Five Steps To Take Before Implementing VoIP

There's something unfair about success and failure in enterprise voice over IP (VoIP). When the system works it's business as usual. After all, there are few things simpler than picking up the phone to make a call, and all of those IP features and applications can help make the flow of business move effortlessly.

When it doesn't work, however, enterprise VoIP can be a disaster. When something as simple, basic and essential as the phone system stops working, business grinds to a halt. There are few things easier than making a phone call and conversely, there are few things more frustrating than picking up the telephone receiver and hearing no dialtone.

"Voice over IP (VoIP) can be very successful in the enterprise," says George Goodall, research analyst at the Info-Tech Research Group. "On the other hand, it can also be a massive failure." The key, Goodall says, is to step very carefully and make sure that everything is in order before trashing that old private branch exchange (PBX) and going to IP.

"Most of the problems with VoIP implementation can be caught before you implement it," Goodall says. "You can save a lot of trouble if you proceed carefully."

So what should you do? Listen to advice from the experts, and follow these five steps you should take before moving to a VoIP system.Make the business case As scandalous as this might sound, not every enterprise actually needs to move to a VoIP system, Goodall says. "The first question that enterprises don't ask is 'do I need VoIP?" he says. "Do you have a PBX system that works? How close is it to the end of its life? Those are questions that people don't value. But they should, because it can determine the benefits of a VoIP migration."

Those benefits, and not the amorphous promises of technological novelty, are what make the business case, says Avaya director of communications applications Lawrence Byrd. "It's always easier to start with the business case of what you're trying to accomplish," he says. "Do you want the kind of features you have now with a PBX, but with VoIP, or do you want something different?"

Moreover, Byrd says that you should have a clear idea of what processes and activities the new communications features are going to enable. Above all, VoIP deployments should be about more than cost savings. "Originally, it was all about converging and reducing networking costs," he says. "But of that's all you're trying to do, then you'll fail to realize the real benefits."

Have reasonable expectations One part of making the business case for VoIP is to have reasonable expectations for the new system. Success and failure, after all, depend to a great extent on how you define success.

"One company I know about wanted to use VoIP to do toll bypass using its home-office workers' DSL lines," Goodall says. "What they didn't factor in was the data demand spike at 3:30 pm every day when the kids came home and started using their PlayStations. So people switched over to their traditional telephones, charging the company for their long distance at full price."Indeed, Goodall is quick to point out that the real killer apps of VoIP are not cost savings or even network simplification, but things like messaging, voice mail and call forwarding --- features usually considered only as added value to network convergence. "VoIP isn't going to make your business," he says. "But it can make things easier. The key is to know what those things are."

Test your network All the benefits of VoIP will mean nothing, of course, if your data network collapses under the weight of the voice traffic. Just about any IP network can, in theory, run VoIP. However, almost none are properly configured for voice traffic right out of the box, Byrd says.

"Typically, it's very easy to run one or two VoIP calls over a network," he says. "When you start running hundreds or thousands of calls on a network that can't support them, that's when you have problems."

The mission-critical nature of voice communications --- almost no one can do business without phones --- makes this doubly critical. "Any time you drop bandwidth-hogging devices like IP phones into a network, it becomes an issue," Goodall says. "You don't want that happening on a mission-critical basis."

Consequently, enterprises must put the time and resources aside to evaluate network weaknesses and, more importantly according to Byrd, identify places where the network needs to be upgraded before going ahead with the implementation. VoIP success, he says, is in the planning.

Have a plan It makes sense, then, to have a plan. One approach, which is hotly debated in the VoIP market, is to rip out your PBX, phones and wiring and replace everything with IP phones. That might work in same cases, particularly where the old PBX is long past its expiration date, but it isn't always the best approach."It's often a good idea to pilot or start small," Goodall says. "You should look at where you really need to implement VoIP, whether it's a PBX overlay or at a branch facility. As with any major project, it's often a good idea to start small and plan from there. If you have to pull the plug for any reason, you still have that option."

The plan should also consider the human factor, since VoIP is as much about business processes and the people tasked to undertake the processes, as it is about technology. "Within the organization, you definitely need some kind of driving group from the business side as well as from the technology side," Byrd says. The bottom line, he says, is that, once you have identified the goals of the implementation, you need to plan each step to achieve those goals, and have someone or some body delegated to ensure that you stay on track.

Choose the right vendor(s) "The problem is that there is such a cornucopia of options." Goodall says. "So you need to know what you need a vendor for. You have to be sure that the VoIP system talks to your data infrastructure and that whatever vendor you go with can support your handsets."

Part of that can be addressed in the formal RFP process, but given the scale and complexity of the implementation, enterprises have to approach vendor selection with very fine granularity. "Application functionality and value are important," Byrd says. Services are important. Support is important. You have to evaluate how reliable a vendor's solutions will be in the worst case."

Above all, you have to choose a vendor in terms of how its technology and features work for you. "The big question is 'will it change the way I interact with customers or the way I do business?'" Byrd says. "If it will, are these changes for the better?"0