That isn’t to say there has been a dearth of strong competitors. The U.S. cellular industry once featured large independent wireless carriers. Only Sprint remains. The U.S. dial-up Internet access business once was lead by independent providers. Today the largest providers are at&t and Comcast.

At one point, AT&T launched vigorous assaults on the local exchange market, leading the whole U.S. competitive local exchange carrier business in the process. Those assaults failed to dislodge the leading local exchange carriers.

The independent CLECs fared no better. Then Vonage and a few other independent VoIP providers set out to prove there was a market for mass market VoIP. They succeeded, only to watch cable companies assume industry leadership.

And despite all the challenges, the former Bell Operating Companies are creating new revenue streams to replace their legacy services. Broadband services appear to be responsible for an uptick in Verizon Communications average revenue per user, says equity analyst William Trent.

At&t also can point to broadband services, both video and access, as contributors to their performance, but Qwest, which really isn’t pushing video, also is seeing some firming of line losses and strengthening of sales in the small and mid-sized business segment, for example.

In Qwest’s case, though broadband is getting serious marketing effort, abated competition probably accounts for a goodly portion of the revenue support. And smaller operators such as CenturyTel also say they are seeing heightened marketing effort from cable, but little revenue impact as of yet. By implication, CLECs are not much of a factor.

Broadband is key for Verizon. Increases in broadband and video revenue and more bundled customers helped drive Verizon consumer retail average revenue per user up $5.64 or nearly 11 percent year over year.

The 3.3 percent sequential growth in Verizon ARPU in the most recent quarter also is primarily broadband and video-related. In Texas the retail ARPU growth was in excess of 20 percent. New York, New Jersey and Virginia experienced double-digit growth.

Most of at&t’s line losses in the business space are caused by technology migration, at&t maintains. Only about 30 percent of its overall access line disconnects are caused by competitive threats, the company maintains. Of course, the cable company impact has yet to be fully felt. So far, only four to five percent of the churn is caused specifically by cable companies, at&t says.

That pressure is expected to grow especially at the lower end of the small business space, in the under-10-lines segment, and probably most of that will be concentrated in the four-lines-or-less segment.

If history proves an accurate guide, the telco incumbents will survive the cable challenge as well, as hard as the challengers will work. IP

In 2007, mobile will cement its position as the normal way to make phone calls, leading to a “sharp decline” in landline calling, says Deloitte & Touche USA. That challenging prediction is arguably the most urgent of all telecom trends the consulting giant expects will shape the new year.

Moreover, viewers 35 and older watch more streaming video than users between 18 and 34, the study finds. About 31 percent of 18 to 34 year olds watch streaming video, while 69 percent of consumers ages 35 and older view streaming video online.

Approximately 83 percent of surveyed respondents indicated that their online video usage in 2007 has either stayed the same or increased since 2006. About 36 percent of consumers have increased their consumption of online video, with an even breakdown between men (36 percent) and women (37 percent) viewers.

The majority of consumers are streaming online video at home rather than at work or school, with 45 percent of streaming activity taking place in the evening. The survey found that 95 percent stream at home; four percent at work; one percent at school or university.

The findings for “school” or “university” usage probably are based on streaming from personal computers in dorm rooms, rather than institution-owned machines.

More than 62 percent of consumers said they are most likely to stream news clips, with movie trailers and music videos next in popularity. Compared to the second half of 2006, consumers are streaming fewer music videos and streaming more news clips, user-generated videos and sports clips. However, these consumption behaviors vary dramatically by age.

As you would expect, viewers in the 18 to 34 age range watch more user-generated fare, fewer news clips, more music videos and more TV shows.

Streaming Selections (% of respondents)
	All		18-34 year olds
	1st Half ‘07	2nd Half ‘06	1st Half ‘07	2nd Half ‘06
News clips	62%	49%	44%	34%
Movie trailers	38	33	40	35
Music videos	36	47	54	65
TV shows	33	26	51	33
User generated videos	29	21	42	26
Movies	25	20	32	19
Sports clips	21	11	14	10
Other	8	9
Source: Advertising.com, September 2007

Social elements of viewing also are increasing. Overall, 42 percent of consumers have forwarded a video clip to a friend. Consumers who view content more than once a week also forward more clips, with 55 percent forwarding clips vs. 34 percent and 20 percent for those who view content once a week and once a month. Women (47 percent) forward more clips than men (36 percent).

The study findings also suggest that online video usage has not yet begun to displace linear TV viewing. Or at least that is what people think.

Some 80 percent of respondents say that online video usage does not cut into their TV time. But there might be gender differences. Some 29 percent of men say online video usage cuts into TV viewing, while just 16 percent of women say online video usage cuts into their TV viewing.

These results also vary by intensity of online video usage, however. About 12 percent of those who view content once a month claim that video usage cuts into their TV time.

Consistent with most such research, about 94 percent of respondents indicate that they would prefer to view ads than pay a fee to watch video content online. Of course, actual behavior varies from that expressed opinion. People generally say they don’t like ads, but that never actually seems to dissuade them from watching video when advertising means the viewing is at no incremental cost, and when the alternative is the payment of an extra fee.

Not surprisingly, 63 percent of respondents would prefer online video ads that are shorter than TV ads. And though there is widespread concern in the content and advertising communities about users zipping right past ads when possible, 65 percent of respondents say they watch online video ads through to completion, including the 72 percent of consumers who view streaming content more than once a week.

Lighter viewers say they are less likely to watch ads through to the end. Of those who view content only once a month, 49 percent say they view advertising through to completion.

And there is some evidence that shorter spots get more viewing. In the study, consumers suggested they were eight percent more likely to view 15-second spots to completion than 30-second spots.

On the other hand, 30-second pre-roll formats seemed to slightly outperform the five-second and 15-second ads when measured in terms of click-through rate.

The study suggests that older consumers use streaming video to get information, as shown by their preference for online news clips. Younger consumers tend to stream content for entertainment purposes.

Also, as you would expect, consumers between the ages of 18 and 34 are using more streaming content on a habitual basis. Sixty-nine percent of their online video streaming occurs more than once a week, while 47 percent of those ages 35 and older view streaming video multiple times a
week. IP

Even so, in its financial results for the year ended March 31, the India-based operator reports that its wholesale voice revenues were up from 21.62 billion rupees (US$528.99 million) to INR22.16 billion ($542.21 million), but operating profit fell from INR4.6 billion ($112.55 million) to INR4.23 billion ($103.5 million). Formerly India's monopoly provider of international long-distance services, VSNL has positioned itself as a major player in the global connectivity market by picking up the beleaguered Tyco Global Network and Teleglobe for bargain basement prices.

VSNL is focused on increasing its share of global voice traffic to push down its operating costs. Since acquiring Teleglobe in late 2005, the carrier has grown from handling 13 billion voice minutes a year to 20 billion. But with wholesale voice prices continuing to fall, the operator is working to diversify its services into enterprise and managed IT services such as TV uplinking, transponder leasing services, retail Internet access and satellite services.

