Who Owns the Internet?

Ted Lewis

for Wired Wired World

The Battle Royale

While everyone’s attention is glued to the Microsoft v. DOJ contest, the really important contest is taking place amongst the teleco, telecommunications, and software giants. Microsoft’s desktop revenues are small potatoes compared with potential big bucks from the Big Network - the Internet. Like the range wars between cattlemen and sheepmen of the 1880s, the range war between telecommunications giants is over real-estate. Only in 1998, real-estate is virtual-estate, because battle lines are being drawn in the virtual space of the Internet.

With most IP data traveling over long-haul lines, the Internet is actually the telephone companies. But who owns the telephone companies? Because of the Telecommunications Act of 1996, ownership of the Internet is up for grabs. Vernon Keenan of Zona Research, Redwood City, CA., doesn’t mince words, "It’s a battle royal between them [Cisco, Bay, Ericsson, etc.] and traditional telephone equipment providers such as Northern Telecom and Lucent for billions of dollars worth of business as service providers and metropolitan areas upgrade their telephone systems." [Tom Quinlan and Jon Healey, "Cisco to gain ammo in voice/data war," San Jose Mercury News, Dec. 23, 1997, pp. 10C]. According to Quinlan and Healey, computer network companies like Cisco are positioning themselves for dominance in the new economy, because mixing their network equipment with traditional telephone switching equipment, "...will enable Cisco to work with existing telephony switching equipment while it seeks to ultimately replace it."

Cisco isn’t the only contender. Microsoft may be getting a lot of press attention because of its predatory pricing practices on the desktop, but the controversial company has something even more diabolical up its sleeve. As the plot thickens, you will see what I mean.

The New Robber Barons

The Internet used to be a collection of UNIX servers connected via T1 lines to college campuses that served the role of mini-ISPs. That was then. Now, the Internet is rapidly being privatized just like the railroads in the 1850-90s. No longer populated by the hippie nation, the Internet is as three-piece suit as J. P. Morgan, Edward C. Harriman, James J. Hill, and the Vanderbilts were over one hundred years ago. Remember them? They were to so-called railroad baron robbers.

Enter the new robber barons. Everything is going IP and our rates are going up, but the Internet is undergoing a more dramatic transformation that digitization and rate increases. The Telecommunications Act of 1996 has put traditional telephone companies in danger of going extinct. Upstart companies like Intel, Cisco, Bay Networks, and Microsoft are competing against the traditional telecommunications companies like AT&T, Nortel, and Lucent, for ownership of the $1 trillion global industry.

The mega-billion dollar merger of Worldcom and MCI, and similar smaller mergers going on around the globe don’t reveal the fact that AT&T’s grip on telephony has slipped from a monopoly to nearly 52% market share. And the slippage is far from done. Within a few decades, AT&T won’t exist, or else it will exist as an entirely different company. Why? In the modern telecommunications industry you are either a digital networking company or you aren’t even a player. This implies a rapid rebuilding of the out-moded infrastructure, which in turn implies lots of cash. And the old guard doesn’t have it. The new guard does. Companies like Cisco and Microsoft have the cash as well as the technology needed to replace analog with digital hardware and software. If they move fast enough, they can put themselves in the cat bird’s seat. And by the time the Justice Department wakes up, Cisco and/or Microsoft may have monopolized the Internet.

One reason to believe that the new robber baron’s will be successful is that we are living in a new era - an era of the friction-free economy. The new economy means rapid replacement of out-moded business models. Even the politically liberal governments of Germany and England are capitulating to friction-free capitalism. They too are deregulating their telecom monopolies out of fear that the USA giants will soon completely dominate the global Internet. But, it may be too late as powerful US computer companies continue to buy up the main hardware and software lines connecting countries and continents to the emergent private Internet. It is a repeat of the railroad robber baron story. Most telecom companies throughout the world have reacted too slowly and too unenthusiastically to IP. This sluggishness will be their ruin. But then, lets not get sidetracked by politics.

