31 December 2007

Miscellaneous topics ...

I had a few items that I wanted to blog about over the past couple of weeks. So, in an effort to clear out my backlog, let me summarize them here.

  • The US GAO published this report, which is fairly critical of the US FCC and other agencies with regard to the upcoming conversion to digital television. The "switch off" is just over one year away, and consumer education has been basically non-existent.

    Update (2008-01-03): The FCC published their Third Periodic Review of the digital TV transition. No mention of the GAO report that I could find ...

  • On a similar note, the UK regulator Ofcom issued this report that examines alternatives for using the "digital dividend", which arises from the sale of the spectrum formerly used by analog television.

  • Changing gears a bit ... this article in BusinessWeek gives hope to the fans of Municipal Wireless. The Spanish WiFi operator FON seems poised to enter the void that Earthlink is leaving after it reevaluated its participation in this market. They have a different business model. Will they find more success?

  • There are a couple of interesting items on the international front.

    • Take a look at this (from Swivel, one of my favorite sites), which shows the dominance of prepaid wireless in China's market.

    • Gordon Cook unearthed a fascinating, detailed report on providing wireless to Nepal. See Gordon's post here. I have been unable to locate it on the World Bank's website.

    • In this article, the possibility of a "natural limit" to Kenya's mobile market is raised. I am always suspicious of these kinds of projections, because they normally assume that the future will be driven by forces similar to those driving the markets today. History has shown us the disruptive power of wireless technology, after all ...

  • Speaking of natural limits, I have been interested in the pending conversion to IPv6, which has been prophesized for several years already but which has not yet happened. This article over at CircleID provides more grist for this mill. Is it different this time, or should we expect more of the same?

Happy New Year, dear readers!

Australia censoring the Internet?

This article over at TechCrunch reports on a new law in Australia designed to provide a "clean feed" of Internet content into the country. This law is an initiative of the newly elected Labor government. I had blogged earlier about the implications of this election on telecoms ... but I hadn't been aware of this element of the electoral campaign.

I think this points to the tension between information openness and social preferences that every country has to deal with in the Internet age. Do you think that censorship of this kind is an effective way of dealing with it?

Inside the spectrum auction

This article in Forbes provides an interesting insight into the workings of the upcoming 700 MHz auction. Since Forbes doesn't do permalinks, I will quote from the article more liberally than I prefer to:

If you're a 150-person start-up going up against the likes of AT&T, Verizon and Google in the upcoming auction of wireless spectrum, you look for any advantage possible.

Towerstream, a tiny Middletown, R.I.-based fixed-wireless Internet provider, is hoping telecom lawyer Dee Herman will give it a leg up against the competition. Herman, a principal at Bennet & Bennet, is part of a cottage industry of lawyers, consultants, engineers and economists that companies are hiring to boost their chances in the closely watched Federal Communications Commission auction that could redefine the telecommunications industry.


Experts say larger companies are more likely to have in-house teams that assess the value of particular slices of spectrum and come up with different bidding scenarios Those with fewer resources are hiring outside consultants and lawyers, who may, in turn, work with bankers or economists to devise strategies. Even Google, which has set up a war room at its Mountain View, Calif., headquarters, is consulting with auction experts and game theorists to gain an edge over the competition.

Consultants and lawyers usually file the applications on behalf of companies. They also get tapped for a multitude of other tasks, including supervising auction deposits and payments and deymystifying FCC regulations, such as the FCC's anti-collusion rule. Consultants and lawyers also may help craft business plans, determining, for instance, how particular licenses complement a firm's projects or other spectrum holdings, how much a particular slice of spectrum is worth and what an appropriate bidding strategy would be.

These outside consultants describe their role as streamlining a somewhat Byzantine process. "The process of bidding, looking at results and strategizing with our clients to figure out their next move can be constant, consuming and all-encompassing," says Herman, Towerstream's outside legal counsel, who is also representing several other bidders interested in different geographic areas. Some clients, he says, enjoy the frenzy and get intimately involved in the details; others prefer to have updates e-mailed to them.

Experts differ on the advantages of outside help but agree that the real payoff may simply be saving time in what can be a weeks-long process. In spectrum auctions, bidding stays open until no new bids are received, at which point the entire auction ends. That helps ensure no one cuts in and outbids someone at the last minute. It also means serious competitors must keep a close eye on the auction until the end.

Compounding the complexity, FCC auctions tend to begin slowly, with two to three rounds per day, then ramp up to as many as eight to 14 rounds. A smaller spectrum auction held last fall with 168 bidders went 161 rounds over 29 days. The 700 MHz auction is expected to last longer, given thee greater number of participants and the higher perceived value of the spectrum.

