February 2002  

BY CHRIS DONNER, CUSTOMER INTER@CTION SOLUTIONS

As technology advances make their way into the call center environment, bringing IP networks, multimedia, e-commerce and m-commerce, we see the inevitable displacement, or at least challenge, of the traditional PBX/PBX-adjunct solution. The PCs are over the walls, bringing with them the conveniences and shortcomings of PC-based platforms and their associated operating systems and software. One thing you could say about the traditional PBX: it ran. Enter IP-PBXs, software-based ACDs running on NT, VoiceXML-capable IVR systems, and so on, and suddenly youre not quite so sure that everything is going to keep running the way it should...or where to turn when something does go wrong.

And yet, to one degree or another, this is the future. So, what to do?

That is the issue that this article attempts to address. What are industrial computers and high-availability systems doing in a book that covers CRM, workforce management and e-sales? Well, what theyre doing is keeping all that other stuff running, freeing up call centers from their PBX chains and bringing them new functionality and ease of use while continuing to meet those core requirements of capacity and uptime.

To help with this article, I turned to some key vendors in this space and asked their opinions on high-availability systems in the call center. On the system side of things, I spoke with Jon Shapiro, CEO of Alliance Systems (www.alliancesystems.com), and Greg Colaluca, vice president of services of e-talk (www.e-talkcorp.com), about changes theyve witnessed in this market in the past 3-5 years as well as what they see ahead as customer service morphs in the next 3-5 years. On the power side of things, I turned to Roger Iannetta, product manager of American Power Conversions (APCs) DC Networks Division (www.apc.com).

RESISTANCE
When I first started this piece, I wondered what specific types of resistance resellers and manufacturers ran into in the call center environment when installing mission-critical systems. Did companies balk at the costs? Were they unwilling to consider alternative (and perhaps much less familiar) brands? Were there core issues of trust with PCs and PC-related software versus traditional PBXs that needed to be overcome?

And what about the invisible cost cutting on things like backup power, and even port availability and wiring. I had in mind an image from Christmas Vacation, where Clark Griswold clearly spends a fortune on lights for his house and yet has stuck together a pile of bargain-basement extension cords and power strips to carry all of that current. Despite all of his effort and frustration dealing with hanging the lights, the real problems began when he finally flipped the switch. Stuff that worked initially just began to stop working quite fast, in fact.

In response to my question about resistance, Jon Shapiro suggested that there were a few areas of primary failure when attempting to use an OTS system rather than an industrial computer: incompatibility with vendor software, driver support issues, and inconsistent or unreliable internal system power due to the demands of CTI cards compared with consumer-style PC cards. He also suggested that on the enterprise level there was a resistance to unfamiliar brands. This was less true when dealing with larger call centers, who are generally more familiar with supporting multiple brands and are able to trust (read: demand) that their various partners or resellers will provide the necessary service and support.

Greg Colaluca of e-talk discussed ROI and costs a bit more, especially on the smaller scale. Most pushback from industrial computer/high-availability systems comes from the initial sticker shock of such a platform, not necessarily the technical merits of the solution, he said. It is more challenging to approach a smaller contact center than a large enterprise due to the ROI the entry point for such a solution could be viewed as prohibitive for a smaller center that typically relies on redundant systems being separate, smaller NT servers. He also made an interesting point about costs, pointing out that contact center systems acquisition is not necessarily a call center staff decision, since the funds for such systems usually come from the enterprise IT budget rather than from the contact center budget.

PUSH VERSUS PULL?
One of the main questions I had was regarding the necessity of high-availability, PC-based systems given the increase in PC/IP-based solutions for next-gen, multimedia contact methods. Sure, a traditional inbound/outbound voice-only call center might get by with a traditional PBX and maybe an ACD or a voice mail add-on running on an NT server, but what about when customers suddenly start contacting this place via e-mail, the Web, IM, Internet chat, wireless Web, and so on? What then?

This issue resulted in a kind of mixed response, and understandably so as I considered the question more carefully. On one hand, Jon assured me that Alliance continues to see the acceptance of all-in-one solutions like Interactive Intelligence, Telephony@Work and other PBX replacement systems. But he suggested that the issue with CRM evolution was not the method of contact itself, but rather capacity, uptime, service/upgrade assurances. No one will just stick the Windows 2000 upgrade disk in one of these systems without a complete written upgrade plan, said Jon. Multimedia is not the issue; uptime and reliability are the key issues.