It’s among several operators in Asia to sign a bilateral Wi-Fi roaming agreement with, and it has a contract with Skype Ltd. to terminate calls from the SkypeOut VoIP-to-PSTN (public switched telephone network) service.

In April it teamed with Microsoft Corp. to launch the white-label ReadyAccess business messaging service for mobile operators, for which it's signed up Thai mobile operator Advanced Info Service as a customer. IP

The first issue is that G.711 is bandwidth intensive; it uses 64 kbps per channel, whereas other codecs such as G.729a use only 8 kbps. The large amount of bandwidth for G.711 has limited the number of channels that can be transported through asymmetric broadband access links, which are sometimes limited to 128 to 256 kbps in the upstream.

The second issue is that G.711’s voice quality, when transported on a network with packet loss and jitter, isn’t very good. Even low amounts of packet loss or jitter can cause substantial degradation in voice quality from one side of a call to the other.

To overcome these concerns, carriers looked to the G.729a codec. However, while G.729a solved the bandwidth concern, it delivered a lower level of voice quality than G.711, which was compounded even further when affected by packet loss and jitter. G.729a also was plagued by an encumbering and expensive licensing arrangement that could cost vendors and service providers upwards of $.60 per VoIP channel.

The advent of a new codec type, called the Internet low bitrate codec (iLBC), offers great promise for utilizing less bandwidth (13.3 kbps), while providing a very good level of quality, including built-in protection against quality degradation when packet loss and jitter are present. In addition, iLBC is a “royalty-free” codec, so it lowers the economic impact for large-scale VoIP deployments.

While iLBC’s benefits are attracting attention, the codec does have its drawbacks. The first issue is the amount of media processing resource that is required for an iLBC channel. While G.711 requires 0.2 MIPS (million instructions per second) of processing power per channel, for example, iLBC requires 12.5 MIPS per channel, or nearly 60 times as much. This increase in MIPS often has substantial impact on media gateway and/or media server channel densities, forcing carriers to invest in additional media gateways and media servers to accommodate the reduction in supported channels.

The second issue is that iLBC is not universally supported by all media gateways and media servers in the carrier’s network.

But instead of adding or upgrading media gateways to support the new codec, providers now can easily add support for iLBC by deploying a purpose-built media processing platform into their networks. These products deliver far higher media processing performance and can offload transcoding functions from media gateways and media servers. With a media processing platform in place, existing media gateways can continue to perform the G.711 packet-to-circuit conversion while the new media processor handles the complex transcoding functions that require additional MIPS.

The basic architecture of a media processor includes a bank of digital signal processors (DSPs) to form a large MIPS processing pool, using architectural economy, density and packaging to deliver far more MIPS per cubic inch than in any other device in the VoIP network. This new element allows carriers to have a cost-effective MIPS resource at their disposal to improve voice quality, support diverse codec transcoding, and increase service flexibility. Even with MIPS-intensive codecs like iLBC, a media processor is more than capable of supporting tens of thousands of calls per system in a small footprint.

A media processor is often deployed between the customer’s premises device and the media gateways in the carrier’s central office. This is often called the “access border” location, similar to where a session border controller may also reside. Alternatively, a media processor may be deployed at the carrier-to-carrier demarcation point or “peering border” to handle peer traffic from other carriers.

Using a media processor to provide transcoding for iLBC and other complex codecs not only provides additional flexibility but also makes good economic sense, with a typical cost per channel of $50 to $70 for transcoding functions. This may seem to be expensive at first glance, but when you consider that the media processor need only be involved in the fraction of iLBC-encoded calls that require extra processing power, the economic benefit is significant.

For example, in a network where 10 percent of calls require iLBC transcoding, the carrier would purchase one media processing port for every 10 media gateway ports that are needed to support the new iLBC service. As a result, using a media processor costs one $50 port (on the media processing platform) compared with 10 $20 ports (for upgrading media gateways). The oversubscription benefits of the media processor do not even include the savings from operational expenditures and complexities realized, because it eliminates the need for the carrier to upgrade all media gateways, media servers and other backbone equipment.

By adding the iLBC codec with media processing functionality, carriers not only gain the numerous benefits of iLBC, but they also can overcome the challenges of deploying iLBC within their networks. IP

Matt McGinnis is director of product marketing for Ditech Networks.

What Ethernet hasn’t been—at least so far—is a primary driver of retail or wholesale revenues for most providers, though it has high importance for some specialized providers.

But that seems finally to be changing. Worldwide revenue for business Ethernet services is projected to climb to $30.7 billion by 2012, according to Vertical Systems Group. Business Ethernet services include dedicated Internet access, Ethernet transport for IP or MPLS virtual private networks or frame relay, Ethernet private lines and Ethernet local area network or virtual private LAN services.

Still, in a global business whose annual recurring revenues approach $1 trillion, U.S. Ethernet services revenue only contributed between $1.5 billion and $1.7 billion in 2006, according to analysts at In-Stat.

That explains the relative sluggishness with which tier one providers have been addressing the market. Up to this point, there just hasn’t been enough revenue in that one segment to convince tier one providers that the return is worth the investment. Predictably that has created market space for specialized providers.

But as is always the case, once a market begins to ripen to the point that a single provider reasonably can expect to capture about $1 billion a year in incremental revenue, the competitive dynamics shift as tier one providers begin to assert themselves. And we are nearing the point where that should be anticipated.

Still, there appears to be some significant degree of buyer confusion at the moment. One always hears that Ethernet is the more affordable way to source business data capacity.

Surprisingly then, nearly half (46 percent) of enterprise network decision makers don’t know how pricing for business Ethernet compares to their existing network services, according to recent survey results from Vertical Systems Group.

Another 34 percent of customers say that Ethernet is in fact “not lower priced.” About 20 percent say business-class Ethernet is a less costly service alternative than services such as frame relay, private lines, or dedicated IP VPNs, which is the message Ethernet proponents have been pounding on for quite some time.

Of course, one reason many decision-makers aren’t completely able to assess Ethernet value and pricing is that the actual services they buy, even when based on Ethernet, are not simple “capacity” products.

Even the simplest of Ethernet “private line” decisions requires some modeling to assess how the additional capabilities will match up with the costs. Given a big enough network with enough discrete applications, protocols and locations, it can be time consuming to evaluate how a proposed Ethernet solution matches up to what already is in place.

The capacity and price gap between T1 and 10 Mbps Ethernet is sizable.

Legacy network services including frame relay, private lines, ATM, IP circuits, and PBX trunk lines currently reach 1.4 million U.S. business sites over T1 access links. An additional 1.6 million sites use 56 kbps or fractional T1 connections.

