Marketing Strategy
– Segmentation
Li Yang
Dec 2019
Last Time …
• Deriving value from customers
– Customers are assets
– Acquisition, retention and margin
• Creating value for customers
– Conjoint analysis
Course Roadmap
Customer and Developing Advanced Marketing
Marketing Strategies Marketing Plans Topics
Sessions 1 - 4 Sessions 5 - 8 Sessions 9 - 12
1 Deriving Customer Value 5 Pricing 9 Internet Marketing
2 Creating Customers Value 6 Channel 10 Social Media Marketing
3 Segmentation 7 Communication 11 Big Data & AI
4 Positioning & Targeting 8 Branding 12 Blockchain
Today
• Understand how to segment markets and select
appropriate target markets
• Identify the challenges associated with segmentation
• Understand how segmentation influences marketing
strategy
• Learning a segmentation tool – K-means clustering
Strategy as an Integrated
Cascade of Choices
Strategy is an integrated set of choices
What is our that uniquely positions the firm in its
winning industry so as to create sustainable
aspiration? advantage and superior value relative
to the competition.
Where will
we play? Lafley and Martin (2013, p. 3)
How will we
win?
What
capabilities
must be in
place? What
management
systems are
Source: Lafley and Martin (2013) required?
The Heart of Marketing Strategy
What is our
winning
aspiration?
Where will
we play?
How will we
win?
What
capabilities
must be in
place? What
management
systems are
Source: Lafley and Martin (2013) required?
Foundation of Marketing Strategy
Where will we play? How will we win?
Segmentation Targeting Positioning
S T P
Discover & profile Evaluate segment Define value
groups of attractiveness and proposition for target
customers with target ones that segments and develop
common needs are most attractive an action plan
How Do You like Orange Juice?
A. Sweet with pulp
B. Sweet without pulp
C. Less sweet with pulp
D. Less sweet without pulp
Pulp and sweetness: Who’s who?
Ideal Amount of
Pulp
Segment B
Segment A
Average Age: 40
Male : 49% Average Age: 26
Male: 52%
Ideal Sweetness
Some people like pulp in their orange juice
Some people like their orange juice sweet
Segmenting Europe
Based on different stereotypes …
Levitt on Segmentation
If you’re not thinking segments, you’re not
thinking. To think segments means you have to
think about what drives customers, customer
groups, and the choices that are or might be
available to them.
— Levitt, Marketing Imagination
What is Segmentation?
• Market segmentation is the subdividing of a market
into distinct subsets of customers
• Segment members are different between segments
but similar within
Why Segmenting?
Consumers are not the same!
Consumers are not the same!
They are very different from you!!
Why Segmenting?
Market Situation is Complex
Hospital A Hospital B
Died 123 72
Survived 3777 3744
Total 3900 3816
Death rate 0.032 0.019
Day Time Night Time
Hospital A Hospital B Hospital A Hospital B
Died 9 56 114 16
Survived 891 3552 2886 192
Total 900 3608 3000 208
Death rate 0.010 0.016 0.038 0.077
Benefit of Segmentation
To the Firm … To the Customer …
• Identification of valuable • Customized products &
customers services
• More targeted promotions • Personalized experience
& communications • Increased customer
• Higher CLV satisfaction
Sustainable Profit Growth Customer Loyalty & Retention
Bases for Segmentation
• Bases — characteristics that tell us why segments
differ (the “why”)
– E.g., Sensitivity to price, sensitivity to screen size
• Descriptors — characteristics that help us find and
reach segments (the “who”)
– E.g., high-income professionals vs. low-income students
Segmentation Bases:
Relevance and Observability
Segmentation Describer
Effective Descriptor Ineffective Descriptor
Do Not Own Own Low Education
Microwave Microwave High Education
20% 80% 30% 40%
Likelihood of buying Frozen Food Likelihood of buying Frozen Food
Characteristics of Good Segmentation
• Substantial
• Identifiable
• Distinctive
• Stable
Demographic Segmentation
• Based on age, gender, income, generation, marital
status, occupation, education, ethnicity …
Where Are We Going, Dad?