My Own Private Net

The global struggle will be waged by big players, but not without assistance from the small potato players. High profile companies like Cisco, Microsoft, TCI, and Intel, will provide the big bucks, but small companies like LightSpeed (bought by Cisco in early 1998), Starvox Inc., and DGM&S, will provide the innovation. The big fish will eat the little fish to accelerate their expansion. Otherwise, they won’t be able to expand rapidly enough to beat their competitors. Who are these minnows?

Take TeleHub Communications Corp. [www.telehub.com] as an example of how a small-fry hardware infrastructure company is providing the innovation. Tiny two-year old Walnut Creek CA TeleHub has built a coast-to-coast ATM network for carrying voice IP, provide local number portability, and other services which should appeal to local markets. TeleHub uses Cisco-Stratacom ATM switches, but operates them over Signaling System 7 (SS7) services to smoothly integrate their digital IP into existing analog infrastructure. SS7 is the protocol of the traditional (AT&T) telephone ESS infrastructure. TeleHub’s strategy is to usurp the ESS system, replacing analog voice with IP voice, and eventually displace the long-haul carriers, altogether! Thus, TeleHub is positioning itself to swoop in and provide access to LECs (Local Exchange Carriers) in place of AT&T, MCI, and Sprint. And because TeleHub runs billing on a Sun web server system, it can do real-time billing cheaper, faster, and before the big telecom companies can react.

Innovative competitors like TeleHub are cropping up all over the Internet. VIP Calling, Inc. [www.vipcalling.com] recently unveiled its US-to-Hong Kong Internet voice service, which sells to carriers and resellers, including pre-paid calling card companies. Inter-Tel, Inc. [www.inter-tel.com] launched a commercial service in seven US cities in September 1997. Global Gateway Group is offering calling services throughout the US, Japan, and Europe. InterNex Information Services Inc., Santa Clara, CA. runs a Sonet network in six US cities and four international cities. Exodus Communications is building a national network over ATM and 45 Mbps lines and runs data centers for major Silicon Valley companies. Internet Systems Inc., Sunnyvale, CA. does the same thing for Netscape Communications, Yahoo!, and others. Little-known backbone provider GST Internet Inc. is spending $250 million to string fiber across the Pacific Ocean connecting Hawaii and Asia to the US. GST wants to replace existing voice telephony with IP networking that will extend high-speed Internet to Asia.

"These companies could ultimately become the data communications equivalent of the competitive LECs," said Gordon Werner, vice president of NetEdge Systems, Research Triangle Park, NC. "A lot of these companies did not exist six months ago, but they are a whole new data market." [John Mulqueen, "Exodus to network centers," Communications Week, Apr. 21, 1997, pp. 69.]. In other words, Internet hardware infrastructure is being rebuilt by small potatoes looking for downstream big bucks sometime in the distant future. They are doing this with computer technology - not traditional telecommunications technology. Their business model is to create a local monopoly on a niche, and then sell out to Cisco, Microsoft, or Intel. But, they aren’t the only ones who have heard about the Telecommunications Act of 1996.

Weeding Out the Middle

It is not surprising that massive disintermediation appears to be going on as a result of flattening of the established telecommunications industry caused by IP networking. Disintermediation is a polite word for "elimination of the middleman." With companies like PSINet "going around the long-lines" and young upstarts re-inventing telephony in the image of ethernet, it would seem like the middleman is doomed. PSINet, for example, recently announced an aggressive Internet-based telephony strategy to double the company’s revenue every year for the next three years through acquisitions and a $310 million investment in its own Internet backbone. PSINet wants to eliminate the middlemen LECs - moving customers directly onto its tier one super-ISP network. PSINet will acquire ISPs, add 20 POPs (Points of Presence), spend $100 million on fiber, and another $120 million on upgrades. This kind of integration is reminiscent of AT&T’s integration of local exchanges during its buy-up of the telephone network circa 1898 - one hundred years ago. Is history repeating itself? Is the LEC and traditional long-lines telephone company as obsolete as MS-DOS?