Ethernet classically ranges from 10 Mbps up to gigabit speeds. The disparity explains why mid-band Ethernet alternatives are touted, but also indicates why it can be challenging to compare a legacy network with a new Ethernet alternative.

Even when that is done, there is the further task of evaluating similar implementations offered by different providers, each of which might be sufficiently different to require another level of analysis.

That done, there is another issue: service pricing is custom, so it’s nearly impossible to figure how an Ethernet solution is going to measure up without getting a number of actual detailed quotes. When considering the costs for different T1, digital subscriber line or cable modem options across lots of branch locations, planners at least can ballpark costs without lots of effort, as pricing is standardized. Not so for carrier Ethernet solutions.

Then again, service providers won’t be going out of their way to pitch Ethernet alternatives where they now provide legacy services with higher margins and gross revenue contributions. Buyers are more confused than they someday will be simply because a clear enumeration of value and price is not being offered readily.

Also, with some exceptions, business Ethernet providers are being deliberate about the “bits per dollar” metric. In no hurry to commoditize pricing more than they already assume will be the case, even many Ethernet proponents arguably are being less aggressive on the pricing front that might be possible.

Nor, given the relative spottiness of access fiber coverage, are many suppliers touting Ethernet access services as aggressively as they could and would if fiber access were simply ubiquitous.

In that regard, the situation is akin to that faced by cable and telco broadband providers. Marketing tactics must be controlled whenever service availability in a market is constrained by construction timetables, coverage limitations or other network impediments that gate availability.

Business fiber access to network services reaches 13.4 percent of U.S. buildings with 20 or more employees, for example, Vertical Systems Group says. Under those conditions, it is not helpful to stimulate demand for services most customers cannot buy, even when they
want it.

Given all those issues, “it’s understandable why customers aren’t sure if business Ethernet is a bargain or a budget-buster,” says Rosemary Cochran, Principal at Vertical Systems Group.

One might superficially think that with the explosion of Internet Protocol traffic, Ethernet must already have claimed a much-more-substantial share of the transport business. But IP is a layer three protocol that simply rides on top of layer two. And most of the layer two transmission and access technologies simply are something else, in other words.

Then there’s the layer one issue: Ethernet is most compatible and most affordable when fiber layer one facilities are available. And that remains a doggedly tough proposition.

That is not to underestimate Ethernet-over-copper options. But there are lots of deployment scenarios where legacy protocols must be supported for technical, business or cultural reasons, rendering Ethernet over any medium a less-desirable choice.

Mobile carrier cell site backhaul is one obvious example.

Still, the core driver for native Ethernet remains enterprise data traffic, as important as consumer applications are becoming. The simple reason is that organizations rapidly are exceeding the capacity limits of their legacy frame relay, ATM, and private line networks, and as they do, they are increasingly replacing these legacy services with Ethernet and IP MPLS VPN services, which are more flexible and less restricted than their legacy counterparts, say researchers at Infonetics Research.

“After more than doubling in 2005 to $5.7 billion, worldwide Ethernet service revenue jumped another 65 percent in 2006 to reach $9.4 billion, according to Infonetics Research.

The Ethernet service market is forecast by Infonetics to surge 171 percent from 2006 to 2010, when it will top $25 billion worldwide.

Service revenue for mid-band Ethernet, including the 1 Mbps to 10 Mbps category, will nearly triple from 2006 to 2010, Infonetics forecasts. Oddly enough, as Ethernet pushes into the gigabit range, there is a new “mid-band,” the space between 10 Mbps and 50 Mbps.

If U.S. service providers aggressively pursue service expansion, revenues will reach $6.1 billion in 2010, In-Stat estimates.

But there are more aggressive forecasts. Researchers at Infonetics think global Ethernet revenues could reach $20 billion or so by 2009.

It looks like there might finally be some fire where all this smoke has been. FAT

Enter Critical Path, a messaging software and services company that has developed a solution for broadband Internet providers to consolidate most of those windows. Supported by Critical Path’s Memova Messaging platform, the Contact-Centric Communications solution gives Internet subscribers unified access to email, instant messaging, multimedia content sharing and other communications through a central console.

That means customers can go to one place on their computer to carry out such tasks as send and receive emails, IM buddies on the Google and Yahoo! networks and deliver and receive an SMS to and from any cell phone. Users can discover which contacts are available and are notified when new messages and shared content arrives.

Critical Path also is working to establish relationships with AOL and MSN to broaden the potential reach of instant messaging contacts on the console, says Bruce Warren, senior director of product marketing with Critical Path. The Memova platform also features VoIP capabilities: customers can click on a link to make a computer-based call through Critical Path’s relationship with CounterPath Solutions, Inc., a developer of session initiation protocol phones. Warren says Critical Path also could integrate its solution with an Internet operator delivering its own VoIP service.

The solution comes at a time when broadband operators “have been struggling to build customer loyalty in a market that is increasingly facing price pressure and commoditization,” says Teney Takahashi, a senior market analyst with The Radicati Group.

“By integrating a wide range of features, like VoIP, IM, email and SMS into a single interface …, [Critical Path] can help carriers to differentiate themselves from others and build brand loyalty,” says Takahashi.

Storing and controlling the contact information of subscribers’ friends and family also has its upsides, says Jennifer Simpson, emerging consumer applications analyst for Yankee Group.

“Incumbent carriers face the challenge of consumers using just the Web or devices as a repository for contact information,” she says. “Increasingly, services that connect individuals to one another through the address book or contact list reduce churn by capturing not only a single consumer but his or her friends and family as well.”

All the while, “The future direction of messaging will be a consolidation of messaging functions into a single tool that will allow individuals to gain more control over how they receive messages,” says Mike Osterman, founder and president of Osterman Research, a market research firm in the messaging and collaboration space.

With a single address and access to a directory, individuals could control how they would receive communications, he says. For example, a person at his desktop could specify that non-urgent communications would come in as an email while an urgent message would be conveyed through an IM.

San Francisco-based Critical Path is optimistic that the operators around the world using its email platform and messaging components will adopt its latest solution, which took nine months to introduce following its conception and also features such capabilities as anti-spam, anti-phishing and virus protection.

“I don’t really see this as being a paid service,” Warren says. “I think this service would be targeted at reasonably sophisticated online Internet users.”

Reaching a state of absolute convergence takes time, but companies like Critical Path have certainly begun to embark on the journey. FAT

The problem, quite frankly, is that businesses with more than 100 employees—some even would say with more than 50 employees—operate very differently from the typical "small business."

"The 100-person business has an organizational chart and departmental breakdowns," says David Cork, Natural Convergence CEO. "They truly are little versions of larger enterprises.

"People that build things for the 100-person business could easily sell those capabilities to companies with 1,000 or 10,000 people," he continues.

