September 1999

The Impact Of The Internet — Moving Beyond POTS

BY BROUGH TURNER

IP telephony will win in the end. That's been my view since March 1996 (when Intel and Microsoft first agreed on H.323 for the desktop). I've written on the subject on several occasions since then, most recently in my February 1999 CTI� column. Today, the idea is no longer controversial. Voice over IP (VoIP) is generally accepted, even among old-line telecom companies. The remaining arguments are about how soon it will happen.

But even VoIP advocates still don't get it. Today's focus is on recreating the quality, features, and functions of traditional telephone service, only using IP. Recreating plain old telephone service (POTS)? This is nuts! I'll explain what's really coming, but first we need a little background on technical revolutions.

STUDYING THE RECENT PAST
Clayton Christensen has written a fascinating book, The Innovator’s Dilemma — Why New Technologies Cause Great Firms to Fail (1997). Christensen, a Harvard Business School professor, studied technology and business changes in various high- and low-tech industries over time spans of several decades. He concludes there’s a difference between sustaining innovations, which improve product performance or features for an existing market space, and disruptive innovations, which create new markets and new business models. In his book, Christensen discusses several of the industries he studied in detail, including fixed disk drives for computers and excavation equipment for the construction industry. His case studies provide insight into what’s in store for the telecommunications industry.

Fixed Disk Drives
In the mid-70s, disk drive vendors were selling 14-inch disk drives to the mainframe industry. They were listening to their customers, responding to customer requests and advancing their technology. By 1980, several new entrants had arrived on the scene with 8-inch disk drives. The new disks were of lower capacity and lower performance, but they appealed to a new market — mini-computers.

The new disk drives weren’t perceived as a threat to the 14-inch disk drive market. But the mini-computer market was a much higher volume market than the mainframe market. With growth came the ability to improve the technology at a faster rate, to the point where the 8-inch drives were able to exceed the capacity and performance of the 14-inch drives.

Within a few years, all of the independent 14-inch disk drive vendors were out of business, or at least out of the disk drive business. And IBM, which remained in the disk drive business, did so only through an autonomous organization elsewhere within their corporation.

This dramatic industry restructuring was not a unique event. In disk drives, such upheavals occurred again and again. For example, 5�-inch disk drives got adopted for PCs, and then 3�-inch and smaller drives got adopted in portable applications.

At each transition point, the new technology was not a breakthrough for the old customers and their established applications. Each new disk drive form factor started out as lower cost and smaller, but also with lower performance. It was not viewed as an immediate threat to the existing players in their existing markets — their customers were happy. Their customers were not asking for smaller and lower performance disk drives. It was a new customer base that found the small size advantageous.

At each transition, the majority of the companies from the earlier era failed to make the transition. When Seagate (a leader in 5�-inch disk drives) started to make 3�-inch disk drives, the company equipped them with 5�-inch adapters and sold them to its existing desktop computer customers rather than focusing on the opportunity in portable and laptop computers. Seagate pulled through, but even four years after they finally started shipping 3�-inch drives in volume, they had failed to sell any drives into the new portable computer market.

Excavating Equipment
Another industry that Christensen studied was mechanical excavators. This industry took 20 years or so to go through one major transition, but that transition was just as disruptive as any of the transitions in the disk drive industry; that is, most of the original manufacturers went out of business. The innovation that caused this disruption was hydraulics.

In 1950, most mechanical excavating equipment used cables to control and lift the bucket. But the first hydraulic excavators had already appeared on the scene. The new machines (also called backhoes) were too low performance (small capacity, small reach) for use in general excavation, but they were inexpensive enough to appeal to residential contractors — contractors who formerly had to hand dig small trenches. For example, contractors digging narrow utility trenches in new housing tracts found the small buckets on hydraulic equipment very useful. The new equipment was affordable and could fit into narrow spaces.

As the use of hydraulically controlled buckets in small-scale excavation projects grew, volumes grew and the rate of improvement in the hydraulics technology surged. By 1970, hydraulics performance had overtaken traditional cable technology for even large machines. And, by the mid-70s, only four of the original thirty or so cable-operated equipment vendors remained in business.

LESSONS LEARNED
These disruptions came about because the new technology happened to match the needs of a new application or new customer base. The first lesson is you can’t only listen to your customers. Your existing customers are a great source of information for their current application needs. But they are unlikely to anticipate new applications or opportunities in other markets.

The second lesson is that new things start small. If a business is structured to sell into billion-dollar-per-year opportunities, it may have a hard time going after new opportunities that are only worth a few million dollars per year in sales at first. But that’s how disruptions start.

