Editor's Outlook
January 2002

Kevin Mayer

Bred To A Harder Thing

BY KEVIN MAYER


Sad to say, but this column will be my last on behalf of Communications Solutions magazine, which ceases publication with this issue. Having served as the magazines editor since its founding, over six years ago, back when it was called CTI magazine, I will miss it at least as much as anyone else. I imagine that I may, for some time to come, interpret the communications news I hear with an impulse towards thinking of how it could inform or enliven the magazines coverage. This impulse, however, will simply hang in the air, like a ghost, like a shadow sensation of a lost limb.

No doubt others who have known the magazine will also feel its absence, even if that absence is accompanied by nothing more than a mild sense of loss. The sort of feeling you might get were you to walk past a boarded-up storefront: a brief pang of sympathy, which dissipates as soon as you begin wondering what a new tenant might bring to the neighborhood.

Mourning, whether long or brief, is succeeded by other thoughts, the most edifying of these, perhaps, is the idea that while some things may pass, others last. With or without Communications Solutions, the magazines subject, voice/ video/data convergence, remains as compelling as ever, inspiring suppliers, distributors, and users to stay current, to discern the relevance of continuing development, and to recognize emerging opportunities for more natural and effective communications.

When I look to the future of voice/video/data convergence, I cannot help but see it through my editorial perspective. I think of the educational challenges that remain, the continuing need to filter through a mass of information, to acquire insight, instead of being swept along by a sea of unqualified claims and isolated facts. Nothing is so elusive as insight, particularly in a discipline that is so alive, that evolves so quickly, that is animated by the contributions of so many constituents.

Going forward, industry constituents will struggle, as always, for knowledge, comprehensive knowledge, while staving off confusion. Unfortunately, industry constituents, Ive noticed, persist in habits that all but guarantee confusion. I dont suppose these habits will fade. In fact, with the passing of Communications Solutions, they may grow all the stronger.

ALL THE TRUTH IS OUT
The magazines readers and contributors (and would-be contributors) often betrayed how narrowly they perceived their own interests. Readers would often call or write to me and ask very specific questions, unique to their situations, as though I were qualified to act as a free consultant, as though I were at liberty to tell them which product in a given category was the best, never mind that I tend to doubt there is ever such a thing as a single, undisputed, all-purpose best.

Prospective contributors would call or write, demonstrating little or no knowledge of or interest in the magazines objectives, while hoping that the magazine would nonetheless deem this or that piece of news, any random snippet, worthy of front-and-center treatment. Many a would-be contributor would ask, Whats the focus? Exactly the wrong question. Better to have asked, How may we establish relevance? Then, the ensuing discussion could have broadened, establishing connections between the interests of multiple constituents, creating a picture, creating a story, simply by connecting the dots.

Although many contributors were more than capable of connecting the dots, the shame of it all is that most resisted, and will continue to resist, anything but the most narrow self-focus. That is the danger. There will always be the temptation to intensify chaos, to succumb to perpetual novelty. When contemplating the diverting world of chaos and novelty, I sometimes remember a famous scene from a film called Sammy and Rosie. In this scene, Sammy flops down on a couch and clicks on the television. Then he flips open a magazine. And then he puts on a pair of headphones. And then he begins stuffing his face with junk food. (And thats not all he does, but decorum prohibits mention of these activities.) No doubt, if Sammy had access to a wireless PDA, hed have been logging on to the Internet and checking his e-mail, too. If that had been the case, Sammy would have been even more oblivious to the mayhem in his apartment. (A pair of enraged, knife-wielding lesbians were chasing his father, who had provoked the attack by making free with insults.)

Equally dangerous is an entirely different response to perpetual novelty. Instead of simply swimming in novelty, one can attempt to absorb it, in anticipation of filtering and processing it, until one imagines being able to act as a channel or medium, articulating the essence of what at first seemed so resistant to summation. Such a response to novelty, I fear, appealed too much to me, in my capacity as editor. So comprehensive an approach is beyond what is practical.

A LAUGHING STRING
To demonstrate how such an extreme response to novelty can tempt fate, I cite the example of Joe Gould, a real-life figure who had his story retold in Joe Goulds Secret, a recent film. Gould described how his life changed upon reading a passage of Yeats writing. The history of a nation, said Yeats, is not in parliaments or battle-fields, but in what people say to each other on fair-days and high days, and in how they farm, and quarrel, and go on pilgrimage.