Despite the plethora of options, many players will probably go it alone. The FCC says it has taken pains to make the auction process easy to navigate. A Jan. 22 mock auction is scheduled to familiarize participants with the bidding software, which has been fine-tuned over the past two years and includes a section of frequently asked questions. Like the real auction, the mock auction will have anonymous bidding, which means companies will know the amount of the highest bid and whether they submitted it, but not who they're bidding against. Limiting information in this way could help level the playing field, experts say.

While I believe that auctions are the most efficient way to do spectrum assignment, I think this article highlights some of the shortcomings of this mechanism. It seems clear that smaller participants stand at a relative disadvantage. Can you imagine how it might go if governmental agencies (eg., police and fire departments) had to enter this auction for their spectrum needs as well?

18 December 2007

US consumer expenditures on telephone services

This item has gotten a lot of press in the blogosphere. It represents what we have all been experiencing. If you look at the raw data, adjusting for inflation, the total expenditures on telecom have remained steady. There has been a substitution effect, though, where expenditures on mobile telephony has increased to 0.9% of income from 0.4%, while expenditures on landline telephones have decreased to 0.9% of income from 1.5%.

If these trends continue, we can expect expenditures for mobile phones to exceed landline phones in the US this year (2007).

13 December 2007

Thoughts on Google and business strategy

I succumb to the temptation to write about Google from time to time (why? search me... ha ha). This article was prompted by this item by Nicholas Carr and this one by Om Malik. Both of these analyses are focussed on strategy rather than technology.

In his analysis, Malik writes:

By squeezing the supply chain as hard as he could, he turned Dell into a fearsome (and loathsome) competitor. With his help, the supply chain for the PC era came to consist of foundries, ships, U.S. assembly plants and UPS trucks. Google (GOOG), with over $200 billion in market capitalization, is following a similar strategy, fine tuning and adapting it for the Web & broadband.

Instead of trucks and assembly plants, however, Google’s supply chain is made up of fiber networks, data centers, switches, servers and storage devices. From that perspective, its business model is no different than that of Dell’s (DELL): Google has to deliver search results (information, if you want to be generous about their other projects) as fast as possible at as low a cost as possible.

To better understand Google and its business model, one needs to break it down into three data inputs.

  • Relevancy of results.

  • Speed of search.

  • Cost of executing a search query.

Malik goes on to argue that these foci are the driving force behind Google's efforts to build their own switches, data centers, etc. Indeed, Carr refers to this as Google's third innovation: "Google’s third great innovation — and it may well be the one most critical to the firm’s future success — is the design of its parallel-processing computer system." However, Carr notes that:
The way Google makes money is actually straightforward: It brokers and publishes advertisements through digital media. More than 99 percent of its sales have come from the fees it charges advertisers for using its network to get their messages out on the Internet.

This is useful because profit is, in the end, the difference between revenues and costs. I don't think that Google's efforts to build a switch can be explained in terms of cost alone, because the market provides switches at competitive costs. Instead, following Carr, I can only rationalize this effort as one that gives Google a special leverage in the way that it is able to improve its revenues. Going back to Carr:

Any understanding of Google as a business has to begin, I'm convinced, in the idea of complementary goods. In The Google Enigma, an article in the new issue of Strategy & Business, I argue that the wide scope of Google's interest and activity is a natural and inevitable result of the fact that everything that happens on the internet is complementary to the company's core business.

Regulation or micromanagement?

This item from Forbes reports on the process of gaining regulatory clearance for a sale of access lines. FairPoint Communications had agreed to purchase Verizon's 1.6 million landline operations in northern New England for US$2.7 billion (or about US$1700 per access line). Transactions such as this are subject to regulatory approval because they involve the transfer of an operating license. According to the article,

An MPUC staff report last month recommended that the proposal be rejected unless the companies satisfy dozens of conditions.

The most controversial conditions would require Verizon to lower the selling price by $600 million and make FairPoint cut its dividends to shareholders by 30 percent, spend more to expand high-speed Internet and meet stronger quality standards for service.

According to the Examiner's Report, this recommendation is a result of concerns about FairPoint's ability to meet the needs of the residents, i.e., FairPoint is highly leveraged, and so will be limited in its ability to accommodate adverse results. This despite the fact that the examiner's report mentions that the price of the transaction is "considerably less than the price of other recent transactions".

Do you think that this is reasonable regulation, or is this micromanagement by a governmental agency? If this transaction is not approved, would the citizens of Maine be better off with an operator who (apparently) does not want this business (since they are selling it for a low price), or with one who wants to be there and isn't financially as solid as regulators would like them to be?

