Wednesday, December 19, 2007

Sale Ahead for EarthLink?


EarthLink Executive Vice President Mike Lunsford, who acted as interim CEO after the death of former CEO Garry Betty, is leaving the company at the end of the year. Earthlink CTO Jon Kerner also is said to be leaving, as is Vice President of Production Operations Eric Alfaro. Kip Morgan, former EarthLink Vice President for Direct Marketing, Access and Audience, also has gone elsewhere.

When such things happen, one normally expects a sale of assets, which is what many observers expected when Rolla Huff took over at EarthLink.

T-Mobile, 3 Join 3G Networks


T-Mobile and 3 are pooling their U.K. 3G transmission networks, a move expected to reduce mobile tower sites by about 5,000 and save £2 billion in capital spending.

Kevin Russell, 3's UK chief executive, said the joint venture deal includes contingencies should either company be taken over, but both expect it to be a long relationship.

The move is not unprecedented, but still is unusual. Though not dictated by regulatory requirements, the move essentially creates a wholesale entity both retail networks will use to operate their businesses. It is not a structural separation, but certainly a functional separation.

By the end of 2009 the two companies plan to have 13,000 sites, covering 98 percent of the population with a mobile broadband network capable of speeds up to 7.2 Mbps.

Wireline Substitution, Mobile Plans, Broadband


KPN seems to have found a way to take market share in the German wireless market: give customers unlimited calling for a flat fee, avoid phone subsidies or selling phones, and keep things simple. The growing number of wireless-only customers apparently is helping, as one obviously needs more minutes in the plan to cover the additional volume when all calls in and out are taken on the mobile.

For $108 a month, Base subscribers can make unlimited free calls anywhere in Germany. A comparable offer by Vodafone costs $144. The sister E-Plus brand KPN supports also has shifted to this "no frills" approach.

In the third quarter 2007 subscriptions wereup 16 percent year-over-year, to some 14.1 million. E-Plus operating profit also rose 79 percent over that period, with profit margins of 38 percent.

Of course, KPN will have to figure out how to translate that success into similar good fortune in the mobile broadband segment, where it might not be quite so easy to maintain robust margins of this sort. Still, KPN's approach to the market is an example of what a carrier can do in an environment where phones are unlocked.

As Verizon moves to "unlock" its CDMA network, and as the C block 700-MHz spectrum goes into operation, also with an "unlocked" approach to device use, one wonders how soon somebody will try this in the U.S. market as well. Cricket Communications, one might argue, already has been chasing the wireline replacement market, but without the unlocked phone component.

Tuesday, December 18, 2007

FCC Reimposes Market Share Cap

The U.S. Federal Communications Commission has voted to impose a limit on the size cable operators can reach on a nationwide basis, limiting any single company from controlling more than 30 percent of total subscribers. The FCC in the past has maintained such a rule, but the limit was invalidated by a court decision in 2001.

Consumer groups say a strict limit on cable television system ownership is needed to prevent them from dominating television programming and Internet services and from blocking video competitors.

As a practical matter, the FCC action could affect merger deals Comcast Corp. would like to pull off, as Comcast already has about 27 percent. The rule might also affect smaller operators like Charter Communications and Cablevision , as it might rule out their acquisition by Comcast.

New Role for ENUM?

Analysts at the Yankee Group think there might be a new market developing for network elements or functions that provide the electronic numbering function that today is provided by peering federations. Such IP Routing Directories functionally provide the basic information required for interconnecting discrete VoIP networks.

Proponents of ENUM have argued that the business benefits derive from operational cost savings and service quality made possible by avoiding traditional PSTN routing infrastructure (SS7) to complete VoIP calls destined for a non-local VoIP
endpoint.

To date, business issues and volume issues have proven to be stumbling blocks. Large carriers make enough money from interconnection that any move to models that dispense with such payments are undesirable. That's the business issue.

And though native VoIP networks obviously require some sort of interconnection fabric, the fact remains that VoIP still is a small amount of total volume.