Keep in mind, as much as 97 percent of firms fall into that 2-to-99-employees range, according to Yankee Group, so even within just "small business," dramatic variation in needs and behavior require much deeper sub-segmentation. Even the common break-out of "very small businesses," say between 2 and 20 employees, also can be inadequate, as nearly 90 percent of all businesses fall within that category.

Total Installed U.S. Base, in Millions of Lines
Year	Traditional PBX	IP/Converged	KTS	Centrex	Total
2000	48.4	0.5	39.6	17	105.5
2001	49.3	2	39.3	17	107.6
2002	50.1	4.1	38.9	16.5	109.6
2003	49.9	7.7	38.4	15.8	111.8
2004	49.1	13.2	37.8	14.9	115
2005	47	20.8	35.7	14.3	117.8
2006	44.6	29.6	33	13.7	120.9
2007	40.9	39.4	29.2	13.3	122.8
2008	36.5	49.9	24.9	12.8	124.1
2009	32.3	60.8	20.9	12.3	126.3
2010	28.3	72	17.4	11.7	129.4
Source: TIA, InfoTech

And make no mistake, this segment of the market holds huge potential moving forward. Researchers at ATLANTIC-ACM, for example, have pegged firms with 11 to 100 employees as the "fastest growing demand segment for VoIP services," representing a compound annual growth rate of 29.7 percent from 2006 to 2012.

And if we can assume that early adopters, heavy users and small businesses with clear needs for advanced applications already have made the switch to IP telephony, it’s also probably safe to assume that sales cycles are going to grow longer and more arduous as sales staffers attempt to convert the more general user population. Therefore, deeper segmentation that can help tailor more specific messaging and quickly qualify or disqualify leads will be increasingly valuable to customer-facing sales representatives.

As Susan Hobart, vice president and senior program director for InfoTrack, warns, a key challenge to penetrating the vast small business segment is "keeping the cost per sale down by shortening the cycle and minimizing premise visits."

One tack that appears to be taking off among wholesale providers is the framing of marketing efforts around the actual voice platform that is being replaced rather than around the size of a company.

Indeed, surveys of Canadian small businesses performed by Yankee Group indicate a direct correlation between a service currently deployed and a preference between hosted and premises-based VoIP. Among respondents that use basic telephone sets or multi-line phones as their primary type of phone system at a headquarters location, 76 percent would prefer a hosted VoIP solution, say Yankee figures. Conversely, only 35 percent of businesses that use some type of PBX (private branch exchange) would prefer hosted.

In the case of access technology, 63 percent of digital subscriber line users and 72 percent of cable modem users also would prefer hosted VoIP, while 70 percent of businesses with a T-1 connection would prefer a premises-based offering.

Of course, one could argue that the size of a company often will play a large part as to the type of legacy service in use. While it’s true the two often won’t be separated, the point is that messaging and product packaging needs to be framed around existing services rather than simply number of seats.

A prime early example of this tactic would be BroadSoft, which has found success marketing legacy Centrex customers with an IP-based Centrex replacement-type service.

Elsewhere, Natural Convergence is marketing its platform as a type of legacy key system replacement, says Cork, offering small businesses a hosted IP service that looks and acts in familiar ways to those currently using key systems, while offering more for less.

"It feels like a simple migration," he says, "because it’s the stuff the business already is used to."

Growth in the Total U.S. Installed Base
Year	Traditional PBX	IP/Converged	KTS	Centrex	Total
2001	1.9%	300%	-0.8%	0%	2%
2002	1.6%	105%	-1%	-2.9%	1.9%
2003	-0.4%	87.8%	-1.3%	-4.2%	2%
2004	-1.6%	71.4%	-1.6%	-5.7%	2.9%
2005	-4.3%	57.6%	-5.6%	-4%	2.4%
2006	-5.1%	42.3%	-7.6%	-4.2%	2.6%
2007	-8.3%	33.1%	-11.5%	-2.9%	1.6%
2008	-10.8%	26.6%	-14.7%	-3.8%	1.1%
2009	-11.5%	21.8%	-16.1%	-3.9%	1.8%
2010	-12.4%	18.4%	-16.7%	-4.9%	2.5%
CAGR ’07 to ’10	-10.7%	24.9%	-14.8%	-3.9%	1.7%
Source: TIA, InfoTech

In 2006, about 33 million small business and branch office locations employed a key system for their voice needs. By 2010, the Telecommunications Industry Association and InfoTech estimate that close to half will replace their system, presumably with some flavor of IP telephony, at least for the vast majority. Among the legacy services of traditional PBX, Centrex and key systems, the shift to IP is expected to move most rapidly among key system users, says TIA.

By comparison, Cork estimates that legacy PBX deployments within businesses of 100 employees or less number around 8 million, amounting to about a quarter of the size of the small business key system market. And since Natural Convergence offers a hosted solution, "We were looking for people that had no IT guys on staff," says Cork, "and key systems were a really good way to find that."

Cork believes about 70 percent of small businesses would be interested in a hosted voice solution that is offered as part of some type of service bundle.

The vast majority of key systems in place, says Cork, serve 48 desktops or less, leading the company to narrow its target to companies with 5 to 50 employees, with most prospects likely falling between 20- and 50-employee sites.

"We don’t see a lot of key systems below the 20 mark," says Frank Paterno, vice president of marketing for Intelliverse.

When segmenting small businesses by their existing services or infrastructure, it becomes apparent that the prevalent IP PBX pitch—whether it’s premises-based, hosted or managed—offering "enterprise-style features" at small business prices simply lacks relevance for the vast majority of small business decision-makers, particularly at a time when small business owners are looking more to bring the simplicity and convenience of consumer-oriented services into the office rather than retro-fitting scaled-down versions of enterprise applications.

That’s not to say many of the IP PBX offerings are unable to properly and affordably serve all sizes of small businesses. It’s just that messaging that centers on "PBX" or "enterprise features" will come across as overkill to the vast majority of firms with 2 to 50 employees, especially when considering that a good percentage of small businesses are looking purely for voice-only solutions, say researchers at AMI-Partners.

"Initially, we came out calling our service ‘hosted PBX,’ but really we are displacing Centrex or POTs (plain old telephone service) business lines," says Paterno. "That is our competition."

The typical business POTs user, which tends to be at the lowest end of the small business segment, may not be familiar even with basic Centex features, say Paterno, so terms such as "hunt groups" may have little impact. "You have to be careful about the wording," he says.

For this type of prospect, Intelliverse recommends feet on the street stick with case scenarios, such as asking: "Do you have home-based workers? Do you have a sales staff? How do you keep in touch with them, and how do customers reach them?"

Find me-follow me, properly described, tends to resonate well at this low end of the market, says Paterno.

"The real value from us comes when the guy using a POTS line can be introduced to concepts like four-digit dialing and call transferring, which he never considered before but now can have and have it be cheaper than his POTS line," he says.

