Summary Customer Analytics 2017
Period Semester 4 of 2016-2017
Lecturer Dr. G. Knox
Summary made by Mirte van Schaijk
Including • 7 lectures given by Dr. G Knox
• The weblectures
Excluding • 2 guest lectures
Version 2
,SUMMARY CUSTOMER ANALYTICS 2017 STUVIA
Index
Lecture 1 Introduction ........................................................................................................................ 3
Lecture 1 Test and Roll Decisions ....................................................................................................... 4
1 Mailing to all customers (No test) Case E-Beer Mailing ............................................................. 4
2 Untargeted campaign: all-or-nothing Case E-Beer Mailing......................................................... 5
3 Targeted campaign ...................................................................................................................... 6
Lecture 2 Recency-Frequency-Monetary (RFM) analysis ................................................................... 7
1 RAW Use of Recency and Frequency........................................................................................... 7
2 Explaining response rate by RFM .............................................................................................. 10
Lecture 3 Logistic Regression ........................................................................................................... 15
Lecture 3 Lift Curves and the Gini .................................................................................................... 22
Lecture 4 Decision Trees................................................................................................................... 25
1 Motivation ................................................................................................................................. 25
2 Algorithm Growing bushes & trees ......................................................................................... 26
3 Decision trees in SPSS ................................................................................................................ 27
4 Interpreting Results ................................................................................................................... 29
Lecture 4 Overfitting and Cross-validation ....................................................................................... 33
1 Overfitting ................................................................................................................................. 33
2 Cross-validation ......................................................................................................................... 35
Lecture 5 Introduction to CLV........................................................................................................... 38
1 Case 1: CLV for new customers ................................................................................................. 40
2 Case 2: RLV: residual lifetime value........................................................................................... 42
3 Case 3 CLV, profit comes at the end.......................................................................................... 43
4 Managerial issue........................................................................................................................ 44
Lecture 6 Next level CLV calculations ............................................................................................... 46
1 Introduction to non-constant retention rate ............................................................................ 46
2 Building a better model BG Model .......................................................................................... 49
3 Using BG Model for CLV/RLV ..................................................................................................... 52
4 How to estimate parameters a and b........................................................................................ 54
Lecture 7 CLV in a non-contractual setting ...................................................................................... 59
1 Non-contractual settings ........................................................................................................... 59
2 Model 1 Parameters: p and q ................................................................................................... 60
3 Model 2 BGBB........................................................................................................................... 62
4 CLV and RLV under BG/BB ......................................................................................................... 63
Page 2 of 64
MADE BY: MIRTE VAN SCHAIJK
, SUMMARY CUSTOMER ANALYTICS 2017 STUVIA
Lecture 1 Introduction
Date 11 April 2017
Reading Chapter 9
Introduction - What do these companies know about you?
Companies such as T-Mobile, Facebook, Google, Amazon, ING, Bol.com, Netflix, etc have a lot of
information about you. a lot of data.
Example How target figured out a teen girl was pregnant before her father did?
A woman search for two different products. These products “say” that this is probably a
woman who is pregnant start advertise for coupons. Even before her father knew it.
Free service If something is free for you, you are not the customer
Problems with If your ad before a youtube video, it could be before some content you do not
targeting agree with (racism).
Introduction – The age of big data
• Massive customer data sets available now. But same old questions; whom do I target and
how? Which customers are the most valuable?
• Companies are drowning in data but starving for insights. Companies have a lot of data
about you, but that doesn’t mean they always have the knowledge or actionable insight to
do something with it.
Introduction – Customer Analytics
The aim of this course is to introduce you to methods to better understand your customers.
Customer Analytics Using (simple) models
and customer data to
make smarte
rmarketing deciscions.
Introduction – Scheme of lectures
Lecture 1-4 Lecture 5-8
• “Next period analytics” • “Long-term analytics
• Test marketing: why test? How large • Customer lifetime value (CLV): who are
should the test be? the most valuable customers: how do
you calculate the value of the firm of
the customers over his or her lifecycle?
• Models for selecting customer to • How does the portfolio of customers
target: Which customers should be change over time as customers drop
selected for e.g., acquisition, retention, out?
cross-selling, direct mailing? • Customer life time value (CE): What is
the value of a firm’s entire customer
base, i.e.: customer equity. CE = sum of
customer value.