Similarly, Greg assured me that application availability was key, but also said that it has been proven through high-availability systems. He suggested that as expectations of customer service evolve, this will provide an additional entry point for high-availability systems, and that the key question will not be about availability but about bandwidth. Jon seconded this: We have installed over 13,000 of these systems in major accounts worldwide and have enough references on uptime data. Most issues still boil down to capacity, not uptime.

POWER
One of the things we woke up to this past year or two (besides the fact that the economists who claimed we had broken the cycle of recession and boom should all be dismissed for either lack of veracity or lack of insight) is that power is not something we can just rely on to come out of the wall. The rolling blackouts and price gouging in California were a frightening demonstration of just how unreliable power can be. Needless to say, without power no system can be considered mission critical for very long. Usually its only a matter of seconds before a system needs to be back up and running to meet service needs.

Jon touched on this when he discussed the added power requirements of call center/communications systems over more traditional consumer PCs. But before power reaches the system it has to come out of the wall and make its way through a maze of circuitry and wiring. For more on the requirements regarding power protection in a contact center, I spoke with Roger Iannetta from APC. He made this point even clearer: Power protection plays an important role in helping to ensure availability for the various equipment and critical applications that make up or support the network, from where the utility enters the building, to protecting the servers that house CRM and information sources, all the way to the desktop where transactions occur.

Regardless of the system they are running their applications on, network and telecom managers must be ready for any situation that threatens the integrity of their network or call center infrastructure. This doesnt only mean complete loss of power either. Roger mentioned power anomalies that can destroy sensitive hardware (chips, disks, etc.) in an instant, as well as the steady stream of dirty power that gradually wears down components until they fail well before their life cycle suggests they should.

Imagine the frustration of your customers being disconnected or placed in long hold queues (or worse, being in a long queue and then being disconnected) simply due to nuisance problems like keyboard lockups or loss of data integrity, resulting in system crashes, all of which can be caused by unreliable power. Roger suggested some components to look for when considering a power protection system, and I have collected them into the sidebar to this piece (opposite).

RESELLERS
While Ive already mentioned resellers in terms of the expectations for service and support placed upon them by their customers, it is also important to talk about the opportunity and the other challenges that the call center market presents to a reseller. The reseller can be a key contact between the manufacturers/distributors and the end customers. As such, the reseller becomes much more than a so-called middleman he/she becomes a long-term partner with both sides of the equation.

Jon Shapiro emphasized this when I talked with him. Resellers want one-stop shopping and a guarantee from the supplier that the system will work and the supplier is certified with the hardware and software solutions, he said. The VARs need a complete and available provider to help support these complex systems. Our value is making all this work and helping the channel partners with service and support.

VERTICALS
Greg Colaluca and I also talked about the importance of applications in certain vertical markets and how that affects system decisions. As an example, Greg discussed financial institutions, where certain transactions can be performed online or over the phone. The application that supports this business function will be deemed mission critical, Greg said, hence, more importance is placed on this function. Further, e-talks logging solution is deemed mission critical due to the nature of the transactions that our clients record. Many are financial or transaction-based interactions that require easy access to recorded data for order/account verification and transaction compliance.

CONTACT
Whereas a companys internal network and its external links to the Internet and VPNs allow communication within an enterprise collaboration and knowledge-sharing between employees call centers allow an enterprise to communicate with the outside world through sales, service and other relationship-developing activities. Network availability is crucial to the internal processes of a company; the availability of the call center is at least equal in importance to the external processes.

I suppose what it boils down to is that each enterprise or agency must determine what is mission critical for them, and what is not. After all, no one wants to spend money on something they dont need, especially these days. For example, while it might be inconvenient for an consumer e-commerce site to drop a few connections mid-transaction, it becomes much more serious when dealing with things like banking. These issues must be considered when deciding what applications need to be maintained as mission critical, and which applications can be trusted to a standard consumer-style platform.

An off-the-shelf (OTS) PC is always going to be cheaper than an industrial computer with a UPS that not only ensures against power failures, but also conditions dirty power to minimize nuisance crashes and lockups. Cheaper in the short run anyway. Often, the difference in costs between a mission-critical system and an OTS one can be recouped with the first failure that is avoided. Hows that for ROI?

The author may be contacted at lguevin@tmcnet.com.

[ Return To The February 2002 Table Of Contents ]