The datacom industry may be paying attention and attempting to address these issues through acquisitions. Cisco, for example, does its best to acquire a wide range of new technologies, in order to have all possible innovations in its portfolio.
Why have I included a “review” of Christensen’s book in my column? Because “disruptive innovation,” a term coined by Christensen, is an important concept for anyone dealing with Internet-based business or technology change — we in the telecom industry, for example.

VOICE OVER IP AS SUSTAINING INNOVATION
A policy paper prepared for the Canadian Federal Department of Industry by T. M. Denton Consultants, with Francois Menard and David Isenberg, entitled "Netheads versus Bellheads: Research into Emerging Policy Issues in the Development and Deployment of Internet Protocols," is very relevant. It points out that much of the VoIP protocol development (in ETSI’s TIPHON group, in the ITU, and even in the IETF’s Megaco group) is dominated by traditional “Bellheads” who are trying to recreate the structures of the existing telephone network in a slightly altered form. In other words, the Bellheads in these standards groups regard IP networks as a sustaining innovation, one that can be used to provide POTS service at a lower cost while (eventually) matching the performance of the existing telephone network.

Countering this view, Denton, Menard, and Isenberg claim that SIP should win out over MGCP, and that intelligence should move to the edges of the network. They worry, however, that politics could muddy the waters. I am less worried about politics because I am confident the Internet represents disruptive innovation.

DISRUPTIVE INNOVATIONS IN THE INTERNET
The Internet has already proven its ability to be disruptive to most of what it has touched. The advent of e-mail has fundamentally changed business communications — certainly in the high-tech industry, but increasingly throughout the world. In earlier days, I seldom wrote formal business letters, but today I check my e-mail, even before I check my voice mail (potential correspondents take note!).

Similarly, the Web has changed the way people get information. At least in the high-tech industry, every company has a Web site, and Web sites are the primary source of product information. For some companies, Amazon.com or Cisco, for example, the Web is already a primary sales channel. I credit David Isenberg with first pointing out to me that, given the Internet’s history of disruptive change, the telephone industry needs to watch out! (Among other things, David is famous for his June 1997 article The Rise of the Stupid Network.)

How could anyone think that the Internet will be just a new way to provide traditional telephone service?

PREDICTIONS FOR THE FUTURE
Within three to five years, we’ll see several dramatic new telecom business models and at least one completely new form of telephone service. Telephone sets with wide-band audio performance will appear next year. While they will interoperate with narrow-band POTS telephones, they will provide high-fidelity telephone service within their user groups, within an enterprise IP-PBX, for example. High-fidelity telephone service will then creep outward as more of the new telephone sets get deployed.

Differential services will emerge on the public Internet in the next 12 to 18 months, but only for services entirely carried by one ISP. This means you will be able to get preferential low latency service for your telephone traffic over the Internet at least between two corporate sites served by the same backbone ISP. Two, or at most three, grades of service should handle all needs, but we can expect diverse service offerings, at least as complicated as today’s mobile phone service offerings. It will take more than another year for inter-ISP relations to settle out so you can purchase preferential low-latency service that work across two or more ISPs. But once that happens, IP-only phone service will go wherever there are new IP telephones.

Interestingly enough, if the premium traffic is less than 10 percent of the total traffic, the cost of providing a two-tier differential service is negligible once “diffserv-enabled” routers and policy servers have been deployed. At the rate data traffic is growing, voice traffic — traffic that might want premium IP service — will be a relatively small portion of an enterprise’s total traffic. So the incremental cost of VoIP should be quite small. The only time you will have to pay by the minute is when you want to reach old-style telephones within the legacy telephone network. Here you will need to use a gateway controlled by Bellhead protocols based on traditional telephone network design principles.

Communications over the Internet are going to sound much better than what we are used to, even before we add video and new user-defined services. These are the real reasons the PSTN will eventually disappear. The driving force will not be the ability to more cheaply provide “toll-quality” speech, otherwise known as 19th century communications quality. The real innovation will be 21st century quality and features. The Internet will finally let us break out of a telephone service model that has seen little change in a century. The exact path should become clear in the next five years. Hang on for a very exciting ride.

Brough Turner is senior vice president of technology at Natural MicroSystems, a leading provider of hardware and software technologies for developers of high-value telecommunications solutions. For more information, call Natural MicroSystems at 508-620-9300, or visit the company’s Web site at www.nmss.com. E-mail to the author is also welcome.