Don't Have It, Don't Want It

One line of reasoning about broadband deployment is that carriers and government have failed to push it hard enough. There's another way of looking at matters, however. In the U.K. market, for example, respondents who don't buy broadband access services recently told researchers that they didn't want broadband, or didn't own PCs.

Some 27 percent just didn't want it, and some 22 to 23 percent don't own PCs in the home, so broadband has no value. About 15 percent said they couldn't afford it. This will change as compelling services aimed at TVs and game consoles, for example, become available.

In some cases, the ability to use mobile handsets inside the home for a flat fee each month might be so compelling that will drive broadband adoption. The point is that people buy things when they find things valuable, and many end users still haven't been persuaded.

Wireless IS Voice

This probably won't be the year that everybody agrees mobile voice is THE way to do voice. In fact, mobile might never be THE way to do voice. But there's no question that mobile handsets are becoming the convenient single device to use voice and other apps. There's no question the amount of conversation initiated and terminated on mobile devices is increasing. Nor is there much doubt that tethered devices and apps have to start offering value beyond "simple" conversation to remain highly relevant.

P2P Could Drive Architectural Change

Some trends we now are seeing in the Japanese broadband access market raise potential issues for all access network architects. Specifically, the issue is that consumer networks are designed asymmetrically, and P2P uploading requires symmetry. One might like to design a network so flexible that bandwidth can be provisioned "on demand."

The problem is that such a network typically implies an optical infrastructure with a fairly sophisticated degree of intelligence, to supply granular increases in bandwidth, on a "dial it up and it happens" basis.

And that's expensive, particularly so since the network has to be built past all potential end user locations, even when actual demand will vary widely, stranding some of the assets.

Some networks, with short access loops, might finesse the issue by deploying VDSL. Others are simply going to have to look at fiber-rich networks. Expensive stranded assets might result, in that case.

Oddly Enough, Access Is Where QoS Really Helps

Brough Turner, NMS Communications CTO, argues that quality of service measures in the Internet backbone provide negligible benefits. It's in the access links where QoS really can make a difference, and that's the area where at&t's merger approval agreement with BellSouth prohibits it from doing so. Oddly enough, the inability to "discriminate" between packets prevents users from experience improvements they might like to have.

Once packets get beyond the access network, every link in the Internet is carrying multiplexed traffic that is statistically relatively predictable, Turner argues. "Traffic volumes vary by time of day, but these links don't saturate, except as a result of poor engineering or forecasting on the part of the ISP or failures in other parts of the network causing rerouted traffic," says Turner. "Either case generates a rapid response from any ISP that expects to remain in business."

"In short, except at the edges of the access network, Internet links may be heavily loaded but are not saturated," says Turner. "Best effort is good enough even for low latency applications like voice telephony" in the backbone, as a result. "Except during major failures, the effect of QoS in the Internet backbone is negligible."

That isn't the case for access links, Turner argues. "There is one place in the public Internet where limited, highly specific QoS measures make sense and are being deployed," and that is "at the consumer end of an asymmetric broadband access link."
Specifically, it is the upstream links that can saturate. The other issue is that it frequently is not possible to buy additional capacity in the upstream direction, at any price. And that makes routing policies very valuable.

His full comments can be viewed at http://blogs.nmss.com/communications/.

Incentives Still Remain the Issue

Nobody yet knows how much bandwidth typical consumer users will want, and be willing to pay for, in the future. Demand may prove less robust than many now believe. But if customers really want, and will pay for, 50 Mbps or 100 Mpbs access pipes, then a fiber direct to residence network is about the only thing that really works, if demand is strong enough that little sharing is possible.

So the problem is how to get such networks built, at a time when capital providers are skeptical, to say the least, about the wisdom of building such networks.

Perhaps we should simply acknowledge that some partisans in communications policy debates seem to prefer government solutions that remove profit motives entirely from broadband infrastructure efforts. There is a logic there. Some sorts of infrastructure might indeed be natural monopolies that simply can't be created by the private market. And if that is the case we are wasting our time thinking such sorts of networks can be created any way but sanctioned franchises and monopolies.

And that remains one of the issues at the heart of the network neutrality debate. To the extent that we honestly expect private capital to flow into infrastructure, we have to honestly confront the issue of whether profits can be made, and how they can be made, by entities trying to supply such infrastructure, especially when we haven't tackled the issue of how different regulatory regimes can be reconciled. Why, in the same market, do we regulate providers of similar services using entirely distinct frameworks? It's a rhetorical question, of course.

Some contestants come out of a broadcast mass media model with greater degrees of content freedom, while others come out of a common carrier model with drastically less freedom. But by universal agreement, all the services provided by all of these contestants will be the same. Because of the immense commercial interests in play, nothing signficant is going to be done about this growing inconsistency any time soon.

Nor is it likely private communications firms will simply be nationalized and operate as regulated utilities. That being the case, the issue of incentives will remain key if we are to have any hope at all at mass and robust broadband availability.