When approaching a customer who understands Centrex, typically a location with 5 to 15 or 20 seats, "our sales job gets easier," says Paterno. Here, sales pitches can include discussions of hunt groups, auto-attendants and maybe transferring calls to a certain extent.

As far as devices, Paterno believes distributors should lead with ATAs (analog terminal adapter) when approaching both business POTS and Centrex users, which likely are using analog phones.

"Small business employees tend to hate change, and managers know this," he says. "They don’t want any business disruption. All they want their employees to know is that they have to reset their voicemail when they come in the next morning."

Meanwhile, key system users (say firms with 20 to 50 seats) will be interested in the more collective, group-based features, such as transferring calls, extension dialing and conferencing bridges, says Paterno. And now that digital terminal adapters have become available, a decision has to be made whether or not to scrap existing phones.

"We cannot find a compelling reason why someone would want to stick with digital phones," says Paterno. "If cost is an issue, we’ll go buy you $20 phones."

Beyond 50 seats, independent providers and their distributors run the risk of bumping heads with some pretty big brands. "Cisco is going after that 50 and up market," says Paterno. "And some other big guys even have small blades that will do pretty well for that market."

Most small businesses already are confused about the multiplicity of IP telephony solutions available and are unable to evaluate and decide whether a premise-based, managed or hosted solution, or maybe broadband VoIP or peer-to-peer, is the best fit for them, argues Sanjeev Aggarwal, AMI-Partners’ vice president. It’s likely wide-net marketing efforts that attempt to include the entire SMB segment only are adding to the confusion.

"The complexity involved in understanding and evaluating the different solutions and vendors, coupled with a lack of voice-savvy IT resources, is delaying what is an otherwise rapid-paced migration of SMBs to VoIP," says Aggarwal.

Sub-segmentation based solely on the number of employees or desktops is only the start. Because size does matter, but one size rarely fits all. IP

His first speed tests after the install showed 35 Mbps to 43.5 Mbps downloads and 28 Mbps to 33 Mbps uploads. "Infriggincredible," he says. Suffice it to say that this is the first time Dipert ever has had to take steps to ensure that his local network wasn't some sort of inadvertent bottleneck (he says he's running category 5 wiring from his PC to the router.)

As you might expect, "For everyday Web surfing, the incremental speed over my 1.2 Mbps (downstream) and 300 kbps (upstream) DSL connection actually isn't all that noticeable," he notes.

“Access lines” used to be a highly reliable measure of telco revenue and growth. The measure began to lose its predictive power when “alternate access” companies began selling high-capacity connections to business customers. The metric became even less relevant as competitive local exchange carriers began selling a mix of services based on NxT1 access loops.

The mashup also illustrates how easy it is for new communications-centric applications to be created by third party developers, hosted commercially by application providers and provided directly to end user enterprises and consumers without necessary involvement by network service providers.

“I wrote the application in a matter of days, and I didn’t need anything but my credit card to charge the $50.00 for the SMS messages. I didn’t need any contracts or hardware,” says Thomas Howe, owner of Thomas Howe Consulting.

It also, therefore, suggests why it is so important for service providers with platforms based on IP Multimedia Subsystem to spend significantly more time thinking about application development. Developers might not need IMS to create tomorrow’s communication apps. And if the intention is to avoid becoming a “dumb pipe,” Web 2.0 will do that, barring significant movement by “carriers” to integrate IMS networks with Web apps.

Basically, the application transcribes a doctor’s office inbound patient voice mails into text. The messages then are sent as text messages to a doctor, after a bank of nurses has performed “triage” on the calls, sorting the cases into “urgent” and “can wait until tomorrow” baskets.

Say a patient doesn’t feel well and calls his doctor to schedule an appointment for the next day. An IVR answers, asking if the call is a routine business matter or a health issue. If the caller reports it is a health issue, the IVR asks if the caller is on a mobile phone. The IVR then takes a voice message and hangs up.

As soon as the call is over, the patient gets a text message on his cell phone telling him that the call was received, and that the call is being forwarded to nurses. If the caller is calling from a landline phone, the patient gets a return call.

The message is sent to nurses who determine whether the case is urgent or not. If they think it is routine, they indicate that on their console, which results in another message being sent to the cell phone telling the patient that the matter is probably routine, and they would get a call in the morning.

If urgent, the patient would get a message like “A nurse thinks you need to speak with a doctor. We are looking for one now – stay near the phone,” says Howe.

“Urgent issues are forwarded to the doctor’s mobile, with a summary of the call prepared by the nurse. “The doctor’s message would be ‘Mr. Kraus – 40 WM - left side weakness, nasuea -508 364 9972.’”

The doctor would be able to press the send button on his phone to call the patient.

Howe wanted to “demonstrate that integration of real time communications into the business process saves money for the business, while increasing quality of service for the customer.”

He also wanted “to explore the use of artificial intelligence, by way of Amazon Turks, and its possible role in telephony applications.” The basic application was created in about 40 hours, except for the outbound IVR portion, which Howe says he isn’t enough of a programmer to pull off.

There are some issues here for conventional service providers. It didn’t cost Howe much to create the app. As a commercial venture there would be some incremental spending, based on usage, paid to a hosted IVR service. But Howe says the big points are creation of the app quickly, with no large capital investment and no involvement by a network service provider. “Traditional carriers will get their clocks cleaned” by apps such as this, he says.

“According to the National Center for Health Statistics, 55 percent of the 90 million visits to hospital emergency departments in the United States in 1996 were unnecessary,” says Howe.

“According to Blue Cross, Blue Shield, an average emergency room visit costs $1,049.00, whereas the average doctor’s office visit costs $149.00,” Howe notes. So “we have a $44 billion dollar problem with unnecessary emergency room visits.”

Getting the application to work in the real world would require creating a bank of on-call nurses. Howe says that’s where the Amazon Turk system comes in. Turk could aggregate tens of thousands of stay-at-home nurses raising small children.

The incentive is the ability to make several dollars in two minutes by listening to a voice message and prioritizing it. Insurance providers might provide the business model, paying the system to keep patients out of emergency rooms. IP

In fact, of the $407 billion North American service providers now earn from recurring services, mobile represents $164 billion, or about 40 percent. Wireline access revenues now stand at just 24 percent of overall revenues.

And mobile revenue is growing at 7.7 percent on a cumulative annual growth rate, say analysts at the Yankee Group. By 2011, wireless will account for 43 percent of North American service provider revenue, while landline voice will be 21.5 percent of total revenue.

By 2011, in other words, North American service providers will be making nearly 80 of their revenue from wireless, video, enterprise services and broadband.

SNL Kagan predicts wireless growth at about a three percent annual rate over the next 10 years.

SNL Kagan expects total industry average revenue per user to grow at a compound annual growth rate of 1.5 percent over the next 10 years, from $52.38 today to $61.09 by 2017.

