Corporate View

Voice And Data: Together At Last?

MIS manager: "So, what you're saying is that they're not integrated."

Network manager: "Well, they are in that the voice and data ride on the same T1 from each of our sites."

MIS manager: "But on different channels and from there they go to different networks. Wouldn't we save money if we could combine the traffic and the voice could ride for free?"

Network manager: "But really, it's not that simple....."

And so on and so on.

I get the impression that this conversation is being repeated at many corporate sites. The hype for integrated voice and data is everywhere. But in practice, it's a very different story where the reality falls short of the hype. Even integration in good-old ISDN never meant voice and data together.

Why all the fuss? It's simple: Managers are looking for the least-expensive way to handle voice calls. They watch cash fly out the door every month to the carriers and they hear things like "free calls" or "as little as a half cent per minute," and the promise to consolidate their networks and gain the advantage of fewer WAN links. You can't blame them for being interested and for trying.

Desires and Assumptions Now that you know their desires, it's time to find out about their assumptions. MIS managers assume that these new services cost the same to install and administer and have the same reliability as phone services. However, the reliability and administration costs of phone systems are anywhere from difficult to impossible for PC-LAN systems to beat.

On the technology side, the voice-over-the-Internet folks have viable products that are selling well. The frame relay vendors have released a second generation of the technology based on interoperable standards. Probably the most exciting thing in the field is the introduction of products such as the VocalTec Internet Phone Telephony Gateway Server, which enables the integration of legacy handheld voice PDAs--telephones to you and me.

I believe the long-term advantage lies with the Internet-based solutions, basically because they don't dictate a solution. Unlike the frame relay solution, they don't require any particular type of physical or transport network. Any network, given available bandwidth, that can carry IP will do. Also, since the desktop is responsible for compression, silence detection and playback, the routers, bridges and other network devices in your network remain as is--black boxes as far as voice and data integration is concerned. Conversely, frame relay solutions require dedicated multipurpose hardware to integrate and route voice, LAN segments and serial lines.

However, whatever technology you choose, you can be certain th at it will require more management and troubleshooting than the equivalent separate voice and data networks.

Hand in Hand Reliability and cost are interrelated. You know in your gut that the most stable networked PC is far behind the reliability of the conventional phone.

Think of this: You don't even know who maintains your phone--that's how reliable it is. Compare that with LANs where users have the helpdesk hotline number posted above their monitors. The new technology behind the Internet phone and voice over frame relay is great, but neither addresses the cost of maintaining desktops. Whatever happened to the Desktop Management Task Force (DMTF)?

The phrase to watch out for is "no extra charge," since we all know there's no such thing as a free lunch. If you crunch the numbers in a simplistic fashion, it looks like the savings in ongoing long-distance costs would pay for capital investments and higher support costs. You could start saving money in as little as a year. But even the most ambitious adoption plans would not discard the PBX and other voice equipment. That equipment and associated wiring will stay in place and continue to depreciate.

In addition, you have to consider the cost to install and maintain new or additional desktop components. Installing multimedia capabilities will cost you. Most corporations haven't been installing Soundblaster cards in every PC, so this is a new net cost to the desktop, not simply an exploitation of an investment in place. Finally, the cost in time and technical support for maintaining an Internet connection and downloading new drivers and plug-ins is not the same as supporting just another telephone on your desk.

A Call Is Just a Call What about functionality? A call is just a call, right? This type of comment drives telecom managers straight up the wall. Right now you're probably reading this at a desk that has a telephone with some combination of ACD/voice mail/call pickup, for instance. It may be boring and decidedly unsexy, but i t will take a lot of code and integration to make PC-based voice systems provide the same level of functionality as the off-the-rack PBX.

I'm not pessimistic--I'm just realistic. Voice over frame relay or IP will seem most attractive to you if you have underutilized local-loop T1s. Since local-loop fractional T1s are not usually offered, you might have local T1 service even if your data rate is well below T1 capacity. The service should be targeted at remote offices of moderate size. Keep the traffic on the periphery of the network and stay away from using this on your backbone.

Although there is little doubt that integrated voice and data, most likely voice over IP, will play an important and increasing role over time, there is also little doubt that it will be an important player in the tag-team match in which the local exchange carriers, competitive access providers, regional Bells and long-distance carriers will duke it out in the "all services" market. In all likelihood, the smaller resellers of long distance (you know, the ones that have filed protests with the FCC) will wind up as the losers. Either way, though, you will end up the winner.

Brian Walsh is a senior consultant with Cap Gemini America in Portland, Ore. He can be reached at bwalsh@nwc.com.