Sprint WiMAX "Functional Separation"?

With new reports out that Sprint might try to resurrect its failed alliance with Clearwire, this time perhaps with new minority investors (Intel, Google and Best Buy), one wonders whether Sprint ought to do what wireless operators are doing elsewhere and create a separate transmission company from which it can buy wholesale capacity.

Functionally separating retail operations and network functions might make sense in this case, given the other pressing demands for capital Sprint also faces. It isn't clear that Sprint derives significant competitive advantage from retaining ownership of the transmission facilities.

Best Buy might also be more inclined to invest in the new WiMAX network if such wholesale access were built into the investment agreements, as Best Buy might want to brand its own services.

Google is more interested in fostering an open networks environment and might not be that interested in any sort of Google-branded service. But wholesale access might be interesting if Google wants to experiment with applications that require such access, and wants to do so in a "real world" environment.

Ethernet Keeps Growing

Ethernet continues to gain a more prominent role in networking capabilities being deployed by service providers in North America, Europe, and Asia Pacific, with most carriers reporting 90 to 100 percent increases in Ethernet traffic for the past two years, according to analysts at Infonetics Research. IP and MPLS traffic has grown 70 to 80 percent over the same period, Infonetics says.

A new optical transport layer also will emerge, Infonetics believes. This new layer will be a fused Ethernet-WDM packet transport with circuit-like capabilities via Ethernet transport tunnels, also known as COE, or connection oriented Ethernet. That means more adoption of T-MPLS and PBT, Infonetics believes.

There was a time when some argued that "connectionless" protocols such as IP would replace "connection oriented" protocols such as time division multiplex, SONET and asynchronous transfer mode. As it turns out, there's a reason why connection-oriented protocols, or at least protocols that emulate connections, are important. Some traffic types, especially video and voice, are susceptible to impairments that can arise when connectionless protocols are used.

Value-Based Pricing

Communications services in the past have been priced based on usage (minutes or distance, for example). These days, there's a greater range of retail pricing, including flat fee for buckets of usage, for example.

But with the advent of IP-delivered video services, service and application providers have a chance to price services in a more natural "value to me" basis. If one looks simply at retail pricing and "bandwidth consumed," text messages "cost" the most based on bandwidth consumed, with voice second. Video and Internet services "cost" very little on a "bandwidth consumed" basis.

But that isn't the point. Text, voice and other communications applications are valued one way; video entertainment or simple Web access another way. In other words, in the communications space, "cost" is not the same thing as "price."

Target and Wal-Mart sell some products as loss leaders to get traffic into the store, so people buy lots of other products with varying profit margins. Communications services aren't any different. Some have have margins for providers, others have slim margins. The key, though, is value to the end user, not "bandwidth consumed."

Text messages "cost" very little, as the network to enable sending and receiving them is a sunk cost. But cost isn't price. Users demonstrate by their behavior that they value text messaging highly, on a "bandwidth consumed" basis. If one letter of a text message requires one byte, then sending or receving 6,553 messages consumes about a megabyte of data.

So a 160-character or smaller message might "cost" 10 to 15 cents. So a megabyte worth of text messaging represents as much as $655 to $983 for domestic usage. International messages obviously "cost" more.

A month's unlimited usage of entertainment video might cost $50, consuming more than a megabyte in a single second of use.

The point is that the retail price of any particular message or service has little to do with the actual "cost" of providing it, any more than the "price" of perfume, luxury automobiles, shoes or applications.

"Price" for a video service has to be vastly lower, on a "bandwidth consumed" basis, than texting, instant messaging or voice. That has little to do with user-perceived value, though.

Nokia Gets Cross Platform Support

Nokia has acquired Trolltech, a development platform for applications that can run on the Internet, accessed from a PC or a mobile phone. That might mean support for applications that run across operating systems, for example. The acquisition illustrates a couple of trends.

Mobiles need to acquire the ability to run Web applications, and need to do so in ways that are similar to the use of those apps on a PC, so users don't have to relearn a behavior. Cross-platform support also means Nokia can benefit from the huge numbers of developers working in the C, Flash, Java and other environments, for example.

Sloooow 700 MHz Auction Pace

The high bid on the national 700 MHz spectrum block stood at about $3.4 billion on Tuesday. Bidding is proceeding slowly and about all we know is that it will go higher until the minimum to compell an open network framework is reached. Google has pledged to bid at least that much and we have no reason to think it will not. Another billion and some odd to go until that point is reached, though.

Competitive Cable Developing Too Slowly?

One year after the passage of a law designed to ease the entry into the cable market of competitive providers in Michigan, only 110 of 2,000 communities in the state have a choice of cable providers, according to Multichannel News.

That is to be expected. Cable choice requires construction of brand new networks, not just the granting of a franchise. That takes immense amounts of capital and time, as well. A single new network in a single community can take three years or more to build, if there are no competing demands on construction and installation resources.

And there is history to consider. New video service providers have been attempting to so just this sort of thing for several decades, using a variety of methods, of which the most successful so far has been the use of direct broadcast satellite. There have been scattered regional efforts to duplicate cable networks, but overbuilders have not been notably successful, in large part because it is difficult to justify building a network that gets less than 30 percent penetration, which is what overbuilders largely have been able to attain.

Voice and cable modem services have helped the business case, but overbuilding remains a challenging financial proposition, and few expect widespread new competition in the terrestrial space from any other than incumbent local telephone companies.

The point is that nobody should be surprised nothing much has happened in just a year. New ubiquitous broadband networks take time to build, as well as lots of capital. If it were easy, lots of people would be doing it.

T-Mobile USA: Strong Quarter

Of the four largest U.S. mobile carriers, all but Sprint seem to be posting strong gains. T-Mobile USA says it had added 857,000 net new customers during its most-recent quarter. Average revenue per user also was up to $53 in the quarter, rom $52 in the third quarter of 2006.

Operating income was up 15.1 percent compared to the same quarter of 2006, while churn was down to 2.0 percent from 2.3 percent in the third quarter of 2006.

Contract customer net additions in the third quarter of 2007 made up 65 percent of customer growth, down from 80 percent in the second quarter of 2007 and 96 percent in the third quarter of 2006. Prepaid additions are the reason.

Contract customers represented 84 percent of T-Mobile USA's installed base.

Blended churn, including both contract and prepaid customers, was 2.9 percent in the third quarter of 2007, up from 2.7 percent in the second quarter of 2007 and down from three percent in the third quarter of 2006.

Blended ARPU was $53 in the third quarter of 2007, the same as the second quarter of 2007 and up from $52 in the third quarter of 2006.

Contract ARPU was $57 in the third quarter of 2007, the same as in the second quarter of 2007 and up from $56 in the third quarter of 2006, driven by increasing data services revenues.

Data services revenues were $666 million in the third quarter of 2007, representing 15.4 percent of blended ARPU, or $8.10 per customer, compared to 14.7 percent of blended ARPU, or $7.80 per customer in the second quarter of 2007, and 11.3 percent of blended ARPU, or $5.90 per customer in third quarter of 2006.

Text messaging still is the most significant driver increasing data ARPU. The total number of short message service and multimedia messaging service messages increased to almost 21 billion in the third quarter of 2007, compared to 18 billion in the second quarter of 2007 and 10 billion in the third quarter of 2006.