Mobile data ARPU will be a significant growth driver. Mobile data revenues expanded at a 45 percent rate from the first quarter of 2006 to the first quarter of 2007. Of course, the amounts still are small compared to voice revenues, but so are wireline video and broadband revenues compared to voice. Mobile data ARPU grew from $5.92 to $8.58 over the last year, says SNL Kagan.

SNL Kagan estimates data revenue will grow at a compound annual 14 percent rate over the next 10 years, rising to at least 22 percent of service revenue, compared to under 10 percent today.

Mobile providers also seem to be attacking their churn problems smartly, with leading providers now experiencing churn of much less than two percent a month. Bundles may be helping, in that regard, say analysts at the Yankee Group.

A bit over a third of churning mobile customers claim price was the reason. Bundles should reduce that sort of churn, to the extent that there is a perceived increase in value and relative decrease in “price” irritation. Also, bundles will attack churn simply because the complications of losing three or four services at once are a deterrent to change.

Still, reliability, voice quality and signal coverage account for more than 65 percent of the reported churn.

There still is no significant amount of churn caused by “handset envy,” feature or application uniqueness, it appears. Data on iPhone adoption should start to give us some idea of how big a factor that might be. Should a rumored Google Phone actually come to market next year, and should it be available from a single provider, we might get more churn based on handset availability.

So far though, wireless does not register as a component of consumer bundles. Compared with 2006, multi-service subscriptions provided by a single provider increased by 23 percent to a penetration of 79 percent of all respondents, say Yankee Group researchers.

The two most common bundles are subscription television and Internet access; as well as a two-product bundle featuring home phone and Internet access.

Penetration of the triple-play bundle (subscription television, home phone and Internet access) has grown to 14 percent. IP

If you want to know why specialized, focused organizations capable of providing a more personalized degree of service are an evergreen business in the communications business, just look at what Yankee Group found in a recent survey of small and mid-sized business executives with 250 employees or less, in the U.K. market.

Yankee Group took a look at buying preferences for managed services, including broadband access, email hosting, Web site hosting, computer support, phone service and wireless, in the SME (small and medium-sized enterprise) space.

The conclusions might not surprise some industry executives. Put simply, more than 90 percent of U.K. small and medium enterprises buying managed services don’t trust either telecom or cable operators to support and handle their accounts effectively.

Separate Yankee Group studies of Canadian SMEs have found similar reluctance to outsource services, and trust might be an issue there as well.

The majority of Canadian small and medium businesses are not ready to have a third party manage their IT or telecommunications environments, says Carrie MacGillivray, Yankee Group analyst.

It isn’t that the businesses Yankee surveyed do not grasp the value of being able to focus more on their actual business functions, rather than messing around with phone and data systems. They get that, says MacGillivray.

For some reason, though, smaller businesses still do not see the value proposition clearly enough, she says. “Smaller businesses have difficulty recognizing the cost associated with technology management and administration,” she says.

Note the issue found in the U.K. market survey: the customer service interface and processes are the rub, not the actual attributes of the services themselves. In other words, “it is not what I am buying; it is who I am buying it from.”

The research simply confirms an observation: for decades, ever since it became possible for all sorts of companies to offer communications services to the SME customer segment, smaller and specialized firms have prospered.

The stakes are extremely high because capturing managed services income from an SME “is an all-or-nothing effort,” according to Camille Mendler, Yankee Group vice president, who notes that 90 percent of SMEs deal with one or at most two suppliers of managed services at any one time. So selling anything at all to an SME entity is tantamount to owning half to all of the communications spend, for a period of two years.

Enterprises, on the other hand, tend to deal with five or more managed services suppliers. U.S. providers might want to compare the metrics with their own, but Yankee Group says acquisition cost for a single SME account is $1,079.

That creates a permanent “cost of sales” problem. Collectively, SME investments in information and communications technology are growing proportionally faster than other enterprise segments, the last two years worth of Yankee Group surveys have shown.

They also are quite numerous. In Europe, as in North America, the overwhelming number of businesses and organizations are small or smallish. Companies with fewer than 250 employees represent more than 95 percent of European businesses, Mendler notes.

BT Business, the division responsible for smaller businesses, has about 1.1 million customer accounts. Of those accounts, 82 percent are companies with one to nine employees, and these accounts generate 40 percent of the division’s revenue.

Some 15 percent of its customers include businesses with 10 to 49 employees, generating 20 percent of BT Business revenues. The remaining 3 percent of accounts includes companies with 50 to 5,000 employees, and 40 percent of BT Business revenue.

This is no 80/20 rule; this is a 40/3 rule. Some 40 percent of returns are generated by just 3 percent of accounts.

So BT Business structures its channels as you might expect. It uses telesales for accounts with one to nine employees; indirect channels for firms with 10 to 49 employees. It sells direct to accounts with 50 or more employees.

And though the costs arguably are higher for tier one providers, arguably lower for independents, competitive local exchange carriers, value added resellers, interconnects and others, acquisition costs are high enough that no tier one provider actually assigns a dedicated account manager to any single SME, that leaves an opening for suppliers that actually can afford to do so.

Assessing tier one provider pursuit of sales to SMEs, Mendler says “we believe that the poor functioning of sales hierarchies actually predisposes SMEs to reject managed services propositions irrespective of any real technical or financial benefits that may be on offer.”

It’s all a matter of client touch, says Mendler: “How often clients are contacted, who does it and what they communicate about.” To be sure, there are other potential pitfalls.

Excessively tactical approaches are almost a given in a customer segment where, despite what everybody says, consultative selling costs too much. And while a service provider might well like to sell a more complex, higher-margin product, that typically requires more time and effort, so the default is to the easy-to-explain propositions rather than the managed services sales everybody might prefer.

Still, there’s another important observation here. “The belief now driving BT Business is that winning and retaining SMEs is not just a function of price but also the service wrap on offer and the number of customer touch points it is able to support beyond the direct sales force,” says Mendler.

It might not be easy, but specialists thrive in the SME space because the buying decision is not about the solutions; it is about the support of the solutions and the ease of doing business with the provider of the solution.

So as the complexity of consumer, SME and enterprise solutions increases, buyers are going to be more likely—not less likely—to want to buy from people they can rely on to keep solving problems after the service is turned up. IP

About 49 percent of very large enterprises, with 10,000 or more employees, are receptive to IP communications. That compares with 42 percent receptivity among companies with 2,500 to 9,999 employees and 39 percent among companies with 500 to 2,499 employees.

As is generally the case everywhere, VoIP implementations are driven initially by cost savings, though interest in applications development is growing, as is the case elsewhere. Also, the lure of lower capital expense and operating expense is leading to stronger adoption in some vertical segments including retail, manufacturing and the public sector, for example, say Yankee Group researchers.