For the moment, legacy interconnection requirements remain essential. The percent of originating VoIP calls that are actually destined for an IP endpoint are sufficiently small that it might not make terrifically great sense to shift to an IPRD function of some sort.

Ideally, IP-to-IP connections are preferable. But the cost and quality issues might be a growing irritant rather than a compelling necessity for a large carrier, at the moment.

The pain of media conversion and database dips might not be creating enough pain to require an immediate shift to ENUM, in other words. Not to ignore the revenue implications for large carriers, either.

Whither WiMAX?


It might seem odd to question just how big the WiMAX infrastructure market might be, given strong support from the likes of Intel and others who see a data device optimized broadband network as a huge opportunity. And maybe WiMAX ultimately will create a large enough global base of infrastructure that handset and device manufacturers will have a large opportunity.

But potential end user volumes matter, and matter a lot, in today's world. The issue isn't whether WiMAX will work. CDMA works. But global volumes for GSM networks are so large that device innovation is higher on the GSM than the CDMA sides of the ledger. Volume also is a factor for software developers, who prefer larger markets to smaller markets.

Sprint got people excited with plans to build a $5-billion, nation-wide network in the U.S. but that strategy is now in question. Sure, there's the rest of the world, but if you have followed mobile technology trends for any time, you are more careful about the installed base, and the potential installed base.

In the third-quarter, Infonetics Research says, worldwide WiMax equipment sales climbed a mere six percent to $206-million. Meanwhile, worldwide unit shipments of fixed and mobile WiMAX equipment rose 16 percent in the third quarter of 2007.

Still, Infonetics is looking for Wi-Max to see annual growth of 87 percent between 2006 and 2010 as more carriers embrace the fourth-generation technology.

The number of worldwide WiMAX subscribers (fixed and mobile) is expected to skyrocket to close to 60 million in 2010, led by the Asian region, Infonetics says.

Still, there is the historic example of iDEN and CDMA to consider. Devices are more important than networks these days. And one has to contend with the issue of sheer mass, in that regard. There's no question that WiMAX will work. But that's not the crucial question. The issue is how large the market for WiMAX devices might be, compared to GSM and its derivatives.

What Next for Sprint Nextel?



Sprint Nextel has turned to a wireless industry veteran in naming current Embarq Corp. CEO and Chairman Dan Hesse new CEO and President. So what might we expect from him? Perhaps a focus on the many details of execution that seemed lacking in Sprint, of late. Hesse gets high marks for execution at Embarq.

Hesse also was considered a top candidate for the Qwest CEO post as well. And in some ways, Qwest and Sprint are in similar situations. Qwest does not have the financial ability to do some things one might expect from a former Baby Bell. Sprint likewise is in desperate need of serious attention to its core business, even as it contemplates a fourth-generation WiMAX rollout.

Neither company seems suited to a major acquisition that would fix the basic problems each faces. Qwest lacks scale to make some strategies work (it does not own a wireless network and arguably can't afford a major fiber-to-home video initiative).

Sprint remains the third-largest U.S. wireless carrier, but is feeling a rejuvenated T-Mobile nipping at its heels and has to do something really serious about its churn problems. Beyond that, Sprint is looking at some very basic decisions about future technology direction.

Volume in the global markets clearly is in GSM, and Verizon, the other major CDMA-based carrier, has made clear its decision to migrate to LTE, a GSM platform, for 4G. That leaves Sprint even more out on the fringe, as it now supports iDEN, which no other carrier uses, and CDMA which is losing traction in the U.S. market, if not yet internationally.

Before those issues can be tackled, Sprint has to stabilize itself. And Hesse is an adroit manager, most observers probably would say.

Before taking the helm at Sprint spinoff Embarq, he spent 23 years at AT&T, serving as President and CEO of AT&T Wireless Services from 1997-2000, then the nation’s largest wireless provider.

It is probably fair to say Hesse will have to right the ship before considering launching a new vessel.

Have LLMs Hit an Improvement Wall, or Not?

Some might argue it is way too early to worry about a slowdown in large language model performance improvement rates . But some already voic...