Demographic Segmentation
Demographic Segmentation
Post 70’s Post 80’s Post 90’s
Hard-working Enthusiastic Non-mainstream
White-collar Open Personality
Self-control “Moonlight clan” Homeboy/girl
Working extra- Against work extra- Against working
hours hours
Have savings Have debts Have Dads
Demographic Segmentation
Skincare products for men: purchase pattern by age group
Percent Growth
rate
Geographic Segmentation
• Sellers of Taobao
Motivation-based Segmentation
Social
Shoppers
Price-
Sensitive 7%
Affluents Brand-aspirationals
11%
Convenience 29%
Seekers 11%
Trendy Quality Price-Value
Seekers Shoppers
12% 16%
Conscientious
Objectors
14%
Behavior Segmentation
Be careful about
“stability”
Segmenting U.S. Millennials
Psychographic Segmentation
Segment Market by Consumers’ “Purchase Journey”
Mercedes Benz China
Segmentation by OS !!
Source: WSJ Aug 23, 2012
Segmentation on Weibo
Nike’s Weibo Campaign
Sport
Activeness
Loyalty to
Nike
Appropriating Value
Segmentation by CLV
Vulnerable Star
Customers Customers
High
Value
of
Customers Lost
Low
Cause
X
Free Riders
Low High
Value to Customers
Creating Value
Segmentation Methods
• Cluster analysis
• Cross-tab
• CHAID/Regression Trees
• Discriminant analysis
• Logistic regression
• Further readings:
“Fundamentals of Marketing Research” S.M. Smith and G.S. Albaum
“Handbook of Market Segmentation” A. Weinstein
Segmentation of Tablet Consumers
Price: Screen:
• $100 • 7 inches
• $300 • 10 inches
• $500
Battery life: Cellular network access:
• 4 hours • No
• 8 hours • Yes
• 12 hours
OS:
• iOS (Apple)
• Windows
• Android
The Data I Have
Respondent Intercept Scr: 10" Cell: Y Px: $300 Px: $500 Batt: 8h Batt: 12h OS: Win OS: And
1 3.27 -0.18 -0.33 0.33 -0.78 -0.72 0.61 0.55 0.50
2 4.93 -0.14 -1.14 0.83 -0.98 -1.02 -0.69 1.19 1.17
3 3.83 1.39 0.26 0.17 -1.71 -1.46 -0.96 1.29 -0.33
4 4.75 0.23 -0.96 0.17 -0.51 -1.16 -0.66 1.66 0.50
5 3.43 -0.27 -2.73 0.33 0.45 -0.12 -0.95 2.12 1.67
6 4.22 1.11 -0.50 1.00 0.08 0.25 -0.42 1.25 0.83
7 2.68 0.70 0.64 0.17 -0.11 -0.56 -1.06 3.06 0.50
8 2.24 0.55 -1.03 1.33 -0.33 -0.50 0.00 1.00 0.50
9 3.86 1.66 -0.54 -0.33 -1.91 -0.26 -0.26 0.59 -0.33
10 3.66 0.59 -0.33 1.00 -1.61 0.78 -0.05 1.89 1.00
11 5.38 0.97 0.34 -0.50 -0.89 -0.94 -1.28 0.28 -0.17
12 3.89 -0.63 0.17 0.00 -1.03 -1.97 1.03 -0.03 -1.00
13 3.60 -0.79 -0.57 0.50 -0.74 -0.10 1.07 0.60 0.67
14 6.13 0.53 -1.61 -1.00 -0.73 -0.27 -1.27 -0.06 -0.17
15 4.10 -0.07 -0.31 0.00 0.72 0.11 -0.55 -0.28 0.00
16 1.80 0.73 -0.70 0.00 -0.55 0.22 0.38 0.62 0.33
17 4.15 0.23 -1.47 0.67 -0.59 -0.08 0.42 1.41 -0.33
18 2.41 2.70 -0.37 0.50 -1.27 -0.90 -0.06 0.56 -1.33
19 4.03 1.06 0.29 0.83 -1.71 -0.79 -0.45 1.29 -0.17
20 3.88 0.36 -1.90 0.33 -0.02 0.35 -0.48 1.32 0.00
Segment by Battery Life
and Cell Connection
Cell connect
Battery life
When Bases are Complex
Cluster Analysis
• A class of techniques used to classify objects into
groups
– Objects within a group should be as similar as possible
