Part 4 of a four-part series.
Click here for Part 1, here for Part 2, and here for Part 3.
A customer relationship is a living thing, and can die a sudden, ignominious death if you fail to manage its health at every moment. If you don’t provide top-notch personalized service through the contact center—and every other external touch point—your customers may bolt for the competition at their next opportunity.
If you want to keep the customer happy, you should be providing reps with access to the most current, comprehensive feeds of customer information available. Sure, you’re already providing contact center agents with access to account profiles, purchase histories, service records, and other information. But that background information may not help defuse some acute issue that has the customer incensed right now. The difference between a satisfied customer and an ex-customer may simply be your ability to respond immediately to some make-or-break current event—such as the pain felt by a customer trying to understand her bill or carry out some simple transaction through your wretchedly designed Web site.
One of the new frontiers is a technology often known as complex event processing (CEP), which factors real-time updates of critical information into business intelligence and other decision-support applications. In the contact center, CEP technology is principally seen in applications known as recommendation engines. These automated rules-processing programs sift through dynamic feeds of relevant event information—critical, rapidly changing information—to help agents provide personalized service to each customer.
Much of this event information is generated in the context of whatever interaction your customer is currently engaging in through various channels (e.g., a contact center agent, Web self-service features, etc.). Recommendation engines may be able to monitor the full range of ongoing interactions, including text-based discussions such as instant messaging, email, short message service, blogs, and social networks. In addition, speech analytics technology could feed the recommendation engine with real-time updates on customer emotional states as expressed through over-the-phone voice inflections.
Crunching a vast stream of event information, CEP-enabled recommendation engines calculate the next best action suited to each customer, within the context of a current interaction. The engines apply predictive-analytics models and rules to each new event-data input, and can dynamically change recommendations based on customer responses, actions, or inertia. For example, an agent can be prompted to extend offers that are personalized to the specific profiles, interests, requirements, and circumstances of each customer—and a recommendation engine can take automated, event-driven actions, such as granting customer rebates or extending warranty periods, and then notify the customer and/or the contact center agent. It may even prompt the rep to indicate the customer’s current emotional state or buying propensity, which then become fresh events that factor into fresh recommendations.
“Complex” may appear in the name of the technology, but none of this interaction needs to be complicated where the customer or the agent is concerned. CEP-driven recommendation engines can be set up to present a simple, uncluttered interface to all involved. For the agent, it can be as simple as an autogenerated script on a display screen to guide the conversation. For the customer clicking around on your self-service Web site, it’s usually invisible, generating a personalized navigation path of which the customer may be blissfully unaware.
Recommendation engines can also be applied to batch-oriented customer interactions, such as up-to-the-second personalization of direct emailings. That creates the dangerous potential for launching personalized spam that appears to originate from a familiar, well-intentioned organization. So consider yourself cautioned: If you’re not careful, you can sour customer relationships just as easily as you can strengthen them through this newfangled synthesis of CEP and predictive analytics.
James Kobielus (jkobielus.blogspot.com) is a senior analyst at Forrester Research. You can email him at firstname.lastname@example.org.
Every month, CRM magazine covers the customer relationship management industry and beyond. To subscribe, please visit http://www.destinationcrm.com/subscribe/.