Update (2007-12-27): According to this article in Forbes, Vermont regulators have rejected the Fairpoint bid. Quoting the article:

The ruling by the Vermont Public Service Board cited FairPoint's financial viability.

"The Board found that FairPoint had not demonstrated that it would be financially sound as it seeks to operate the newly-acquired territories in Vermont, Maine and New Hampshire -- a service territory that has five times the number of access lines as Fairpoint presently has," the board said in a prepared statement.

FairPoint, based in Charlotte, N.C., would have to borrow $2.5 billion to complete the transaction, the debt service on which could exert "significant financial pressure" when combined with operating costs and revenue projections, the board said.

State regulators, however, left the door open to a revised bid.

The article mentions that the union representing some of the employees was opposed to the sale. This also comes through in the papers from Maine. I wonder if they were shareholders ...

Tower sharing in India

This article in Forbes contained some interesting tidbits about tower sharing (in India, in this case). Quoting the article:

Last week, Bharti Airtel’s subsidiary Bharti Infratel, Vodafone Essar and Idea Cellular said they would merge their wireless-infrastructure businesses to cut costs and improve efficiencies. Vodafone Essar and Bharti will each own 42% of the new company that results--Indus Towers--and Idea will get 16%.

The companies all operate on the global system for mobile communications standard. They will merge their infrastructure assets, totaling 70,000 towers, in 16 of the 22 demarcated telecommunication zones. Media reports said Indus Towers would fund expansion through debt and could venture a public offering two years on.

Other facts about the Indian market from the article are:

India’s mobile market added nearly 8 million subscribers in October, the largest gain in the world, along with China. But intense competition has ensured tariffs are as low as 2 cents a minute.

[... snip ...]

... as record subscriber growth overcrowds the airwaves, Indian telecom players are now engaged in a legal battle over precious spectrum allotments from the government.

11 December 2007

AT&T network upgrades

Stories like this one in Forbes don't get a lot of press attention, but I think that they are worth tracking anyway. Since Forbes doesn't do permalinks, here are some key excerpts from the article:

AT&T Inc. said on Monday it has switched on its high-speed backbone network, which is designed to ferry data traffic across the U.S. four times faster.

AT&T has begun placing traffic on its so-called "ultra-long haul" network, which boasts a capacity of 40 gigabits per second, meaning consumers will be able to download large files quicker and more easily stream online videos to their computers. Carriers have been upgrading the backbone network - the underlying pipes needed to move data across extremely long distances - to meet the increasing demand in bandwidth-intensive programs and videos.


The company, which is deploying routing equipment supplied by Cisco Systems Inc., has upgraded 50,000 miles of its network and plans to connect 25 major metropolitan areas in the next several months. ... In addition to a faster connection for consumers, the upgrades will help ease the capacity requirements for the company's U-Verse Internet-based TV system.


While the network is the first in the U.S., Verizon Communications Inc. said that this month it would begin building a 2,000-mile backbone network connecting major cities in Europe.

Both companies plan to push the 40-gigabit standard in the U.S. and eventually upgrade to 100 Gbps.

The article doesn't mention it, but I think it is safe to assume that the "40 Gigabit" standard is, in fact, OC-768 (this article in Network World confirms this). NW also reports that this is AT&T's MPLS network. I'm not sure what the "100 Gbps" is ... OC-1536 comes in at approximately 80Gbps. Wikipedia reports that the OC-3072 standard is a "work in progress".

Could the 100Gbps bit rate be referring to 100 Gbps Ethernet (as this article in Wikipedia suggests)? That would be quite a departure ... and would suggest an explicit strategy to integrate local and long distance network standards. Ethernet has truly come a long way (pun intended)!

In light of the Comcast "network management" discussion, this is an interesting development. Do you think AT&T would be credible if they employed similar techniques on this new network?

06 December 2007

Wireless network openings

AT&T, of course, could not be left out of the limelight that has surrounded Verizon's recent announcement. As TechDirt reports, they now claim to have "opened" their network as well. As a GSM-based operator, it has always been possible for users to purchase an unlocked phone and put in AT&T's SIM card. This is basically what AT&T is announcing to the world.

It is clear that this misses the point a bit. In its announcement, Verizon suggested an "unbundled" pricing scheme for these new devices. AT&T makes no such offer ... as a user of an unlocked phone, purchased separately, I received no break on AT&T's monthly service, so I didn't benefit from an handset subsidy, yet I pay the same amount. Go Figure.

