Today there are multiple ways of carrying voice over data networks. It can be challenging for business people to understand how those differences matter from a business perspective. This month, we'll focus on what you need to know from a voice transmission technology perspective to eliminate potentially substantial and unexpected costs down the road.
Early adopters of voice over IP (VoIP) technology purchased solutions based on the H.323 Protocol. Today's VoIP buyers tend to favor solutions that are based on the SIP (Session Initiation Protocol) standard.
While SIP is rapidly displacing the H.323 Protocol as the VoIP standard of choice, the issue for many companies is that they (or the companies they may acquire) already have legacy investments in H.323-based gateways, IP-PBXs and protocol-specific handsets. The challenge these companies face is how to enable anyone to talk to anyone, regardless of protocol. The answer that most vendors put forward is a complete replacement of legacy equipment. Some vendors, however, have taken the customer-centric approach of building a 'protocol bridge' that empowers interoperability between these previously incompatible standards; the result is co-existence in the network with no forklift of back-end equipment or desktop handsets required.
Another consideration: Even if you don't have legacy investments to worry about today, it might be fair to say that today's investments will be tomorrow's legacy infrastructure. Selecting vendors that are committed to building bridges to older protocols on an ongoing basis (as opposed to those with a forklift mentality) is obviously preferred from the perspective of managing obsolescence risks.
RTP Mixing: Reducing The Cost Of Call Center Interactions
Many VoIP solutions will distribute calls to endpoints (IP phones or softphones) without adding new value to those calls. These IP-based solutions don't have an 'out-of-the-box' capability to manipulate the RTP (Real-time Transport Protocol) streams of calls in order to natively add the value-added capabilities that call centers require. (The transport management in an IP-based call, whether SIP or H.323, is typically an RTP stream.) As a result, such vendors will need task-specific media servers to be added to the network in order to get the job done.
What kind of value are we referring to? For example, voice channels may need to be recorded for quality assurance purposes. Consent to transactions may also need to be recorded. As well, voice channels may need to be 'mixed' so that a conference call can occur. Another example is 'whisper coaching,' which enables a supervisor to whisper private instructions to the agent over the phone without the caller hearing those instructions. Calls may need to be bounced between agents and an IVR application. All of these capabilities can be requirements in a call center environment. The response of vendors that lack the ability to 'mix' RTP streams is to throw more hardware and software at the problem by adding task-specific media servers to the network. Relying on such media servers can be both complex and costly because multiple ports and connections can be required for each caller and capability, involving additional licensing costs and hardware expense.
Of course, these challenges can be most effectively addressed by leveraging solutions that offer inherent RTP mixing ' without the need for task-specific media servers. The inherent ability to manipulate RTP streams as a pre-integrated function is, therefore, something that prudent buyers should look for when selecting an IP-based contact center solution.
"Hard" Phone Support
While some vendors advocate softphones (which transform the computer into a telephone via headsets connected directly to the PC), customers should take great care in evaluating whether the proposed softphone can 'reserve' CPU power for phone calls. As most of us know, desktops can sometimes slow down or freeze momentarily ' resulting in calls possibly dropping due to spikes in CPU usage by desktop applications. That's why most call centers prefer 'hard phones.' The challenge, indeed, is that adding proprietary hard phones can be expensive once you're 'married' to a vendor ' so a vendor that supports a diversity of lower-costing, third-party IP phones offers real value. A related issue, of course, is that you'll also want that vendor to actually provide phone-based controls on those third-party phones (to put people on hold, transfer calls, etc.). Some vendors claim support for open phones but don't invest in developing phone-based controls ' so caveat emptor ('buyer beware') applies.
Soft Gateway Bridging: Avoiding Costly IP "Tromboning"
In most call centers, agents work "off hook" with headsets. This helps to avoid the delays and productivity loss associated with manually picking up and hanging up the phone. But working off hook in an IP contact center can be costly if the solution isn't designed for it.
As you scale an IP contact center solution, you'll want to add additional voice gateways to your back-end infrastructure. This will enable you to scale gracefully instead of buying a 'bigger gateway' every time you hit a growth spurt. While most mainstream vendors are not challenged by this requirement, buyers need to be sure their proposed solution can, in fact, support multiple gateways and off-hook agents without 'tromboning' or 'doglegging' calls (i.e., 'hard bridging' calls across gateways).
Here's the issue: Because off-hook agents maintain an open SIP stream to particular gateways, in a multi-gateway deployment you need to understand how a call received on one gateway can be routed to an agent connected off hook to another gateway. If the solution supports the ability to 'soft bridge' calls across gateways (i.e., enhanced RTP redirect capabilities), then RTP streams can be dynamically reassigned to appropriate gateways. This means that callers and off-hook agents can always be connected in a multi-gateway deployment without tying up extra resources. Less sophisticated solutions will hard bridge calls across those gateways ' meaning that extra resources will be tied up in order to bridge calls across different gateways. Obviously, this would impact costs because additional resources would need to be provisioned in order to accommodate this extra resource requirement.
The use of hard bridging instead of soft bridging has even greater cost implications for multisite organizations with gateways at different locations. You will have these same resource consumption challenges whenever calls are transferred between gateways. Those problems go away if the agent's off-hook IP connection is not fixed but can be dynamically reassigned to the appropriate gateway on the corporate network, regardless of geography.
Fortunately, most solutions do leverage SIP to provide soft bridging by dynamically re-directing and re-inviting RTP streams to different gateways ' so hard bridging won't be required. The few solutions that don't provide capabilities to soft bridge should obviously be avoided by organizations that allow or require agents to work off hook.
While every IP contact center vendor will tout the benefits of converged networks and VoIP technology, buyers need to closely examine how the differences between vendors' approaches to technology will impact on costs over time. That 'what you need to know' mission is what we focus on every month in this column. If you've missed any of our previous columns this year, or if you simply want more information, feel free to contact us and we'll be happy to e-mail our first five columns to you and answer any questions that you might have.
Eli Borodow is the CEO of Telephony@Work, the leading provider of adaptive, multitenant IP contact center technology for contact centers and service providers. He can be reached via e-mail at email@example.com. Kevin Hayden is the Director of Integrated Contact Center Solutions at TELUS Communications Inc., a tier-1 telecommunications carrier in Canada and the Canadian leader in hosted contact center services. He can be reached via e-mail at firstname.lastname@example.org.