As generally is the case elsewhere, security, network resilience and quality of service continue to represent the top objections to IP communications adoption in the Asia-Pacific region, Yankee Group says. Also, as is the case elsewhere, Asia-Pacific enterprises of 500 employees or more are concentrating on IP communications capabilities between enterprise sites, suggesting the importance of the cost-reducing benefits. Just 4 to 8 percent of enterprises use IP telephony for communications off network.

Still, Yankee Group analysts are confident the deployment pace and scale will continue, as objections gradually are overcome and users gain more confidence with the technology.

For the intermediate period, up to about 2008, Asia-Pacific enterprises likely will focus on putting hybrid networks into place and buying IP quality of service features. After 2008, enterprises will start to get more aggressive about IP virtual private networks, wide scale off-network uses for IP communications and integration of newer applications such as video, multimedia and wireless broadband, Yankee Group senior analyst Agatha Poon says.

An earthquake off Taiwan’s coast, which disrupted communications across the Pacific region and Asia beginning Dec. 26, 2006, probably won’t help convince doubters. International phone service was cut off or restricted in some regions, and Internet service slowed to a crawl in much of China after the magnitude 6.7 quake, damaging as many as eight undersea cables.

For the most part, enterprises with redundant and diverse transport services were able to reroute and maintain service, according to published reports by the Wall Street Journal, for example.

Services were gradually restored in the days following the quake but haven’t fully recovered by the beginning of the year. In Hong Kong, Internet connections on
Jan. 1, were running at about 70 percent to 80 percent of normal.

Internet connections in Hong Kong should be fully restored by mid-January, when repair work for the first of six damaged undersea cables serving the territory is expected to be complete. By the end of January or early February, all major cables serving Hong Kong and other countries in the region should be fully repaired. IP

As many as 60 to 70 percent of iPhone customers AT&T signs up will be customers switching from other mobile carriers, argues AT&T chief financial officer Rick Lindner. In fact, some might be surprised if the percentages were that low. So the iPhone might be a huge churn inducer for other service providers.

There are 233 million mobiles in use in the U.S. market. Assume just 10 percent of those users can immediately switch service. That’s 23.3 million users. Assume 10 percent have the appetite to buy an iPhone. That’s 2.33 million users. Assume just half of them do it. That’s 1.17 million buyers.

So if the iPhone proves as popular as some believe, that might mean attrition of a million or so wireless customers from other networks, almost immediately.

A million customers represent just about a whole quarter’s worth of net additions for AT&T. Put another way, that’s possibly $5 billion to $10 billion in incremental revenue on an annualized basis.

To put that number in perspective, consider that AT&T booked $10 billion worth of wireless revenue in the first quarter, and just $1.1 billion in "national mass markets" revenue, representing the old AT&T residential long distance business.

AT&T 1Q07 Wireless Segment Results
Service revenue (billions)	$9.1
Total revenue (billions)	$10
Adjusted OIBDA margin	38.9%
Adjusted operating income margin	25.5%
ARPU	$49.21
Postpaid churn	1.3%
Total churn	1.7%
Subscribers (millions)	62.2

In other words, if AT&T can add a million iPhone customers taken from other carriers, that’s the equivalent of more than two years worth of standalone consumer long distance. And though most people see the iPhone as a device for younger consumers big on their iPods, or bleeding edge gadget freaks, there might be more
to it.

The early adopter crowd, which simply wants the coolest gadget, might be followed by the "I want to carry one device not two, mobile plus iPod" segment.

Other segments might prove unexpected. The iPhone has huge storage capacity and pretty powerful multiple processors. So the issue is how much that power might appeal to non-intuitive users such as specialized segments within enterprise markets, for example.

When the iPhone was announced back in January, Steve Jobs demonstrated a few features of the phone but also said there were many hidden features that were not yet announced. Then one has the almost inexplicable announcement by AT&T that the iPhone is going to be pitched to enterprise customers.

Nonsense, many observers will say. The iPhone is closed, so there is no way enterprises will be able to integrate it with essential business applications. Unless AT&T knows something we don’t. That seems almost certain.

Maybe the iPhone is going to run BlackBerry software. Maybe it will have hooks for enterprise developers. Maybe it will provide terminal emulation. Who knows? The point is that AT&T would likely not be saying iPhone will be sold into enterprises without Apple having something up its sleeve. Barring that, AT&T’s intentions would seem a mistake. And we doubt AT&T will make mistakes of that sort, given its temporary advantage and the importance of wireless overall.

In its most recent quarter, wireless contributed 34 percent of company revenues. Against the backdrop of overall quarterly revenues of $29 billion, mass market long distance, which contributed about $1 billion, has become less important a contributor than the directory business, and a tenth of the importance of wireless.

But there are all sorts of angles to the iPhone initiative. The iPhone might radically change the loyalties of cellular subscribers. Though few wireless subscribers are very loyal to their cellular operator, many people are fiercely loyal to Apple and its iPod media players.

Miro Kazakoff, director of the wireless practice for market research firm Compete Inc. suggests that AT&T is attempting to leverage loyalty to the iPhone as a device, instead of relyng on customer loyalty to AT&T itself.

Verizon Wireless and Sprint, the second and third largest cellular operators in the U.S. behind AT&T, claim to be nonchalant about the potential challenges presented by iPhone. But at least some of their actions belie the insouciance.

"Publicly, they’ve adopted the dual stance of being nonchalant about iPhone and trumpeting their own music offerings," Kazakoff says. "But they’re working to build out their services to compete. They’re not really nonchalant at all."

Alana Muller, Sprint’s director of wireless data marketing, says that her company would have lowered prices, from $2.50 a song to 99 cents, and introduced new music phones even without iPhone’s imminent entry into the market.

The stakes unquestionably are high for AT&T as well. Working with Apple to develop iPhone is so resource-intensive that the project required approval by AT&T’s board, executives acknowledged in January when iPhone was announced. Board approval for introducing a cell phone is unheard of in the cellular industry.

AT&T also is attempting to use iPhone as a huge way to create churn among its competitors’ subscribers and minimize churn for itself. Few devices have been offered so exclusively, and few have proven to have high enough drawing power to drive serious amounts of churn.

And AT&T might win even if customers come into a retail store and decide not to buy an iPhone. They’re in the store. They might see another phone they like and buy that instead. Either way, AT&T wins.

The iPhone also might give AT&T a leg up in the potentially lucrative mobile music market. People are used to sideloading music into an iPod. So what does that do to the walled garden music stores Sprint and Verizon are trying to create?

And AT&T’s wireless business was on fire in any case. First quarter net wireless adds were 1.2 million. During the past year, AT&T has added 6.4 million customers, bringing its in-service total above 62 million. The number of active data users has grown more than 30 percent during the same period a year ago. Where it had 25 million data users a year ago, it now has more than 33 million.

Wireless service revenues grew 13.5 percent in the first quarter while wireless data revenues grew 67 percent.

