that must be added incrementally for each customer (e.g., you'll need another customer service representative for every X number of new subscribers). These factors, combined with the massive amounts of money spent on advertising, raise the cost of acquisition significantly.

Furthermore, the need for resources to _retain_ these customers becomes ever more acute as we reach a penetration rate of 70-80%.

You might consider my site construction figures to be high when factoring in the cost of hardware - radios, racks, shelters, cables, etc. However, the additional soft costs of attorneys, expert testimony for zoning issues, hosting neighborhood meetings to assuage health concerns, etc., adds to that hardware cost significantly. Therefore, I'll stand by my estimate.

Finally, yes, the oil companies do pay royalties. That has not, however, stopped them from making massive profits (although the profit margin is less than one cent per gallon, according to oil industry estimates - giving you an idea of how many gallons are actually sold every year) on a lease that is essentially given to them at the front end of the process. Sure, they're paying, but it's a different business model than having to lay out $6B _before_ even punching a hole (or building a cell site, in our example).

Frank Bulk - 06/03/2008 02:04:58

I do think $800/customer is about double the typical CPGA. I've always thought of it in the $400 range. See page 3 of this report::
ftp://ftp.scansoft.com/nuance/care/cpga_metricreport.pdf
for further substantiation.
Frank

Scott Goldman - 06/03/2008 09:10:26

Thanks, Frank - I appreciate you taking the time to respond. Having looked at the report I'm going to stand by my $800 approximation. The report is now two years old and states that the first reason that costs are in the $500 range is that the carriers are offloading marketing costs to MVNOs. As we now know MVNOs are failing with great regularity and offer little help to the carriers with marketing. While I do concur that the other factors are reasonable (particularly activation cost reduction, if the Apple iPhone activation model is any indicator) I believe that the increased cost of customer service required due to the complexities of smart phones adds significantly to the CPGA.

Frank Bulk - 06/03/2008 13:07:24

I haven't come across any financially-oriented press releases that document a CPGA as high as you state. Do you have any third-party substantiation?

Smartphones still make up a small percentage of all phones sold. There are a few analytics firms that prove that out.

Scott Goldman - 06/03/2008 13:15:24

Only through discussions with cellular carrier personnel.

Without anything other than anecdotal knowledge I'd say that smart phones make up a small percentage of all phones sold but a disproportional percentage of customer service resources.