Reinventing the Call Center
In the era of Do Not Call, predictive analytics can turn inbound calls—even complaints—into potential sales
Just when you think it’s safe to outsource your customer service operation, data mining powerhouse SPSS Inc. has thrown a monkey wrench into the works.
Last week the company unveiled PredictiveCallCenter, a new application that combines bread-and-butter CRM and call management features with SPSS’ data mining and predictive analytics technology. The result, SPSS officials claim, is software that’s designed to enable the dual-use call center—one geared toward both customer service and marketing.
Thanks to the federal Do Not Call registry and other efforts to reign in unsolicited telemarketing, it’s more difficult for customers to connect with customers by means of their outbound call centers, observes Marcel Holsheimer, vice-president of vertical product marketing with SPSS. In this respect, he argues, the inbound call center provides an opportunity for customers to market to customers who effectively opt-in by placing a call in the first place.
How so? Holsheimer points out that companies are already collecting information about prospective customers through Web submissions, direct mailings, and other channels. Therefore, when current or prospective customers call in, he says, there’s an opportunity to market to them. “[Customers] call the call center to complain about something, or have questions, or they’re interested in a product or service, and the idea is to take that information and combine it with things we already know about the customer, like historic behavior or behavior from other channels, and run analytics against it,” he explains.
What kind of analytics? For starters, PredictiveCallCenter can evaluate potential cross-sell and retention offers against an analysis of current and historical customer data, using a combination of predictive analytics and business rules to select an offer that is likely to be accepted by a customer and generate value for the company. The new product can also enhance traditional inbound call center tasks, such as customer compliant resolution, says Holsheimer.
“The system will automatically take the information for the call, like the customer is complaining about something, and then access the database to determine if [the caller is] a high-value customer, has complained in the past, or anything else that qualifies the customer for the retention complaint,” he explains. “A message will pop up on the [PredictiveCallCenter] screen, saying for this particular customer, give them a 10 percent discount.”
The upshot, SPSS officials claim, is that PredictiveCallCenter can increase cross-sell and up-sell hit rates by 50 percent or more.
The new offering draws equally upon SPSS’ own data mining and predictive analytics expertise, along with technology that it acquired from Data Distilleries in November 2003 (see http://info.101com.com/default.asp?id=3628). Earlier this month, SPSS introduced PredictiveMarketing, another product based on technology developed by Data Distilleries.
In this respect, says Holsheimer, PredictiveCallCenter is but one of several planned combined offerings. “What we’re looking at here is a range of applications—PredictiveMarketing, PredictiveCallCenter—and going forward we expect to launch other applications, such as PredictiveFraud. There’s going to be a whole range of applications,” he indicates.
Holsheimer acknowledges that SPSS doesn’t have a terribly large presence outside of its data-mining and predictive-analytic strongholds, but says that the new applications are designed to be deployed in non-SPSS environments. “The system enables you to integrate with existing call center environments, so … [customers] can leave their existing systems in place."
Stephen Swoyer is a Nashville, TN-based freelance journalist who writes about technology.