ILECs will Lose 30 Percent of Households by 2010

By 2010, 30 percent of households in the United States will completely cut their service with their local phone company, predict analysts at The Yankee Group. These households will upgrade to broadband and choose VoIP for their home phone service or become wireless-only households. But there are countervailing pressures.

Some 48 percent of households with both a wireless and landline phone say their wireless phone will never replace their landline phone. The main barriers to wireless-only service include reliability and quality. Some 19 percent of wireless users cited quality and reliability as primary reasons for retaining the landline phone. Twenty-two percent of wireless users cited indoor wireless coverage and 28 percent cited battery life as reasons not to cut the cord. Wireless quality and reliability will continue to improve, and will weaken as reasons to keep the landline phone.

About 37 percent of wireless users say Internet access requirements are a primary reason for retaining the landline phone. About 45 percent of households have DSL or dialup, which both require a landline connection. With the exception of Qwest and a few others who offer it as a retention strategy, local phone providers are still not offering consumers stand-alone DSL service. Over the next few years, Yankee Group researchers expect local phone providers to offer DSL without local phone service ("naked DSL"), reducing the need to keep a POTS line in service.

About 37 percent of wireless users cited emergency calling as a primary reason for retaining the landline phone. Some 26 percent of wireless users cite a unified family phone number as a primary reason for keeping the landline phone. A single phone number that provides a family identity, that can be given as the main contact number, is valued by this consumer segment.

Consumer Directory Services: A Work in Progress

The majority of consumers still use a traditional paper address book to keep the names and numbers of their friends and families but the mobile phone and instant messaging services are increasingly important, and points out the way service providers or portals can use directory services--especially enhanced with availability features--to "glue" customers to their offerings. Some 25 percent of households recently interviewed by The Yankee Group said that they always or sometimes keep contacts on a web-based e-mail account. About 32 percent of users say the mobile phone is the device where directory information is kept.

Directionally Helpful Trend

If one is a would-be provider of IP video content provider in the European market, the trend is helpful, if suggestive of the business challenges. As one might expect, few users say they have paid for downloaded or streamed video content. But the percentage of users who say they have done it, or would do so, seems to be growing. If we assume that the primary audience to this point has been younger demographics rather than mainstream video consumers, that's helpful.

How BT Got Its Voice Revenues Down

A challenge every legacy telco faces is transforming its revenues away from voice. One advantage BT has is that it is a global carrier, able to build a wholesale business, serve transnational enterprises, small and medium business as well as consumer customers. Increasing its efforts in all sorts of new and data-related services has allowed BT to reduce its reliance on voice revenues far below levels typical of other "telephone" companies. Smaller telcos will have a tough time replicating the global enterprise and wholesale parts of BT's strategy.

Email, Web, IM Top Enterprise Mobility Drivers

Email, Web and instant messaging access are the top three drivers for enterprises adding more mobility support for their workforces. Smart phones are the new black, it seems.

Build Strategy Makes a Big Difference

Verizon and at&t are fighting local franchising rules of various sorts that the carriers say are deterrents to building new broadband networks. Without fanfare, other providers such as RCN build Triple Play networks without serving the whole community, and find the going tough even without mandatory communitywide buildout requirements. The problem? The economics of competitive local networks are tough, since no single provider can reasonably expect to get much more than 20 percent of potential customers in a market with four providers, for example.

Another way of putting matters is that 80 percent of the local ports have no customers on them, and no revenue. That's not a prospect likely to make capital providers very comfortable. The fact is that some parts of every community, and some neighborhoods within every community, are more profitable than others, as indicated by Solon Management Consulting. Building in the most-profitable neighborhoods first, to get the revenue flowing, before tackling the more problematic areas is simply a business reality.

Arbitrage Won't Drive VoIP Forever

To this point, global VoIP usage has been driven by price arbitrage. But global calling rates are coming down, so that form of arbitrage will prove less interesting to end users at some point. At the same time, wireless usage continues to climb. And while there might be some room for price arbitrage in the wireless domain, it will be found precisely where arbitrage as proven most significant in wireline calling: international termination.

That probably doesn't mean price arbitrage forfeits its key role in user adoption any time soon, simply because there are other cost elements to arbitrage. The U.S. cable operators, in fact, have extended price arbitrage into the local calling realm, offering VoIP-powered services that simply mimic the PSTN, but arbitrage the recurring access revenue stream, not the minutes. Inevitably, though, at some point, VoIP simply won't be about price arbitrage any longer, because there won't be much left to arbitrage. Users will have found they are paying for services and features way beyond the ability to place or receive a call, and service and app providers will have adjusted accordingly. A "soft landing," you might say.

Piracy Part of the Business Model?

Up to this point, younger Internet users have been the most active abusers of video content rights on the Web. Much content now is pirated, according to The Yankee Group. We doubt this will continue to be so much the case as downloaded video becomes mainstream and much of the demand shifs to high definition TV fare where image quality and consistency will really be important.

There also is some precedent in the video world for a small amount of piracy to be a good thing for the overall value chain and growth of a new business. In its earlier days, the U.S. cable TV industry used filters of several types to control viewing of premium services such as Home Box Office. So piracy was a matter of securing the proper filter and inserting it between the house drop cable and the distribution cable's "tap." In this scenario, manual audits were the only way to determine whether an installed filter was legal or not. And since assigning technicians to patrol the entire plant looking at "traps" costs money, there is a cost-benefit issue.

