Chances are your company views customer interactions with your inbound phone channel as a matter of reducing pain, rather than increasing gain. Inbound call centers can be expensive to maintain, and automation is often applied as a 'Band-Aid' to keep certain crucial transactions under control, including call routing and basic customer information requests. For customers, the experience is similar. Most have come to expect the usual routine of navigating a touch-tone menu tree and then waiting on hold. Frequently, this translates to a frustrating experience for high-value customers, as well as any others. But not all customers are created equal. Let's look at two sample customers from the area of retail finance. One has a mortgage, a healthy savings balance and multiple checking accounts. The other is a first-year college student with $42.19 in a basic checking account. When both customers try to reach 1-800-ABC-BANK, they'll hear identical messages. If you just realized this level of 'bl anket' treatment for all inbound callers represents a missed opportunity for better service and increased revenue, then you've made an important step toward understanding how CRM can work for your business and your inbound phone channel.
The ability to improve the efficiency of a call based on a customer profile, or some other demographic, is a powerful business tool. CRM software vendors have been providing this capability to call center agents and Web sites for years, but such tools have only recently begun to find their way into automated call center solutions. Obviously, the ability to insert content into a call flow requires a degree of flexibility nonexistent in many legacy interactive voice response (IVR) systems. However, the emergence of open standards, such as VoiceXML, for inbound call center automation has brought a significant opportunity for personalization of the caller experience. Call flows are no longer set in stone, now easily lending themselves to dynamic customization.
A simple example from the area of retail finance is the choice of appropriate cross-sell or upsell messages for the two callers briefly described above. One is an established banking customer with several accounts and substantial income; the other is a new customer with a single checking account. Reaching these two customers with appropriate messaging could be based on retrieving a profile, for example, from an existing CRM database, or simply from a set of business rules applied to available balances and account types. To a financial institution, however, there are sales opportunities relevant to both callers. One customer is likely to purchase high-revenue investment products, while the other represents a potential long-term customer who has yet to make use of additional banking services. Even by offering the most basic services, such as a savings account with automatic transfers, the bank might earn revenue on an otherwise unprofitable customer.
Of course, these are well-known premises to anyone with an understanding of basic marketing concepts. CRM has certainly created the potential to understand customers and to analyze behavior across different media to best tailor the enterprise's attempts to reach its customers with the most critical parts of its message. What remains missing in many automated solutions, however, is an integration of available data about customers as part of a holistic approach to customer contact across all available channels, especially the phone.
Surprisingly enough, messaging within a call is generally not regarded by callers as intrusive or solicitation as long as it fits with the overall context of the call; that is, as long as it pertains to products and services relevant to the caller. For example, a recent survey sponsored by a major call center hosting service revealed that companies with call center cross-sell or upsell programs showed a significant increase in caller satisfaction. The key to success here is that the targeted messaging to customers must appear to enhance the caller's level of service or service options of interest. Doing this, rather than simply inserting blatant solicitations, follows the same pattern of good service practices that might be offered by a live operator. Callers expect human operators to have a basic understanding of their relationship with the company, their previous concerns and even buying preferences. A targeted message simply brings this feeling of being understood into an automated solution, rather tha n creating the impression of a 'hard sell' which will likely go unnoticed by most callers.
Be aware, though, that callers will respond negatively to being presented with unsolicited or irrelevant information. An untargeted message such as, 'Can I interest you in signing up for a Gold Card today?' is not likely to appeal to either of our two sample callers. Each of them will have a different, but equally negative, response. The high-income caller likely already holds several credit cards, while the new, young customer may not even consider himself or herself credit worthy. Both will rightly perceive this message as an intrusion or an annoyance. Remember that, unlike a graphical medium such as the Web, it is impossible to 'overhear' a promotional message in the way that one can overlook a banner ad.
Understanding callers does not necessarily mean selling or messaging; however, businesses that can anticipate user calls and the reason for those calls are best positioned to respond to the unique needs and interests of their customers. Callers always have a primary motivation for reaching out to a business, yet most call centers force all incoming calls to self-screen and sort themselves into neat groupings by categories that follow the logic of the business, not of the caller. It is hardly surprising that as many as 30 percent of callers will immediately press '0' in an attempt to directly reach an agent when presented with an automated menu. Here again, maintaining a caller profile can dramatically improve the caller experience by eliminating unnecessary menu items or by preempting customer requests for information.
