Optimizing CRM Using Advanced Classifiers

• Activity Number: 105
• Type: Invited
• Date/Time: Monday, August 2, 2010 : 8:30 AM to 10:20 AM
• Sponsor: Section on Statistics and Marketing
• Author(s): Marianna Dizik*+ and Choongsoon Bae and Adam Ghobarah and Tony Fagan and Jim Koehler
• Companies: Google and Google and Google and Google and Google
• Keywords: CRM ; random forest ; R ; decision tree ; boosting ; bagging

Abstract

An important challenge for sales and marketing organizations is to target their (limited) resources to the right customers. Google CRM environment includes millions of customers and thousand of potentially useful features driving marketing or sales decisions. These decisions include customer service actions, support level, retention efforts, etc. The volume and complexity of available data require advanced and sophisticated statistical analysis and modeling to meet our needs. We address these problems and provide classifiers by advancing boosting, bagging, and Random Forests algorithms. We will demonstrate differences and benefits of different approaches and techniques to real Google problems.