It struck Gould that he could apply Yeats notion in the creation of a different kind of history, an oral history. Gould would listen, simply listen, to everything, no matter how trite or everyday, and synthesize all that he had heard into a grand statement, The Oral History of Our Times. Warming to his vision, Gould wrote, Ill put down the informal history of the shirtsleeve multitude ... a great hodgepodge and kitchen midden of hearsay, a repository of jabber, an omnium gatherum of bushwa, gab, palaver, flapdoodle, and malarkey, the fruit of 20,000 conversations. What people say is history, what we used to think was history, is only formal history, and largely false. Ill put down the informal history, or Ill perish in the attempt.

Gould never did complete the Oral History. Perhaps the demands of doing so overwhelmed him. After his death, in a sanitarium, all he left behind were a few dozen composition books, scattered here and there amongst friends and acquaintances, who held them for safekeeping, since Gould had lived homeless, and had no place to keep them himself.

And yet it could be said Gould left behind something else. Gould, the film suggested, may have infected another writer with the idea of embracing a world of chaos as the way to create a work of profound scope, beauty, and order. The other writer, Joe Mitchell, had written two profiles of Gould, and was, in some respects, a success at what Gould had attempted. Mitchell specialized in writing stories of New Yorks eccentrics. Towards the end of the film, Mitchell was shown describing a book he was planning, a work of unusual subtlety and depth, a work, it seemed, uncannily like the Oral History. Mitchell, however, never published another work following his second Gould story. Instead, he contracted a case of writers block that lasted thirty years, until his death.

A PLACE OF STONE
The lesson I derive, and which I offer to you, is that neither the haphazard nor the overly deliberate may promote understanding, let alone a productive life. On the one hand, you have a ceaseless blare of meaningless noise. On the other, silence. Understanding, as well as a purposeful life and productive work, may depend on a middle course. Something to keep in mind the next time you are strolling the aisles of a trade show floor, surfing the Web, reviewing your e-mail, or reading your newsletters or magazines. You might want to ask yourself if you are, like Sammy, mindlessly adrift in a sea of information. Alternatively, you might want to ask yourself if you are, like Joe Gould, indulging in the notion you can absorb a mass of information, a universe of particulars, while somehow approaching omniscience.

In my tenure as editor of Communications Solutions, Ive been much preoccupied with the problem of charting a middle course. Ive tried to be selective, as opposed to indiscriminate. Ive tried to frame the representative, as opposed to the impossibly comprehensive. Ive tried to see connections, and to encourage others to see connections, instead of isolated islands of self-interest. Lacking the wattage to light up the whole stage, I tried to illuminate subjects as well as I could, so that they might glow suggestively, an approach more sensible, I think, than training a narrow spotlight here or there, or anywhere, letting it swing wildly from one end of the stage to the other.

And now, in my last column for Communication Solutions, I have one more opportunity to demonstrate how a middle course approach might work. In this column, Ive taken up the issue of enhanced communications functionality, and whether this functionality is best based on the premises or in the network. More often than not, each side in this debate would have its say in a separate article. For example, a premises-oriented supplier might brief an editor one day, and a network services-oriented supplier might brief an editor another day. Different days, different briefings, different stories. Each story with a definite focus, one a premises focus, the other a services focus.

It is possible, however, to create a story that would take in both focuses. But why bother? Well, in this case, I noticed that whenever I heard people taking sides in premises/services debates, the next-gen service provider advocates would denigrate PBXs and key systems. Similarly, the LAN telephony or IP-PBX advocates wouldnt stop at showing up the limitations of traditional PBXs. They would also denigrate Centrex. To me, the arguments from both sides seemed off the point. Wouldnt it be more appropriate, I thought, to discuss the relative advantages and disadvantages of IP-PBXs with respect to IP Centrex (or virtual PBXs or whatever theyre called this week)?

Pursuing this question, I thought to poll a range of suppliers, asking them all to respond to the same set of questions. I spoke with a premises-oriented supplier called Sonexis, which provides adjuncts to premises-based systems, such as PBXs, to deliver enhanced functionality such as collaborative audio and Web conferencing. I spoke with Intel/Dialogic, a supplier of component technology that may be applied in either premises-based or network-based systems. And, finally, I spoke to aTelo, a supplier of media servers, which serve as a next-gen platform for enhanced services. My thanks to the individual respondents, which included David Friend, chairman and CEO of Sonexis (www.sonexis.com); Peter Gavalakis, marketing manager, public network products, Telecommunications & Embedded Group, Intel Corporation (www.dialogic.com); and Huan C. Le, president, aTelo (www.atelo.com)

Ive avoided representing just one category of supplier, which would have limited the story to one point of view. In addition, Ive avoided compiling a comprehensive list of vendors from each category. Instead, Ive decided to present a few representative voices on a topic not so preoccupied with a single focus that it would fail to recognize a continuum between topics.

