For the past seven or eight years, every conference relating to telecoms or networking has had at least one presentation on "convergence". It was the miracle ingredient that said: "This conference/seminar/symposium really understands hi-tech stuff."
And usually we all nodded knowingly, took down copious notes of slides, displaying astounding projections that convergence would bring many trillion dollars/bytes/users/connections/market share. And then we'd go back to our day jobs, and continue doing what we'd been doing all along - trying to keep the databases working.
Almost a decade since the hype started, convergence is still magic ingredient X. And it's still everything, and nothing. Unfortunately, it's also having a confusing "deer in the headlights" effect on people, all trying to work out exactly what part of "convergence" they're meant to be doing something about.
Convergence means different things to different people. To network service providers it could mean a low-latency, multiprotocol network. To a hardware provider it could mean PCs with onboard video cam, mic and speakers. To a networking vendor it could mean switches supporting Quality of Service. The basic concept, putting voice, video and data together on a single infrastructure is valid, but not useful for anything except slapping an extra zero on a cost estimate.
Consider the duck. Waterfowl. Family anatidae. It walks, it flies, it swims, some even whistle. It does all of these things, but none particularly well. The same applies to 2-in-1 shampoo and conditioner. Or Swiss army knives. Or SUVs. Basically, anything where you take a number of functions, and try to put them all together into a single thing that does none exceptionally well.
That is why I have a problem with 'convergence'. The most typical definitions suggest that you're taking voice, video and data, and combining them onto a single network. You're not. You're taking a data network, and adding more services to it.
This may seem like a trivial distinction, but it can be summed up in four words... "It's the data, stupid".
You have a network pipe, you design it to have a certain performance characteristics, and then you cram data down it. Whether its e-mail, database queries, packetised voice or highly compressed streamed video, it's all data... requiring different classes of service, of course.
The goalposts have moved in the last two to three years. Where voice was once a real challenge to place on a data network, the shift towards gigabit speeds in network devices and dropping bandwidth costs means that latency, jitter and packet loss are becoming less of a problem all the time. While the consultants pontificated, every Tom, Dick and Harriett was busy streaming YouTube videos and Web cam chats over the Internet. It's still small potatoes, tip-of-the-iceberg stuff, but users are driving voice and video adoption on the network and learning to like it, while many IT departments scramble to cope.
In South Africa, we're lagging our first world cousins in integrating video and voice with our technology. The end-game of bringing together voice, video and data is to drive competitive efficiency or provide new services. There are several milestones along the way, but we're still near the beginning, somewhere between "Least Cost Routing/cost savers" and "reduced Total Cost of Ownership" (unifying cabling, PBX and switching).
Most South African companies have not even started thinking about the next step - "Productivity Improvement", never mind the destination, "Competitive Advantage". Some are getting there, using VOIP to allow call centre agents to transfer calls with "screen pops" so that they don't annoy the daylights out of the customer by asking for the same information again and again ("account number, please..."). Calls are starting to find you wherever you are. But it's still fairly primitive.
The insidious damage that the obsession with "convergence" has done is most apparent in IP telephony strategy. The simple truth is that the world has been invaded by IP. The only number or address that will really matter will be your IP address. Your phone calls, your IM calls, your video conferences, your e-mail - no matter what the numbering or addressing scheme, it will all ultimately resolve to an IP data address. Stop worrying about "convergence", and start worrying about how to build a data network that allows services to be switched to any device in a way that provides competitive advantage. "Click to talk" on your Web site? Sure, and get it to your customer support people in the same way as your incoming PBX landline calls, even if customer support staff are working from home. Convergence is a software application development issue, not a telecoms issue.
Because people are distracted by "merging their voice and data networks", they're losing sight of the big picture. Device mobility and secure "access from anywhere" broadband means that a person is (or will, hopefully, before long) always be at the end of a fast data pipe. By worrying about whether an IP telephone on his desk is important, you lose sight of the fact that his "phone" may soon be in his PDA, on his desktop, and at his laptop at home, all running VOIP.
Stop thinking about "convergence", and start thinking about how to take voice and video streams and send them to the people that matter so that they can do their jobs better. Voice connection? Video image? Customer payment record? Sales priority list? Call centre history? Product information sheet? How can they be linked to let someone do their job more effectively?
Because the longer you spend going around in circles obsessing about "convergence", the less time you'll spend thinking about what you're going to do with it when it arrives - and in many ways it already has. The way that networking technology is developing, sticking voice and video on a data connection is not anything special, it'll be the norm. It's not whether you have "convergence" that's important; it's what you do with it.
Share