A service provider wants to minimize theft of service, but not completely, as the cost to control then outweighs the "losses" from theft. Also, there is a subtle way that some amount of piracy actually creates more revenue, at least for the service provider. The example is subscribers who pay for basic cable access but then are able to pirate HBO. In that scenario, a cable operator might actually get a recurring revenue stream it otherwise wouldn't have gotten, while the economic loss is borne by HBO, which isn't paid. With changed technology, this particular attack is seen rarely, if ever. Also, access to HBO no longer drives cable, satellite or telco TV penetration.

The point is that some amount of piracy must be tolerated because complete eradication costs too much. And there might even be some cases in which a limited amount of piracy actually can lead to a bigger revenue stream for legal uses. Earlier in its development, productivity software suppliers realized that some amount of piracy actually would increase the base for legal sales, because users grew accustomed to the use of particular products and then would buy legal copies.

As the network-delivered video business develops, particularly as an application or service used by large numbers of people who don't want to hassle with technical details, and just want to watch content on their HDTV and other screens, we likely will see a significant decrease in piracy and an equally significant increase in legal downloads. We might even find, as we have found before, that some amount of piracy can actually stimulate the legitimate part of the business.

Both Substitution and Augmentation

The common sense expectation we all seem to have is that VoIP ultimately replaces POTS, and that's correct in a technological sense. Their respective business futures might be less certain. We mean "business" here in the sense of revenues that can be generated from services and features associated with POTS, even if the platform changes. Those of you who lived through the analog to digital switch transformation understand this. Revenues sometimes grow when a new platform displaces the old, because there are new things to sell, and costs sometimes drop at the same time, allowing a retail price to reap more margin, even on the legacy products.

And one of the things about POTS is that there are latent values beyond the switch and transport technology that might be the real reasons customers buy POTS service. If you think in terms of "user communities," the PSTN outstrips nearly anything else we've built so far, with the exception of wireless. But if wireless is seen as untethered access to the PSTN, then the PSTN community is where one wants to be.

In other words, the PSTN directory is far more developed than any directory you or I have built ourselves. A "buddy" list can be built for the small number of people you communicate with regularly. Beyond that, the PSTN still rules. That's valuable. One might argue, at least at this point, that POTS is more reliable, in just about every sense, compared to just about any VoIP alternative. That has value for end users, and might continue to, even when the infrastructure changes.

Also, the business value of the PSTN, which is voice-oriented, arguably is enhanced when integrated with instant messaging, text messaging, conferencing, Web access and other features. All of that might help explain why just half (49 percent) of residential customers who adopt VoIP say they discontinued a traditional phone service when they got their VoIP services, according to a recent survey by In-Stat. Put another way, VoIP displaces half of existing POTS line purchases, but also augments POTS half the time.

Also, in addition to "cord cutters" who rely exclusively on wireless for their personal voice communications, In-Stat finds that nearly 12 percent of respondents say their only VoIP service is based on the use of soft clients. That's further evidence of the growing importance of IM-based voice, presumably including "call out to the PSTN" capabilities.

And blurring the distinction between "business" and "consumer" uses, half of the VoIP users say they use their residential VoIP service either in part or in whole for business purposes. The trend is especially pronounced among VoIPers who are softphone-only users. More than 40 percent of such people use VoIP for business applications.

There are lots of suggestive niches here. Even in the consumer space, it appears there are augmenters, displacers, "no phone" users, "no phone" business users, IMers and maybe people who don't like "gizmos" of various sorts (probably the same people who dislike cable decoders and other CPE).

Metro Ethernet Getting to be a Bigger Deal

As wider adoption of IP phone systems spurs buying of SIP trunks, so IP itself drives demand for Ethernet access. The Yankee Group sees spending by enterprises in the Asia-Pacific markets doubling in four years, for example. "Year over year we see metro Ethernet becoming a greater percentage of our sales," says David Rusin, American Fiber Systems CEO.

The demand is "steady" at about 40 percent growth this year, Rusin says. And Rusin is very clear about how this market needs to be attacked. "After 10 years and billions of dollars of wasted capital we now know you must have your own network."

Needless to say, Rusin isn't a big fan of mandated wholesale access to high-bandwidth facilities. In the U.S. market, mandatory unbundled network element access should go away, maybe over a five-year period. If all competitors understand that they must have their own facilities to compete in the access market, that is what they will do, or they will do something else. "Then capital will come back in to the market," Rusin says. "The incentives need to be there."

In a market as competitive as the metro Ethernet space, how does Rusin answer the objection that there already is too much competition? Simple. "We don't build where there already is capacity," he says. In other words, if there is any existing fiber or capacity in place, AFS simply leases the capacity. The only time it will swing an optical lateral is when there is no other alternative. "If you aren't operating off your own network, IRU or capacity, I don't see how in the long run that's a sustainable business," he simply says. Lots of other providers would say the same thing.