Page 3 of 64
MADE BY: MIRTE VAN SCHAIJK
Period Semester 4 of 2016-2017
Lecturer Dr. G. Knox
Summary made by Mirte van Schaijk
Including • 7 lectures given by Dr. G Knox
• The weblectures
Excluding • 2 guest lectures
Version 2
,SUMMARY CUSTOMER ANALYTICS 2017 STUVIA
Index
Lecture 1 Introduction ........................................................................................................................ 3
Lecture 1 Test and Roll Decisions ....................................................................................................... 4
1 Mailing to all customers (No test) Case E-Beer Mailing ............................................................. 4
2 Untargeted campaign: all-or-nothing Case E-Beer Mailing......................................................... 5
3 Targeted campaign ...................................................................................................................... 6
Lecture 2 Recency-Frequency-Monetary (RFM) analysis ................................................................... 7
1 RAW Use of Recency and Frequency........................................................................................... 7
2 Explaining response rate by RFM .............................................................................................. 10
Lecture 3 Logistic Regression ........................................................................................................... 15
Lecture 3 Lift Curves and the Gini .................................................................................................... 22
Lecture 4 Decision Trees................................................................................................................... 25
1 Motivation ................................................................................................................................. 25
2 Algorithm Growing bushes & trees ......................................................................................... 26
3 Decision trees in SPSS ................................................................................................................ 27
4 Interpreting Results ................................................................................................................... 29
Lecture 4 Overfitting and Cross-validation ....................................................................................... 33
1 Overfitting ................................................................................................................................. 33
2 Cross-validation ......................................................................................................................... 35
Lecture 5 Introduction to CLV........................................................................................................... 38
1 Case 1: CLV for new customers ................................................................................................. 40
2 Case 2: RLV: residual lifetime value........................................................................................... 42
3 Case 3 CLV, profit comes at the end.......................................................................................... 43
4 Managerial issue........................................................................................................................ 44
Lecture 6 Next level CLV calculations ............................................................................................... 46
1 Introduction to non-constant retention rate ............................................................................ 46
2 Building a better model BG Model .......................................................................................... 49
3 Using BG Model for CLV/RLV ..................................................................................................... 52
4 How to estimate parameters a and b........................................................................................ 54
Lecture 7 CLV in a non-contractual setting ...................................................................................... 59
1 Non-contractual settings ........................................................................................................... 59
2 Model 1 Parameters: p and q ................................................................................................... 60
3 Model 2 BGBB........................................................................................................................... 62
4 CLV and RLV under BG/BB ......................................................................................................... 63
Page 2 of 64
MADE BY: MIRTE VAN SCHAIJK
, SUMMARY CUSTOMER ANALYTICS 2017 STUVIA
Lecture 1 Introduction
Date 11 April 2017
Reading Chapter 9
Introduction - What do these companies know about you?
Companies such as T-Mobile, Facebook, Google, Amazon, ING, Bol.com, Netflix, etc have a lot of
information about you. a lot of data.
Example How target figured out a teen girl was pregnant before her father did?
A woman search for two different products. These products “say” that this is probably a
woman who is pregnant start advertise for coupons. Even before her father knew it.
Free service If something is free for you, you are not the customer
Problems with If your ad before a youtube video, it could be before some content you do not
targeting agree with (racism).
Introduction – The age of big data
• Massive customer data sets available now. But same old questions; whom do I target and
how? Which customers are the most valuable?
• Companies are drowning in data but starving for insights. Companies have a lot of data
about you, but that doesn’t mean they always have the knowledge or actionable insight to
do something with it.
Introduction – Customer Analytics
The aim of this course is to introduce you to methods to better understand your customers.
Customer Analytics Using (simple) models
and customer data to
make smarte
rmarketing deciscions.
Introduction – Scheme of lectures
Lecture 1-4 Lecture 5-8
• “Next period analytics” • “Long-term analytics
• Test marketing: why test? How large • Customer lifetime value (CLV): who are
should the test be? the most valuable customers: how do
you calculate the value of the firm of
the customers over his or her lifecycle?
• Models for selecting customer to • How does the portfolio of customers
target: Which customers should be change over time as customers drop
selected for e.g., acquisition, retention, out?
cross-selling, direct mailing? • Customer life time value (CE): What is
the value of a firm’s entire customer
base, i.e.: customer equity. CE = sum of
customer value.
Page 3 of 64
MADE BY: MIRTE VAN SCHAIJK
