Customer Segmentation Using SAS Enterprise Miner

No marketing or customer contact strategy can be effective without segmentation. While the concept of segmentation is deceptively simple, in practice it is extremely difficult to execute. Emphasizing practical skills as well as providing theoretical knowledge, this hands-on, comprehensive course covers segmentation analysis in the context of business data mining. Topics include the theory and concepts of segmentation, as well as the main analytic tools for segmentation: hierarchical clustering, k-means clustering, normal mixtures, RFM cell method, text-based clustering, time-series clustering, and SOM/Kohonen method. This course focuses more on practical business solutions rather than statistical rigor. Therefore, business analysts, managers, marketers, customer intelligence analyst, programmers, and others can benefit from this course.

    No classes are currenty scheduled for this course.

    Call (919) 283-1653 to get a class scheduled online or in your area!

1. Introduction

  • Define customer segmentation
  • Business context of customer segmentation

2. Segmentation Basics

  • Segmentation bases
  • Segmentation descriptors
  • Segmentation methods

3. Art and Science of Hierarchical Clustering for Segmentation

  • Introduction to clustering analysis
  • Similarity, distance, and distance metrics
  • Types of clustering

4. Applications of Hierarchical Clustering

  • Finding segments in B2B customer survey data
  • Modifying segments in B2B customer survey data
  • Profiling segments with bases
  • Applying Ward's method to find segments in B2B customer survey data
  • Validating segments with descriptors and other managerially important variables

5. k-Means Clustering

  • Mechanics of k-means clustering
  • Applications of k-means clustering
  • Scoring new data
  • Evaluating stability of cluster solution over time
  • Customer migration across segments over time
  • Probabilistic clustering

6. A Priori Segmentation Using RFM Cells

  • Application of RFM cell-based segmentation

7. Product Affinity Clusters

  • Application of product affinity in segmentation

8. Segments That Use Customer Comments (Text) Data

  • Introduction to text mining
  • Applications of text mining in segmentation and predictive modeling

9. SOM/Kohonen

  • Conceptual ideas behind the Self-Organizing Map (SOM)/Kohonen node
  • General principles of using SOM/Kohonen for segmentation problems
  • Using the Self-Organizing Map/Kohonen node in SAS Enterprise Miner

10. Time-Series Clustering

  • The business case for using time series data in segmentation/clustering
  • Using the Time Series node in SAS Enterprise Miner

11. Data Preprocessing before Segmentation

  • Variables selection for segmentation
  • Variables selection before segmentation
  • Variables transformation before segmentation
  • Outlier detection before segmentation
  • Missing data analysis

12. Wrap-up and Take-Aways

  • Missing data analysis
  • Split-sample validation
  • The role of segmentation/clustering in predictive models
  • Clustering procedures in Base SAS
  • Big-picture issues in segmentation/clustering

*Please Note: Course Outline is subject to change without notice. Exact course outline will be provided at time of registration.
  • Manage segmentation project cycle
  • Apply both attitudinal and behavioral segmentation tools and techniques on customer or prospect data
  • Use descriptive as well as predictive segmentation
  • Profile and validate segments
  • Evaluate stability of segments over time
  • Assign probability of segment membership to observations
  • Explore customer migration from bad to good segments over time
  • Create segments based on product affinity
  • Analyze textual data (such as customer comments) for segmentation
  • Find segments using time-series data
  • Use segmentation results to build predictive models
  • Perform data preprocessing tasks such as selecting a smaller number of variables from a large pool of input variables
  • Reduce dimensionality of data for building better models
  • Identify outliers in data via density and distance methods
  • Applying scale and shape transformation for better models
  • Handle missing values in your data

Exercises or hands-on workshops are included with most SAS courses.

  • Some prior exposure to SAS is useful but not required
  • No experience with SAS Enterprise Miner, SAS Enterprise Guide, or JMP is required
  • Segment customers based on attitude, preference, or transaction data to develop effective, targeted-marketing communications and promotions for each segment
  • Develop cross-sell and up-sell strategy based on customers' purchase patterns across product classes
  • Track and develop models for predicting customer migration from bad to good segments
  • Develop, deploy, and monitor comprehensive customer segmentation systems in their enterprise

Ready to Jumpstart Your IT Career?