Multichannel data capture matures

answer,” he points out, “and companies frequently do not realize how
costly this can be.” Parascript will work with its customers from the early stages on to maximize the benefits of its investment in the technology.

Parascript is committed to improving the capabilities of its products, with 60 percent of its staff dedicated to advanced R&D. The check processing capability was its first commercial solution, and that was followed by expansion into reading letter mail. Parascript’s product now reads over 40 billion mail pieces per year for the U.S. Postal Service.

For the future, Ball sees good opportunities for applying Parascript’s technology to such issues as fraud prevention. “We have pioneered some new approaches,” he says, “such as being able to validate point-of-sale, online signatures against a control signature.” With biometric signature verification technology already available, retailers and banks can detect fraudulent signatures in real time and prevent the transaction from taking place.

Capturing customer feedback

In the highly competitive telecommunications industry, minimizing customer churn has a significant effect on profitability. Cablecom, a large cable company in Switzerland, offers Internet, television and radio, mobile phone and networking services. The firm has been using several products from SPSS to capture and analyze customer data. The information helps it predict which customers are likely to consider switching to another service, allowing the company to initiate specific actions to retain the customer. In addition, the information obtained from customer surveys allows Cablecom to understand customer needs and improve customer satisfaction, which also contributes to retention.

After using the SPSS data mining workbench, Clementine, for several years, Cablecom in 2005 added SPSS Dimension, a set of products designed for data collection. The ability to extend the capabilities of the SPSS platform is supported by the open architecture used in SPSS products.

“We have two major goals in using the family of products from SPSS,” says Federico Cesconi, director of customer insights and retention at Cablecom. “We want to shorten our time to market with new products, and provide the most responsive customer service possible.” To sustain a high level of customer service, Cablecom insists on a 24-hour callback tocustomers who have reported problems.

Customers provide their input either through a Dimensions form on the Cablecom Web site or by phone through CallRobot, which digitizes phone calls.

“We now have the ability to listen to our ‘customer voice,’” explains Cesconi, “and we can collect and analyze data very rapidly.” Those tools allow Cablecom to identify dissatisfied customers and deal with their problems immediately. The surveys are short—generally only four questions—and provide an opportunity for customers to leave at least two open-ended answers. Customer acceptance of the data capture methods has been good, with 20 percent responding to the online questionnaires and 40 percent to voice surveys.

Cablecom uses Clementine to carry out both numerical and text analyses. “Specific models score the feedback using each type of information,” Cesconi says. The predictive model developed allows Cablecom to anticipate which customers are likely to cancel their service and to intervene prior to that happening. “For example,” Cesconi says, “we are able to provide the customers with specific solutions in order to exceed their expectations and maintain high levels of satisfaction.”

Dimensions was created as a development platform for data capture, and was designed for customization. Although it is used widely for customer surveys, it also can be used for any kind of online data capture. Data is separated from metadata, which allows for multimodal data collection, including online, voice and paper capture.

“No matter what the mode, the data collected using Dimensions goes to the same central location,” says Nancy Dobrozdravic, VP of product marketing for customer feedback and analysis at SPSS. Companies seeking input from customers need a model that allows for input received through any channel. “Customers are more likely to provide comments in the context of an ongoing business interaction,” Dobrozdravic says, “so the survey should be available whether the customer is online or on the phone.”

Some online data capture systems allow the customer to view a form that is already populated with some information from the database. The customer is saved the trouble of re-entering existing information, and can respond to a questionnaire or correct information that has become obsolete.

“Response rates go up when the process is made easier,” says Dobrozdravic. In other cases, a survey begun on the phone by a customer service representative can be completed by the respondent online at a more convenient time.

Such integrated feedback is referred to as enterprise feedback management (EFM) and refers to the ongoing process of integrating multiple types of feedback through standardized collection and centralized storage. It improves corporate agility by providing a more comprehensive view of customer attitudes and expediting the analytic process. The results of the analyses can be used to quickly adjust marketing campaigns or product offerings.

“Obtaining customer feedback is not a productive exercise unless it leads to action,” Dobrozdravic emphasizes. “The results should flow directly back to an organization’s business processes and decisions.”