CATEGORY: CRM Analytics
REVIEWER: Greg James, vice president, CISM Statistical Services, Consumer and Small Business Financial Services for National City Bank.
BACKGROUND: Founded in 1845 and headquartered in Cleveland, Ohio, National City Corporation is a financial holding company with assets approaching $106 billion. National City operates banks and other financial services subsidiaries principally in Ohio, Michigan, Pennsylvania, Indiana, Kentucky and Illinois.
PLATFORMS: Clementine from SPSS is an open system that can access any ODBC-compliant data source and import flat file types. The Clementine server operates on Windows 2000 Professional or Advanced Server; Windows NT 4.0 with Service Pack 6 or higher; Solaris 2.6, 7 or 2.8; HP/UX 11.0 or 11i; AIX 4.3.3 or AIX 5L version 5.1; IBM OS/400 Version 5 Release 1 (V5R1, 5722-SS1) with OS/400 Portable Applications Solution Environment (PASE, 5722-SS1 Option 33) and OS/400 QSYSINC (5722-SS1 Option 13). Clementine client runs on Windows ME, Windows 98, Windows XP, Windows 2000 or Windows NT 4.0 with Service Pack 3 or higher.
PROBLEM SOLVED: The Consumer and Small Business Financial Services division of National City Bank markets its offerings to 5.4 million households and 10 million accounts a dataset that can run from a few megabytes to nearly a terabyte, depending upon the scope of analysis. The division's direct marketing efforts span all channels including inbound and outbound telemarketing, electronic banking and direct mail, and deal with a high complexity of customers, products and channels simultaneously. The challenge of our group Customer Information and Segment Management is to improve the effectiveness of the division's direct marketing processes through more accurate modeling and analysis.
PRODUCT FUNCTIONALITY: Clementine is used to create predictive and exploratory models for customer segmentation and direct marketing. The typical dataset for actual data mining can be 1 to 10 gigabytes. We analyze data prior to marketing campaigns to better target customers and prospects, and after campaigns to determine their effectiveness and profile the targets. While many marketing programs are executed over several weeks as traditional campaigns, a few ongoing ones utilize models that track customer transaction behavior on a daily basis.
STRENGTHS: Clementine's graphical interface pioneered the visual programming approach to data mining, and many other vendors have now followed suit. Clementine is designed to provide easy access to sophisticated data mining algorithms along with necessary support functions such as data access, preprocessing, graphing and reporting. The volume that the human mind can take in is much greater when the information is presented in a graphical rather than numeric format. This approach to data mining saves us a tremendous amount of time. I find Clementine to be one of the most intuitive data mining tools and one of the easiest to use.
WEAKNESSES: Today, marketing professionals are inundated with volumes of customer data. It is no longer a matter of if data mining is going to be a common business tool for marketing professionals, but a matter of when. For this reason, data mining vendors need to simplify the process flow and enhance graphic representations. Clementine has kept this type of business user in mind since its beginning and comes closest to being what we need; but there is still room to improve its ease of use among non-analyst business users.
SELECTION CRITERIA: We needed a data mining tool that produces actionable results without a lot of time spent on the underlying technology. A strong set of import and export functions was another selection factor since we may have up to six people working with the data at one time. Ease of use was also a factor. Clementine has delivered all this and more.
DELIVERABLES: With Clementine, we have made significant strides in improving our direct marketing campaign designs, testing methodology and marketing results. Our predictive modeling efforts are accomplished quicker and with greater accuracy, and we've experienced an annual 10 to 20 percent lift in several models since using data mining. Not every campaign is going to succeed, but now we have much better insights as to why.
VENDOR SUPPORT: SPSS has always been supportive of our work and issues. They've solved what few problems we've encountered over five years of use.
DOCUMENTATION: Overall, it would be a good thing if more information about the "art" of data mining were included to complement the current material about the "science" of data mining. Data miners need help knowing when and how to use various algorithms, not just how they work. The second edition of "Introduction to Data Mining," which is included as part of the Clementine documentation, is quite good.
Register or login for access to this item and much more
All Information Management content is archived after seven days.
Community members receive:
- All recent and archived articles
- Conference offers and updates
- A full menu of enewsletter options
- Web seminars, white papers, ebooks
Already have an account? Log In
Don't have an account? Register for Free Unlimited Access