ComSol: What are the advantages/disadvantages of IP-PBX treatments with respect to IP Centrex treatments?

Sonexis: There are several reasons that people cite for bringing services like PBX, conferencing, and messaging in-house:

  1. Cost. At some point, most services can be brought in-house at significant savings. The economies of scale that service providers used to count on were driven by very high equipment costs, and these costs have dropped precipitously in the last five years. Thus the crossover point for customer premises equipment (CPE) has dropped too much lower dollar amounts.
  2. Privacy and control. Particularly with services such as Web conferencing where you dont want unauthorized individuals viewing your PowerPoint presentations, keeping control of the system behind your own firewalls can be extremely important.
  3. Reliability. Companies worry about the reliability of their vendors and their systems. Some prefer to operate their own CPE for that reason.

In general, a shared resource like IP Centrex has to offer either a functional or cost advantage for the user. Even though CPE prices have fallen along with all computer equipment, there is still a minimum nut to bring any CPE in-house. And most systems dont scale down to very small sizes. Therefore, outsourced services such a conferencing or Centrex work well where usage is low or where the number of users is very small or where there are exotic features not supported by CPE.

An example of an exotic feature might be the high touch conference calls that are used for quarterly analyst briefings on Wall Street, where a live operator is involved. There are roll-calls and a lot of hands-on involvement that wouldnt be necessary or desirable in a regular self-service conference call. Since usage of high-touch conferencing is low and intermittent, it wouldnt make economic sense to bring that capability in-house.

Intel: Your logic is sound. These debates typically revolve around feature comparisons. If (and this is a big if) there is 100 percent feature parity, we move to the classic in-house versus outsourcing issues companies face. Tops on the list would be:

  1. Security (advantage: CPE).
  2. Ability to customize features, and the cost of doing so (advantage: CPE, but less so with user provisioning available via Web-based interfaces).
  3. Capital and time required for deployment (advantage: outsource).

The last point is one of the most significant benefits of the hosted approach. We are increasingly seeing hosting service providers pitching their services as a way to get started quickly. Migration to an in-house solution could be done later. Beyond this, of course, you have to talk in the specific context of a service. Benefits for a network-hosted, cross-media call center to enable remote agents would be different from those for a unified messaging application.

aTelo: Yes, it is only fair to compare IP-PBX systems with IP Centrex services. The comparison, however, is exactly the same as comparing traditional PBXs with traditional Centrex, except for the fact that in an IP Centrex environment, the connection between the customer premise and the central office is still just one pipe.

ComSol: Do you see service providers having difficulty bundling together access services and telephony services? For example, would service providers hesitate to limit themselves to transport services while partnering with, say, a communications ASP (which might take responsibility for such things as telephony functionality, call accounting, and billing)? Is it possible that the issues are not merely technical, but also operational and political, in that it might be unclear in such a scenario exactly who could be said to own the customer? Or is it possible that merely operating a multimedia IP network, accounting for QoS and dynamic bandwidth provisioning, mediating between networks of varying capabilities, is sufficiently challenging to permit providers confined to transport services ample opportunity for differentiation?

Sonexis: This is a very interesting question. My own personal view is that owning and operating a transport backbone, telephony or data, is a very different business from offering high-margin services. One is a commodity that requires superb and efficient execution, whereas the other is much more labor intensive and involves significant customer hand-holding. Many carriers and LECs are enticed by the margins of enhanced services, but more often than not they do a poor job at these services, and they get distracted from being a great operator of pipes. Speaking of pipes, this would be like your local water company also trying to sell you lawn sprinklers, washing machines, and espresso makers. Yes, they use water, but is it really a business that has synergies with running pipelines and pumping stations?

Most carriers have thrown in the towel on directory services, but I see them continuing to get into new enhanced services nonetheless. They dont seem to learn. IP Centrex is a great product, but it wont work unless it is offered by a company that sees this and related services as their primary business.