Wireless data average revenue per user was up 51 percent, and it is now 16 percent of total ARPU. Postpaid data ARPU is during $9, an increase of approximately 55 percent year over year.

Perhaps just as important, data growth, which has historically been driven by text messaging, is now seeing significant contributions from media bundles, email services and data access, the company says.

"If you went back a year ago, I think we would have said data growth was almost entirely due to text messaging," says Lindner, "What we are seeing today is increasingly data growth is coming from other sources."

Customers "are buying a bundled service that allows them a large number of text messages, multimedia messages and Internet browsing capabilities," says Lindner. "It is coming from customers that are adopting email services, like Blackberry, on their mobile devices, and it is coming from customers that are buying data-only packages for their laptops or data-only PDA devices.

"I would have said at one point in time data growth would expand to be maybe 20 percent of total ARPU," says Lindner. "I am not certain that it won’t expand beyond that based on the trends we are seeing."

Right now, the big question is whether iPhone will have the impact AT&T is expecting, first in the gadget freak crowd, then in the music crowd, then in the "I’d rather not carry two devices" segment. Then there’s the enterprise and small business segment.

It could be a very big deal, indeed. IP

So far, AT&T has done that. It has had seven straight quarters of double-digit adjusted EPS growth. The company’s reported earnings per share in the fourth quarter were up 9 percent compared to the same quarter of 2005.

Verizon Communications, in some ways, is a better proxy for how the tier one carrier market is doing, as it largely grows internally. AT&T has grown so much by acquisition that internal growth and prospects are a bit harder to estimate. AT&T simply has more room to cut operating costs, for example, as it consolidates large acquisitions.

And based on fourth quarter earnings, Verizon continues to nudge its business forward financially, whatever issues it continues to have with retail access lines and the maturation of the classic voice business.

In its fourth quarter, Verizon profits fell 38 percent to $0.35/share ($1.03 billion) from $0.59 in the fourth quarter of 2005, while sales rose 26 percent to $22.6 billion. The decline largely comes from charges related to the sale of assets in the Dominican Republic and costs from the spin-off of its directories publishing business, of which net profits were $0.62 (positive 5 percent), in line with analyst forecasts of $0.61.

Verizon wireless brought in 2.3 million new wireless subscribers (2.2 million of which were retail) in the quarter while competitor Sprint Nextel lost 306,000 contract customers, by way of comparison. That’s an increase of about 12.7 percent from the same quarter of the prior year. The company continues to set records for low churn: total fourth quarter churn was 1.14 percent, and 1.17 percent for the year.

Wireless churn rates remained quite low in the fourth quarter, as well. Overall churn was 1.14 percent. Total service average revenue per unit and retail service ARPU, which was $50.78, both were up year-over-year in the fourth quarter and for the full year.

Verizon ended the year with 7 million broadband connections, up 35.7 percent year-over-year, with more than 1.8 million net new broadband connections  (FiOS and digital subscriber line) added in 2006, with 409,000 net new broadband connections in the fourth quarter.

Annual data revenues were $16.1 billion, or 31.6 percent of total annual wireline revenues, the company says. Still, AT&T seems to get more headlines, not the least of which is the Unity calling plan that creates an on-network free calling plan for all AT&T customers, whether wireline or wireless.

One might question how all of that will work, but in principle it is a huge shift. AT&T has 100 million access lines and wireless end points. All now can be a single peered community. Throw in Verizon and Qwest and most U.S. consumers might theoretically be part of a single peered community of users, with “no incremental cost” calling between the end points.

At the very least, everybody that competes with AT&T and Verizon, on either the business or consumer fronts, has to ponder the implications.

“Our view that the industry was beginning to rebound has been validated,” Whitacre says. “I believe the industry has made its positive turn, we are still early in the ramp, and there is a lot of upside opportunity ahead.

“So, today, we are reaffirming the outlook we first outlined for you a year ago, including double-digit adjusted EPS growth in each of the next two years,” Whitacre emphasizes. Some of that is the result of cost-cutting efforts.

One might legitimately note that cost cuts, rationalization and market share gains cannot forever defy a worsening future for classic voice or high-capacity services. Indeed, one would be better off betting that the revenue vise must inevitably squeeze harder. But AT&T seems to be quite well aware of that.

“But I want to make one thing clear,” Whitacre says. “While cost reduction is important, we also plan to grow the top-line at AT&T.” And AT&T has some advantages. Like other global carriers such as BT, AT&T can provide a level of service to the largest of business customes that smaller competitors will struggle to match.

Indeed, to the extent that BT, Verizon and AT&T have begun to wean themselves off the voice market, especially the consumer segment, it is because they have other levers to pull.

“Wireless makes up more than a third of our total revenues, approaching a $40 billion annual run rate, and we expect wireless service revenues to continue the kind of growth we saw in the second half of 2006,” says Whitacre. Likewise, AT&T’s enterprise business isn’t something a more specialized company can match.

“Business and wholesale together make up nearly 40 percent” of company revenue, Whitacre notes. And few smaller providers will be able to match the wholesale volume BT, Verizon and AT&T can boast of.

“Wireline consumer is just under one-fourth of our total revenue base,” says Whitacre, though noting that broadband and video represent opportunities even there. “In total, more than three-fourth of AT&T’s revenues today come from wireless and sales to business, and a high percentage of our growth is coming from advanced data and broadband services; that’s true in both wireless and wireline,” Whitacre says.

AT&T’s EPS growth drivers include wireless, where revenue growth is accelerating and margins continue to ramp up, says Rick Lindner, AT&T chief financial officer. “And at the same time, wireline revenue trends are improving, particularly in business.”

But wireless remains a star. “In the fourth quarter, we had record gross adds, record net adds, and we ended the year with 61 million subscribers,” says Lindner. Wireless data revenues were up 69 percent as well. “Data ARPU was up $0.87 in the quarter and 53 percent over the past year,” says Lindner.

“We expect wireless service margins to be in the high 30 percent range for the full year 2007 and to exceed 40 percent for the full year 2008,” Lindner says.

AT&T’s in-region wireline revenues grew 7.5 percent, that’s up from a 6.6 percent growth rate last quarter, says Lindner. Consumer wireline revenues held steady at 0.1 percent.

About 63 percent of AT&T’s year-over-year consumer revenue declines came from lost AT&T standalone long distance. The smaller and mid-sized business segment, both voice and data, was a continued area of strength.

“We continue to see double-digit growth in small and medium business,” says Lindner. “We continue to see good growth in both voice and data in regional business,” as well. In fact, access lines grew on the SME front.

“At the same time, our regional business data revenues grew 18 percent, with solid growth in transport and 25 percent growth in IP data, led by strength in managed Internet services, virtual private networking and DSL,” says Lindner.

Also, to nobody’s surprise, data revenues continue to grow. “Data now makes