– Objects belonging to different groups should be as
dissimilar as possible
K-means Clustering
• Start with desired number of clusters
• Random assignment at first
• Relocate using the relationship between
variances within and between groups as
criterion
Data
4
3.5
Cell Connectivity
2.5
1.5
1.5 2 2.5 3 3.5 4
Battery life
Iteration 1
4
Choose the centers of the three
3.5
clusters randomly
I made a rather bad choice on
Cell Connectivity
3
purpose
Allocate each point to its nearest
2.5 center
2 4
1.5 3.5
1.5 2 2.5 3 3.5 4
Cell Connectivity
3
Battery life
2.5
1.5
1.5 2 2.5 3 3.5 4
Battery life
Iteration 2
4
Choose the centers again as
3.5 the centroids of the clusters
Cell Connectivity
from iteration 1
3 Allocate each point to the
centroid that is closest to it
2.5
2 4
1.5 3.5
1.5 2 2.5 3 3.5 4
Cell Connectivity
3
Battery life
2.5
1.5
1.5 2 2.5 3 3.5 4
Battery life
Iteration 3
4
Choose the centers again as
the centroids of the clusters
Cell Connectivity
3.5
from iteration 2
3
Allocate each point to the
centroid that is closest to it
2.5
4
2
3.5
1.5
1.5 2 2.5 3 3.5 4
Cell Connectivity
3
Battery life
2.5
1.5
1.5 2 2.5 3 3.5 4
Battery life
Iteration 4 and Final step
4
Choose the centers again as
the centroids of the clusters
Cell Connectivity
3.5
from iteration 3
3 Nothing changed!!
Voila, you’re done
2.5
2 4
1.5 3.5
1.5 2 2.5 3 3.5 4
3
Battery life Cell Connectivity
2.5
1.5
1.5 2 2.5 3 3.5 4
Battery life
Example Using Table Data
Screen: Cell: Price: Price: Battery: Battery: OS:
Segment Intercept 10" Y $300 $500 8h 12h OS: Windows Android Share Name
Market 3.548 0.382 0.522 -0.757 -1.348 1.120 1.547 -0.716 -0.918
Cluster 1 2.079 0.575 0.763 -0.373 -0.965 1.649 2.291 -0.024 -0.424 36% "Road Warriors"
Cluster 2 4.728 0.074 0.399 -0.667 -0.987 0.843 1.146 -2.563 -2.307 25% "I phone only in Iphone"
Cluster 3 4.276 0.389 0.356 -1.215 -1.984 0.759 1.047 -0.198 -0.500 37% "Cheap Bustards"
• We can divide the market into 3 benefit segments
– One segment strongly values screen size, cellular service,
battery life, and is less price sensitive (“road warriors”)
– Second segment cares mainly about an apple device (“I
phone only in iPhone”
– Third segment is very price sensitive (“cheap bastards”)
Today’s Takeaways
• Be aware of averages
• Identify bases to segment the market – focus on
needs and benefits
• Segments should be similar within, distinctive
between
• Segments should be good predictors of behavior
• If bases are obvious, simply spilt; if bases are
complex, use methods such as cluster analysis
Next Time …
Where will we play? How will we win?
Segmentation Targeting Positioning
S T P
Discover & profile Evaluate segment Define value
groups of attractiveness and proposition for target
customers with target ones that segments and develop
common needs are most attractive an action plan
Next Time …
• Case: Chateau Margaux
– Groups 3 & 5
• Targeting and positioning