Take another case study from insurance call centers. Many callers will have questions about claims status, or they may want to file a claim. Those customers with existing claims will almost certainly have called at least once within a recent time frame (30-90 days). Simply tracking the frequency with which a certain caller (identified by ANI, or 'caller I.D.') has been in contact, the system can match that caller against a profile, determine if there is a pending claim and present an appropriate status message. This not only reduces call center costs, decreasing average handling time for calls, for example, but again provides the customer with the feeling that he or she has been heard, understood and appropriately accommodated for his or her time investment.
This type of high-touch approach to callers reflects what has become standard for self-service on the Web. Many consumers have learned that basic inquiries, such as payment due dates or bill amounts, can best be handled through a company's Web site. The majority of customers, even Web-savvy consumers, nevertheless use the phone for certain types of interactions; specifically, those types that are more urgent or complex. Why then would a business intentionally push such callers into an automated system that will, more than likely, frustrate them, by having them experience long hold times? Of course, no enterprise wants its callers to have this experience, yet this remains the overwhelming perception of the call center by most consumers.
The key to breaking the outdated perceptions of automated phone-based service lies in bringing personalization and profiling to sophisticated, voice-automated solutions. Until recently, however, this was a difficult ' and, understandably, low-priority ' task for most businesses with a call center. Prior to the advent of standards-based telephony platforms, call center applications required extensive custom coding, and the focus of most businesses was on implementing technologies such as CTI screen-pops and skills-based routing, rather than on integrating sophisticated back-end systems to the automated telephony interface. Indeed, the automated portion of the call was really only a necessary evil, required to route callers and collect basic information before passing calls to a human operator.
VoiceXML, which has been deemed one of the enabling technologies for the newest round of automated voice systems, changes this by allowing speech and telephony applications to run using the same infrastructure as a company's Web site. Pages both are delivered from Web servers and use the same interfaces to back-end data. This model means that the same technologies which bring customer data to the Web can now bring that same level of personalized data to the telephone. Calls can now incorporate alerts, customize menu options and even retain preferences in the same way that a Web site would when you first login to your account. Call flows have been broken free of the model of static lists of options.
Interestingly, the phone channel has, for the most part, remained isolated from the Web more for historical reasons rather than for any conscious business decisions. Even though more than 90 percent of all customer service transactions still take place over the phone, there continues to be a perception that the phone can never serve as an effective channel for service automation. In many cases, the phone should remain a privileged channel with a reduced level of automation simply because callers expect some degree of human interaction and agent availability. However, businesses can still strive to reduce the need for human interaction by preempting the caller's original intent. And again, the key to achieving this through automation is by applying as much personalization and background knowledge to the call as available technology will allow.
Personalization is not only possible in automated call center solutions, it is now being actively deployed by forward-looking enterprises as a viable alternative to reduce costs, retain customers and generate additional revenue. The factors that have previously worked as barriers to a unified view of customers have now been removed. There is no longer anything hindering the power of today's 'phone channel' as a complement to the other automated solutions used by many of today's enterprises. CIS
As the pioneer of v-Business, Apptera (www.apptera.com) helps businesses intelligently automate and personalize each customer interaction, generating a dynamic and unique experience for callers and ensuring the optimal balance of service excellence, higher revenues and cost savings. Apptera works with enterprises of all sizes to maximize their inbound phone channel to fully extract and understand the needs and interests of their customers, leveraging their profiles to create lasting loyalty and increased profitability.
Dave Holsinger is Senior Product Manager at Apptera. Holsinger has worked in the speech industry for the last five years. He joined Apptera in 2004, having previously worked at OnStar, Voxeo and ViaFone. Currently, Holsinger is responsible for a number of projects that combine speech recognition with search technologies. He earned his Ph.D from the University of Wisconsin at Madison, and taught linguistics at both Humboldt University in Berlin as well as the University of Wisconsin at Milwaukee.
[Return To The Table Of Contents ]