Who Killed the VoIP Revolution?

A host of market and tech forces helped the communications world leap beyond the Web-phone idea

"VoIP is dead," Skype General Manager for Voice and Video Jonathan Christensen declared at an industry conference a few weeks ago. He spoke figuratively, of course, but he may well have been right. While proponents of Voice over Internet Protocol had long promised a decade of creative destruction, they themselves appear to have become the victims.

The full potential of a technology is not always realized once it converges with market forces. In this case, the gravitational pull of the incumbent local exchange carriers (ILECs) has always proved difficult to resist. Most of the VoIP industry, while loudly proclaiming the "session-initiation protocol" SIP era as the beginning of the end for monopoly communications, secretly courted the incumbents in hopes of profiting from replacing their long-amortized investments in the fixed-line business. By tying their fortunes to the whimsy of the ILECs, many of the upstarts suffered, destroying billions of dollars in shareholder value in the process.

Recently, PulverMedia, which spurred the VoIP crowd and rode its financial crest, shut its doors amid a swirl of controversy. As of this writing, Sonus Networks (SONS), once a highflier trading at 95 a share in 2000, goes for about 2.29. Even Cisco (CSCO) has thrown in the towel, discontinuing its BTS series of softswitches, which provide the routing logic for VoIP networks. These dismal stories perfectly mirror the ride of the VoIP industry in general.

Assault on Monopolies

The outlook was once a lot better. In 1999, with the ratification of the SIP specification by the IETF, advocates who wanted to tear apart the monopolies that dominated telecom started to beat the war drums. Following conventional wisdom that the Internet democratizes and deleverages any market into which it enters, they found it easy to persuade investors to pour billions into VoIP products and companies. Regulators seemed to support that theory, too, sealing the deal with the Federal Communication Commission's so-called "Pulver Order," which defended the VoIP industry from over-reaching regulation and tariffing.

The anticipated period of "creative destruction" came, all right. It began in 2001 with the smiting of the competitive local exchange carriers (CLECs) and long-distance competitors, who had not yet even had time to embrace VoIP, by predatory pricing from the incumbents. It continued with the shift from fixed voice lines to wireless phones, as evidenced by the drop in landlines. More recently, the guns have been turned toward the VoIP equipment vendors that begat the revolution in the first place.

So what happened? What clipped the wings of so many VoIP hopefuls can be boiled down to five things:

•Death by Deliberation: The incumbents and cablecos were identified as early targets for the equipment vendors. Their engineers, however, quibbled about curbside protocols and QOS and fiddled with VoIP in the labs, delaying launches by years—far outside the fund-raising cycle of most of the VoIP startups.

•Competition Attrition: The implosion and autopsy of WorldCom signaled to most of the industry that being a competitor in telecom is not healthy. Those high prices were largely arbitrary, and as soon as the market pressured incumbents to reduce them, they did.

•Evolution vs. Revolution: Companies like Nortel (NT), Siemens (SI), and Ericsson (ERIC) rank among the top VoIP equipment vendors today, not startups. Technologists underestimated the sway and leverage that the traditional vendors held over their customers.

•SIP in a Box: SIP might be an open protocol, but networks were built to be proprietary and have not been bridged together. Most telecom services still communicate with each other via public switching, meaning that the wonderful possibilities that SIP might enable are limited by the capabilities of the plain old telephone system.

•Landline Decline: Even as networks were evolving, the number of landlines around the globe was shrinking. People found more convenient ways to communicate via wireless, SMS, instant messaging, or pervasive e-mail.

VoIP technology has clearly been successful in making inroads into traditional telecom networks, but in doing so, the revolution that SIP in particular, and VoIP in general, enables has been largely cast aside, and the entire industry has coalesced in a race to the bottom. With this revolution went the volume of equipment and software sales that could have revitalized the supplier business and stimulated more innovation.

Of course, while the telecom industry was eating itself alive, a plucky little company from Luxembourg called Skype delivered on VoIP's promise by almost completely ignoring the Public Switched Telephone Network, not to mention the pundits and experts that cling desperately to SIP's potential. The point of Christensen's superpoke at what's left of the telecom business is that Skype has been successful because it threw away the playbook, ignoring the obsessions of so-called telecom experts and focusing instead on solving the practical needs of everyday users.

Tens of millions of people use Skype's network today for text messaging, file-sharing, videoconferencing—and, yes, voice calling. All these services are made decidedly more convenient because of presence—you can see who's there before you contact people and use that information to choose what the most appropriate means of communication should be. And with less than a $40 million investment (prior to eBay's rather more substantial buy-in), Skype's user growth has outpaced the entire rest of the consumer VoIP business combined.

The bottleneck for innovation appears to have been Alexander Graham Bell's (no relation) PSTN—the plain old telephone system. By going after low-hanging fruit and forcing their innovations to be defined within the walls of the PSTN, the vast majority of VoIP companies voluntarily muzzled their own revolution and ultimately cost their investors billions.

Before it's here, it's on the Bloomberg Terminal.