Convergence-with a twist

I'll admit it: I'm still a big Perry Mason fan. As I've become a more seasoned fan, I'm less interested in who the murderer is than in the "what-do-we-do-next" chats in Perry's office. I love to watch Perry, Paul and Della huddle in Perry's totally hip, black-and-white 1950s office to figure out their next move to save a hemmed-in client.

That part of the show is an amazing demonstration of an amazingly simple-yet often overlooked-fact of life: To get the right answer, you have to ask the right questions. The puzzles you think you're solving may not really be the right puzzles at all. It's those late night talks in Perry's office that can give away the trademark "twist" at the end-if you're listening closely enough.

I was watching Perry the other night when the proverbial light bulb went on about what's happening with convergence-that long-anticipated, often-discussed puzzle about how the worlds of data and voice will turn out.

In just the last few weeks, the nature of the puzzle has changed. Earlier this month, Compaq released ambitious details of how it will push to make the Internet smart enough to handle real-time traffic (such as voice and multimedia).

Within days, Qwest Communications, a pioneer in the IP backbone business, indicated that there may be limits to how much quality we can expect from an IP backbone by making a bid to buy LCI Communications, a voice-based long-distance reseller.

Compaq wants to make phones, personal digital assistants and servers smart enough to be PBXs. Compaq has a who's-who list of alliances to help, including Dialogic, Lucent and voice messaging innovator Wildfire Communications. One notable fact is that a key to Compaq's push is to use the old standby SS7 voice-switching protocols.

While Compaq is still very much dedicated to Microsoft's Windows NT, Compaq technicians say their servers can support SS7-based software applications. Not to be upstaged, Microsoft also said that it will further boost the next version of Windows NT to make it easier for software developers to support all sorts of streaming applications, including voice and multimedia.

And speaking of borrowing a page from the other side's playbook, look at Qwest. Qwest's long-term intentions in buying LCI are still a bit cloudy to me. It's especially intriguing given Qwest's recent $107 million pact with Cable & Wireless to provide that carrier access to its IP network for five years. It would seem that Qwest is taking its eye off the IP telephony ball. Or is it?

Here's where the "Perry factor" comes in. I would be willing to bet that the buyout could mean that Qwest execs think it is prudent to hedge their bets against IP-only services.

They want to be able to play both sides of the voice/data fence with any flavor of PC or phone technology a business is willing to use.

Setting aside the technology parts of the merger (should it go through), take a look at the economics of it. One of the big appeals of Internet telephony is low-cost long-distance. Well, LCI can already offer rates as low as 6 cents a minute, and that's without a lot of expensive IP retrofitting or fancy voice packetizing, the technology used to convert a phone-originated voice call to traffic the Internet can handle.

Adding up the events of the last few days, I've come to think that convergence will be a lot more messy than we might have expected just a year ago.

Convergence is no longer the domain of start-up Internet companies. Computer and telecom companies with nationwide reputations are now taking their gloves off. And they're not out to change the world like some Internet hard-cores. They're out to build multibillion-dollar businesses as safely and securely-and as quickly-as they can.

That means convergence is not a simple us-vs.-them puzzle anymore.

Convergence over the next year will come to describe not just a convergence of technologies, but a convergence of interests.

Want to know who the winners and losers will probably be? Ask the right questions. Remake your puzzle. And, if that fails, call Paul Drake.