Intel: The bottom line is that end users like to deal with a single service provider if possible. However, offering all services out of the network can be challenging. Weve seen a number of cases where service providers have the ultimate goal of providing service from their network, but the first step is to partner with another service provider. I agree that politics play an important role as well.

aTelo: Just as service providers today have retail and wholesale businesses, so will they have these two businesses in a converged world. Those companies that do strictly wholesale today (Level 3, Genuity) will likely choose that same focus in the future. Companies that are more diversified (WorldCom, Quest, Sprint) will likely have wholesale divisions (which establish relationships with communications ASPs) and retail divisions (which are communications ASPs in their own right).

ComSol: Are there any advantages or disadvantages (with respect to services-oriented versus premises-oriented approaches) when it comes to distribution? Traditionally, you could say few have distinguished themselves in this area. How many customers find would say their service providers have been responsive? How long have we heard that data resellers need to learn about voice, and that interconnects need to learn about data? Might the IP-based communications systems, such as IP-PBXs, be far enough within the comfort zone of data-oriented resellers to allow everyone to move beyond the same old problems?

Sonexis: Centrex and other outsourced services are particularly attractive where users are working in small, isolated groups. Such individuals or groups could be networked together as if they are all in one building. This simplifies matters for individuals, small organizations, or organizations where employees are widely dispersed. IP-PBX vendors, however, will tell you that because their equipment is so cheap, most of the Centrex services can no longer compete. I think there will continue to be compelling reasons for both.

Intel: Many of the traditional telephony CPE distributors have trusted relationships with their customers. This could work to the advantage of the CPE-based alternatives. Ironically, many of these folks are with the ILECs, but its the non-regulated equipment side of the business. It may be a harder sell to get existing companies to migrate to an outsourced model (unless it is an incremental migration).

Regarding your last point: I think this is a dangerous assumption. IP is the transport -- the customers problems and the applications that solve those problems are independent of transport. Perhaps this is an oversimplification, and there are certainly things that the data folks can handle, but there will be a ramp-up.

aTelo: This question really asks whether its easier to sell, provision, and support premises-based services or network-based services. The answer is that it depends on how well the service/application is implemented. The easiest service to sell is one that takes the least amount of effort to set up. One of the advantages of IP-based voice systems versus traditional analog systems is that they require one less set of wiring -- voice is carried over the same wires that carry data and, hence, you end up having only one wire coming out of the wall instead of two. A further advantage of network-based IP services is that you neednt install and configure premises hardware, that is a PBX. Ideally, all a distributor would have to do is install IP handsets and point users to the service providers Web site for self-provisioning, making phone service as easy to procure and maintain as a Hotmail or Yahoo! e-mail account.

ComSol: As for administration, arent in-house MIS and IT personnel already hard-pressed to deal with their existing responsibilities, much less assume new ones, such as supporting telephony functionality? For this reason alone, might not small and medium corporations, the very corporations that might be hard-pressed to retain trained IT staff, decide to outsource communications services?

Sonexis: Youre making the assumption that outsourced services take less administration than in-house services, and thats not always the case. For example, a law firm that I am familiar with used to spend about $8,000 per month on conference calling with an outsourced services vendor. When the bills came in, they had to be apportioned back to individual client accounts, which was a manual process. When they bought their own conferencing CPE, the cost of conferencing (basically the amortization of the CPE) was so inexpensive that they decided to skip the billing altogether, just like they no longer bothering with trying to bill telephony minutes back to the client -- its built into the basic hourly rates. So, they not only saved on the cost of the service, but they actually reduced administration at the same time. The equipment itself is simple enough that a secretary may handle the administrative chores. Most routine tasks, like scheduling calls, are accomplished via a browser by the assistants (they had to do this with the outsourced services, too). The installation was done by a reseller, who may visit if the equipment were to fail (so far it hasnt). Sometimes dealing the bureaucracy of a service provider is worse than the support of in-house CPE.

Intel: This is the logic, but weve been hearing this for a while. There has traditionally been success with things like hosted IVR applications. These will surely migrate to speech-enabled IVR, eventually over an IP transport. However, I think what will ultimately drive the logical trend you point out will be the elusive killer converged communications app.

aTelo: Yes. This is one of the arguments for why IP Centrex will take off when VoIP termination is widely available.

ComSol: A key target -- SMBs -- are often located in corporate parks and multi-tenant structures. Just the sort of companies targeted by broadband MTU/concentrator/aggregator-equipped BLECs, some of which would be happy to bundle together connectivity and communications services on behalf of their subscribers. If a tenant in an office park had access to such a BLECs services, would it really be a good candidate to buy and take responsibility for an IP-PBX?

