By Frederic Dickey
Contact centers are too often associated only with inbound contact. This is not only limiting to the companies using them because it implies that the contact center’s primary function is to process and react to customers’ concerns – this perception also ignores significant proactive contact capabilities of contact center platforms. If those capabilities are not recognized and leveraged, contact centers are underutilized and don’t deliver the return on investment they should.
The fact is that contact centers can and should allocate significant resources to outbound operations – not just for telemarketing or collections campaigns but also for customer service and notification services – in addition to inbound call processing. In this kind of blended contact center, agents perform more varied tasks. That not only improves the service provided to customers, it also makes contact center agents more fulfilled and reduces the risk that they will churn. The result is more satisfied customers and a more satisfied and secure workforce.
Technology is at the core of an effectively designed blended contact center – technology like predictive dialing that allows contact center operators to interact with customers in a variety of ways. When selecting a predictive dialer, it is critical to not overlook the important underlying technology that makes it effective: call progress analysis, or CPA. Understanding the importance of CPA – how it functions and how it continues to evolve – is key to equipping your contact center with the intelligence required for efficient and effective outbound contact.
Intelligent and Adaptive Technology: CPA (also known as call progress detection or answering machine detection) is a sophisticated set of algorithms embedded in contact center software and hardware that help provide context in identifying and classifying outbound calls. CPA automatically detects who or what answers the call in order for the contact center to decide where the call is directed – whether it’s back to a live agent (CPA detects a human answer), an automated message (CPA detects an answering machine), or back in the queue to retry later (CPA detects a busy line).
CPA is such a critical component of predictive dialing because it covers all the bases for what happens when an outbound call is placed – whether it’s answered by a human, goes to voicemail, or gets a busy signal. CPA allows the dialer to recognize when a call is not answered by a live person, thereby contributing to the productivity of the outbound function by screening out the calls that do not need to go to a live agent. The more precise the CPA function is, the more productive the call center will be.
CPA technology is continually evolving to improve the contact center experience. Early versions of CPA (let’s label them CPA 1.0) approached the automatic screening of calls using heuristics-based algorithms, which offered average performance. Heuristics is only average because it doesn’t adapt well to varying calling conditions and must be consistently retuned and re-optimized to account for changing network conditions (such as changing from a residential campaign to one targeted to mobile phones). It’s also implemented on specialized hardware, so it provides less flexibility for contact centers’ IT managers.
Recent evolutions of CPA (let’s label these CPA 2.0) have implemented neural networks algorithms in software, which function like “mini brains” that can be trained to recognize various call conditions. By listening to and processing several different types of calls, neural networks algorithms recognize the type of call with increasing accuracy, much like humans do. As a result, no retuning is required when call parameters change. The result is an estimated 95 percent accuracy in identifying call type, versus 75 to 80 percent accuracy with CPA 1.0. This difference means a dramatic improvement in contact center productivity.
Flexible Contact Center Platforms: The other important technology development for contact centers is an increasing reliance on voice over IP (VoIP) platforms. VoIP technology and applications run, in general, as software on standard servers using standard operating systems. This makes the VoIP contact center much more IT-infrastructure friendly because it runs on the same infrastructure you would use for mail servers, Web servers, and CRM applications. VoIP technology also allows much tighter integration of business processes with the telecom infrastructure. An example of such integration might be generating outbound campaigns directly from a CRM (customer relationship management) system.
As this VoIP transition is happening, it is important the CPA technology integrates easily into this infrastructure through software, not proprietary hardware.
An effective outbound contact center requires a technology provider that understands how technological advancements like VoIP and CPA can make outbound calling more accurate, faster, and more productive, resulting in a more efficient and effective outbound strategy and fully utilized contact center capabilities.
Frederic Dickey is director of marketing and product management at Sangoma Technologies.
[From Connection Magazine – March 2011]