The answer is no. The middleman is not being weeded out. Instead, a new middleman is emerging in the form of new intermediaries - integrated ISPs like PSINet and Worldcom, and the rebel Cable TV operators like TCI, Time Warner, and Comcast. PSINet is bidding to become a LEC as well as filling its traditional role as an ISP. In other words, telephone companies and ISPs are merging - both in terms of their technology and business models. And Cable operators are making their move into this space, too. TCI, Time Warner, and a number of smaller operators agreed to buy at least 15 million digital set-top boxes from NextLevel Systems [ www.nlvl.com a.k.a. General Instruments ] between 1999 and 2002. These boxes will feature high speed IP for Internet access, digital video, e-mail, and games. Microsoft, Microware Systems, PowerTV, and NCI (Network Computer, Inc.) will compete for the operating systems business. Does Microsoft’s $1 billion investment in Comcast ring a familiar chime? The battle for the middle is already raging.

The New Value Chain

Clearly, Cisco Systems has to be one of the upstarts beginning its rise on DOJ’s radar screen, because it stands a good chance of dominating the lowest link in the value chain extending from IP router to Web application software. Ten thousand dollars invested in Cisco stock in 1990 would be worth $500,000, today. Its capitalization mushroomed from $4 billion in 1995 to $43 billion in late 1997. Sales were $382 million in 1992. Sales in 1997 exceeded $6 billion, putting it on a revenue growth ramp that vanquishes the records set by Microsoft and Intel. "The networking segment of the industry is enjoying the biggest growth rate in the history of high technology," says Selby Wellman, senior vice president of Cisco. [Bart Nagel, "Routing the competition," The Red Herring, March 1997, pp. 82-88]. Cisco does it by eating its competitors. They would rather buy than fight. Through a series of acquisitions, Cisco has grown from backwater router manufacturer to top dog in the telecommunications game.

A level upward (operating systems) in the new value chain is where it gets interesting. As I pointed out at the beginning of this rant, Microsoft’s tumble with DOJ over its Internet Explorer browser is small potatoes compared to what is going on behind the scenes, here. "While the US DOJ is busy fretting over Microsoft’s domination of the browser business, the software giant is quietly starting to eat into an arena that is much larger - the worldwide telecommunications market ... which is estimated to be close to $1 trillion for 1997, according to Multimedia Telecommunications Association." [Margie Semilof, "Eye on telco prize," Computer Reseller News, Dec. 22, 1997, pp. 5, 152]. Distracted by Microsoft’s bundling of IE with Windows 95, DOJ has missed recent big moves by Microsoft to dominate this new middleman segment. Soon, with help from Navitel Communications, Microsoft will start adding Web phone functions to Windows CE. Wireless will follow. Windows NT will get voice response capability and become even more of a threat to UNIX as a replacement for carrier switches. Microsoft’s new Telecommunications Solutions unit has two main goals: put Microsoft software on all telephone company switches (eliminate ESS and PBX competitors), and saturate the local loop (LECs) with more Microsoft software. Microsoft is partnering with DGM&S Telecom of Mt. Laurel, NJ. to christen Windows NT with SS7 compatibility. Clearly, Microsoft sees a new source of revenue in the highly lucrative middleman value chain. The software powerhouse has its eye on a bigger food chain than the dwindling desktop market, see Figure 1. "Anyone that isn’t nervous about what Microsoft is doing is crazy," says Scott Wharton, product manager of competitor VocalTec.

Forecasting winners at the next level also gets interesting. No major player is really going after applications, but Sun could position itself here with its Enterprise Computing Platform. Java may be in limbo on the desktop, but Java and JavaBeans with IIOP (Internet Inter-ORB Protocol) could be king of the emerging Internet middleman market, if Sun, Netscape, or IBM were to become telecommunications savvy like Microsoft. This, however, remains in doubt, as they are stuck in a computer industry mindset. Blinded by years of product development on computers instead of networks, they may wake up to the reality of the telecommunications mega-economy just in time to watch as DOJ goes after Cisco or Microsoft for monopolizing the Internet. But then, this is what makes the telecommunications industry even more exciting than the computer industry.

Ted Lewis is co-author (with Hesham El-Rewini) of "Distributed and Parallel Computing", Prentice-Hall, 1997, and author of "The Friction-Free Economy", HarperCollins, 1997.

 

Figure 1. Network spending as a percentage of total IT spending. Desktop spending is declining as network hardware and software spending is accelerating. Source: Based on data published in Computer Reseller News, Dec. 22. 1997, pp. 151.