Sonexis: It depends on the nature of the BLEC. Some have pretty limited physical infrastructure -- they lease other peoples fiber and connect to pre-wired buildings. They are basically a services aggregator and reseller. The essence of the business is customer service. Others are facilities-intensive, with lots of their own fiber, switching centers, and the whole bit. The essence of their business is to keep the infrastructure running. The more your business is high-touch with the customers, the better youre going to be able to do with enhanced services such as shared PBX, collaboration, etc.

A fundamental question, however, lingers. As equipment prices fall through the floor, where are the economies that used to make a shared resource advantageous? It used to be that only the largest enterprises had their own private phone exchanges. Now every little company has an in-house PBX. Having said that, I think a BLEC would be smart to focus on enhanced services.

Intel: Yes, provided that feature parity exists. Ive encountered a number of these types of companies at service provider trade shows in the past few years, but have never noticed any progress.

aTelo: Unless a company has special communications needs/requirements (for example, enhanced security), it would make no sense for them to procure their own premises-based solution when an equivalently functional network-based service is available.

ComSol: Which specific applications will drive the acceptance of premises-based systems? Any standouts? How do you anticipate the situation evolving? A critical mass of bundled offerings? Or a gradual uptake, app by app?

Sonexis: Certain basic PBX functionality would be at the top of the list. In addition, there is voice mail, more sophisticated messaging (unified communications) applications, conferencing and collaboration (now on a very steep curve, thanks in part to September 11), wireless productivity applications (such as Blackberry), and integrated office productivity applications (such as call centers, advanced ASR and IVR applications, self-service telephony applications, and so forth).

I see the PBX vendors incorporating more and more of the basic horizontal applications into their PBXs. Voice mail is a good example of an app that used to be external to the PBX but is now included. But new applications will be coming fast and furious, especially those that integrate voice and the Web. These will be beyond the immediate capabilities of the PBX vendors who have to focus most of their resources on evolving normal PBX features.

In this category, we think collaboration is clearly the killer app; WebEx is a good example of whats happening in this area. Their revenues are doubling each year, and they have a $1.3 billion market cap on perhaps $100 million in sales. The audio conferencing vendors all seem to be enjoying 30 percent or better CAGR, and Wainhouse Research predicts that todays $3 billion conferencing market will grow to $11 billion by 2005.

The next-gen video conferencing products, such as Polycoms, are also starting to see real market acceptance. Over the next several years, youre going to see the virtual meeting displace a lot of business travel. Just about every software company I know uses WebEx for sales demos. Companies that used to fly executives into the home office for monthly (or weekly) meetings now do it on the phone, sometimes with Web conferencing to share documents. This is going to be as common, as simple, and as socially acceptable as making a phone call.

Our goal is to provide simple, easy to install, scalable appliances that allow companies to do for themselves what they used to have to buy from AT&T Conferencing, WebEx, and the like.

Intel: The specific applications we see will be those that offer an evolutionary migration path from current systems and business practices.

aTelo: We believe that a compelling IP Centrex application (which combines the most popular PBX features with e-mail, voice, and fax messaging) will certainly drive mass adoption of provider-based systems. As direct VoIP termination becomes available, and the price of IP handsets begins to compete with the price of analog handsets, we believe this service will take off.

ComSol: Would lower equipment costs undermine some of the economy-of-scale justification for development on the service provider side?

Intel: I dont think that this would significantly affect it. Im sure youve heard the statistic that 80 percent of a service providers costs relate to operations (with only 20 percent related to hardware). I believe a similar concept holds for enterprises, although I have never seen a statistic here. In addition, many of these applications are customer-facing and are therefore strategic to the customers business.

aTelo: Not really. Lower IP handset cost will accelerate the adoption of premises-based and network-based IP services equally. Lower IP-PBX hardware costs should not move the needle much because initial hardware expenditures are actually a small percentage of the overall/lifetime system cost. Much more significant is the cost of ongoing support/maintenance, having personnel trained on the hardware, maintaining an IT staff that is peripheral to the core business, dealing with service outages (that is, lack of five nines reliability), planning for obsolescence, dealing with sizing issues, etc.

If you would like to address any questions or comments to the author, please include them in an e-mail sent directly to the authors personal e-mail account, at kevinmayer2002@hotmail.com.

[ Return To The January 2002 Table Of Contents ]