On a related note, you might find Om Malik's analysis interesting. He writes:

The wireless unit of Verizon (VZ) reported year-over-year subscriber growth of 12 percent, but a mere 5 percent rise in voice revenues. Data revenue saved the day, surging 63 percent and lifting the company to 15 percent revenue growth overall. Data revenue per user increased 43 percent, while voice revenue per user declined 5 percent — pushing data to 20 percent of revenues from 14 percent. The same report revealed a 10 percent decline in residential access lines. The voice business of Verizon Wireless, in other words, seems to have entered the same cycle of contraction suffered by Verizon’s wireline business in recent years.

Joining the open access bandwagon promises to keep data revenues growing strongly, but CEO Lowell McAdam faces some mighty difficult choices as the 80:20 ratio of voice to data revenues reverses. The legacy pricing model incorporates price discrimination that will prove awkward to preserve.

Consider the lucrative SMS business of shipping 160 character messages for 10 cents each, or roughly $1,000 per megabyte. What happens when all devices cleanly incorporate instant messaging? “Any app, any device” means VoIP-capable devices that transparently support voice and web browsing via data plans. Why would someone pay Verizon an extra $40 per month for voice services? Any data plan that makes video affordable makes voice essentially free.

Articles like this one in BusinessWeek lend some credence to this argument. While AT&T suggest that its networks are already "open", they are only open to unlocked GSM devices. Amazon's Kindle is a different kind of device that uses a different business model. If the history of the telephone industry is a useful lesson, we will not be able to predict the kinds of innovation in devices and device/service combinations that might emerge.

Do you agree with Om Malik's assessment? Assuming you do, does Verizon's network opening make more sense?

05 December 2007

Asset swaps

It is not unusual for carriers to swap assets with each other for a variety of reasons. Sometimes swaps allow carriers to rebalance their regional portfolios to gain economies or to align their assets with their strategic objectives and other times it is to comply with regulatory requirements (there may be additional reasons as well). This news release (the quote below is from http://www.forbes.com/feeds/ap/2007/12/04/ap4402372.html?partner=alerts in Forbes) is an example of the latter:
AT&T Inc. said Tuesday it has reached an agreement with Verizon Wireless to swap wireless assets, satisfying regulatory requirements from the company's acquisition of rural cellular telephone carrier Dobson Communications Corp. In mid-November the Federal Communications Commission approved AT&T's $2.8 billion purchase of Dobson but required one of the companies to divest assets in Kentucky, Oklahoma and Texas. Under the terms of the agreement, after Verizon's acquisition of Rural Cellular Corp., AT&T will acquire some former Rural Cellular properties. The properties include licenses, network assets and subscribers in the Burlington, Vt., area and in rural service areas in New York, Vermont and Washington. AT&T will also acquire a cellular license from Verizon in Kentucky. In addition, Verizon will acquire from AT&T some former Dobson properties, including licenses, network assets and subscribers, in some rural service areas in Kentucky.
So, while these companies are fierce competitors in the retail market space, they are also business partners in other circumstances.

Sprint could spin off WiMAX division

This has been quite a month for the wireless industry. Between Verizon's recent announcements (see this and this) and the earlier news regarding Sprint and WiMax. In this article, Sprint interim CEO Saleh said:
... that Sprint was currently examining its plans for widespread WiMAX deployment in 2008 and deciding “if it’s the right course for us.” He then said that in the future, Sprint’s WiMAX division could take “multiple forms,” including one scenario where Sprint would “contribute our [WiMAX] asset to some kind of entity and find investors who are willing to fund the deployment of WiMAX.” Sprint would then buy services from that entity and resell them are on the market, he said.
I find the wholesale scenario most interesting (especially given my interest and work in secondary use and secondary markets for spectrum). This amounts to the creation of a wholesale WiMAX infrastructure from which Sprint would lease capacity for a potential retail operation. This is not unheard of in this industry. Note that Sprint is the carrier for MVNOs (Mobile Virtual Network Operator) like Virgin Mobile. Also, many of the carriers have sold or outsourced their towers to companies such as Crown Castle and American Tower. So, I am wondering a couple of things as a result of this announcement:
  • Is the Sprint retail brand going to separate from its carriage services in its 2G and 3G (i.e., CDMA) operations as well? If they do, then Sprint (retail) will be just another MVNO on the Sprint (carrier) network. If they don't I wonder what spinning off WiMAX really buys them. This is roughly akin to the functional/structural separation that is being discussed (and in some cases implemented) in the wireline industry.
  • Is the Verizon annoucement a signal that the industry is in the process of restructuring itself into separate handset, retail and wholesale services components? If this is the case, it suggests that the economic and strategic benefit of integrated package delivery is coming to an end; that the transaction costs that have driven this integration have decreased significantly so that a new industry organization can become feasible.
It is clear that it could take a few years for this scenario to play out ... we have yet to see what AT&T has in mind in this regard, as well as the carriers in other countries that have a highly sophisiticated mobile industry.