Slide 1

							                 MKT 700
Business Intelligence and
         Decision Models

                         Week 7:
Segmentation and Cluster Analysis
Seen Far
   DB Infrastructure
       Relational databases, data integrity and
        SQL queries
   Data Preparation
       Data cleaning and transformation
 CLV (Customer Expected NPV)
 RFM (Classification)
Outline
   Another classification technique:
     Cluster Analysis
     Theory

 SPSS help function (“Show Me”)
 Demo with Car Sales.sav
 Demo with Student2 with Scores.sav
 Demo with DMData.sav (Lab)
Clustering / Segmentation
 Segmentation, not all customers are
  created equals; no one-size-fit-all
 Basic marketing rules about
  segmentation
Consumer Segmentation
 Family life cycle (stage)
 Lifestyle
 Usage/Loyalty
 Preferred communication
 Buying behavior
Geo-Segmentation
   Environics
       http://www.environicsanalytics.ca/
   Pitney-Bowes
       http://www.utahbluemedia.com/pbbi/psyte/psyteCanad
        a.html
   Manifold:
       http://www.manifolddatamining.com/html/lifestyle/lifes
        tyle171.htm
   Generation5
       http://www.generation5.ca
B2B Segmentation
 Firm size (employees, sales)
 Industry
 Buying process
 Value in finished product
 Usage (Production/Maintenance)
 Order size and Frequency
 Expectations
BI Modeling Techniques
   No Target (Dependent Variable,
    Unsupervised Learning)
      • RFM
      • Cluster Analysis (Unsupervised learning)
   Target (Dependant Variable,
    Supervised Learning)
      • Regression Analysis
      • Decision Trees
      • Neural Net Analysis
Measuring distances
(two dimensions)

               A
      B




                      C




                          10
Measuring distances
(two dimensions)
             dac2 = (dx2 + dy2)
                  A
      B




                         C



             dac2 = (di)2
             dac = [(di)2]1/2
                                  11
Measuring distances
(two dimensions)

          D(b,a)
                            A
      B                         D(a,c)
                   D(b,c)


                                C




                                         12
    Cluster Analysis Techniques
   Hierarchical Clustering
       Metric, small datasets
Cluster Analysis Techniques
   K-mean Clustering
       Metric, large datasets


   Two-Step Clustering
       Metric/non-metric, large datasets,
        optimal clustering
Cluster Analysis Techniques




 See Chapter 23, SPSS Base Statistics for description of methods
Two-Step Cluster Tutorials
   SPSS, Direct Marketing, Chapter 3 and 9
      File: dmdata.sav

   SPSS, Base Statistics, Chapter 24
      Files: Telco.sav and Car Sales.sav
      Help: “Show me”

   http://spss.co.in/video.aspx?id=62
      Car Sales

   http://www.youtube.com/watch?v=DpucueFsigA
      File not available, but good demo
Two-Step Clustering
   Available
       Direct Marketing (Simplistic)
       Analyse  Classify (More Advanced)
   Measurement Scale
       Continuous  Euclidian Distance
       Nominal  Log-Likelihood
       Results will be different
   Repeat for stability
   Explore Viewer model
Top line from Chapter 10-1
Customer Segmentation
   Two kinds of segmentation: learning segmentation divides your
    customers by some criteria and sends them all the same message. See
    which group responds best and use that info. Dynamic segmentation:
    send each group different offers and text to get a better response.

   An ideal segment: large, well defined, can be motivated, and measured.
    Justifies a person’s time.

   Segmentation requires insight, analytics, and anecdotes. Segmentation
    action plan involves a road map, budget, goals, and tests.

   Segments and status levels (gold, silver) are not the same.

   Response rates can be improved by segmentation and RFM.

   Nielsen PRIZM segment codes can be profitable in deciding whom to
    promote to.
Top line from Chapter 10-2
Customer Segmentation
   Demographic data can be appended to any database that has postal
    addresses.

   Dynamic segmentation with direct mail is a winner.

   Segmentation does not work for e-mail promotions if you are mailing
    very frequently. There is not enough time to create dynamic content. The
    tradeoff: with e-mails the lift from segmentation is not as great as the lift
    from frequent e-mails. Result: many mass e-mail marketers do not use
    dynamic segmentation.

   Good e-mails with lots of links are more complicated to create than good
    print copy.

   There are LTV charts that show the lift from dynamic segmentation. In
    general, e-mail marketing staff members do not have the budget or the
    analytic capabilities to get the resources for database marketing.