Topic 6 Customer heterogeneity and segmentation (II) Slides updated PDF

Preview

Summary

Download Topic 6 Customer heterogeneity and segmentation (II) Slides updated PDF

Description

Marketing Analytics and Big Data (MARK3054)

Topic 6: Customer Heterogeneity and Segmentation (II) Dr. Songting Dong Senior Lecturer in Marketing, UNSW Business School [email protected]

Needs-based Segmentation • Customer heterogeneity analysis: modeling the dependence between needs and descriptors, e.g., – Linear regression, t-test, ANOVA – Linear regression modeling interaction – Logistic regression, …

• Cluster analysis: modeling the interdependence among needs, e.g., – Hierarchical cluster analysis – K-means cluster analysis

Dr. Songting Dong UNSW Business School

Interdependence: A 2-dimension Case

Strength (Importance) A,B,C,D: Location of segment centers. Typical members: A: schools B: light commercial C: indoor/outdoor carpeting D: health clubs

Dr. Songting Dong UNSW Business School

.. . . .A. .... . . B. .. .... ... .

Perceptions/Ratings for one respondent: Customer Values

. .

.. ....... ... . C. . .. .... . ... .

. .

D

Water Resistance (Importance)

Distance between segments C and D

A New PDA in 2001: ConneCtor • Combines the features of – Mobile phone – PDA

Dr. Songting Dong UNSW Business School

Telecom Products

Dr. Songting Dong UNSW Business School

New PDA: Managers’ Question • Managerial question: – To whom shall we sell this new PDA?

• Research question – What segments exist in the market? (Features of each segment?) – Who are the consumers in each segment? – Which segment(s) is most attractive?

• Decision – Select segments and allocate resources. Dr. Songting Dong UNSW Business School

A Research Study • 15 needs variables (see data file)

Dr. Songting Dong UNSW Business School

Cluster Analysis • Cluster analysis is a convenient method commonly used in many disciplines to categorize entities (individuals, objects, and so on) into groups that are homogenous along a range of observed characteristics (variables) • In marketing, clustering is used to partition data such that the resultant groups are internally homogenous (cohesive), but externally heterogeneous (separated)

Dr. Songting Dong UNSW Business School

An Simple Example

Dr. Songting Dong UNSW Business School

Text Analytics and Sentiment Mining Using SAS

Another Example: How to Group?

Dr. Songting Dong UNSW Business School

Text Analytics and Sentiment Mining Using SAS

Another Example: Best Grouping?

Dr. Songting Dong UNSW Business School

Text Analytics and Sentiment Mining Using SAS

Another Example: OR …

Dr. Songting Dong UNSW Business School

Text Analytics and Sentiment Mining Using SAS

Another Example: OR …

Dr. Songting Dong UNSW Business School

Text Analytics and Sentiment Mining Using SAS

Conducting Cluster Analysis Formulate the problem Select a distance measure (we use Euclidean distance) Conduct hierarchical cluster analysis Decide on the number of clusters Conduct K-means cluster analysis Interpret and profile clusters

Dr. Songting Dong UNSW Business School

Adjust N of clusters if necessary

Euclidean Distance (2-dimension)

Dr. Songting Dong UNSW Business School

Euclidean Distance (n-dimension) • If  =  ,  , … ,  and  =  ,  , … ,  are two points in Euclidean n-space, the distance ( ) between  and  is given by:  ,  =  , 

Dr. Songting Dong UNSW Business School

Hierarchical Cluster Analysis • A method seeks to build a hierarchy or tree-like structure of clusters.

Dr. Songting Dong UNSW Business School

Hierarchical Cluster Analysis

3

Importance of Quality

2

.. 9

4 8

. .

. . 5

.

. . 1

7

Importance of Price Dr. Songting Dong UNSW Business School

6

New PDA Data (SAS)

Dr. Songting Dong UNSW Business School

New PDA Data (SAS)

…………

Dr. Songting Dong UNSW Business School

New PDA Data (SAS)

d = 0.6112

Dr. Songting Dong UNSW Business School

Conducting Cluster Analysis Formulate the problem Select a distance measure (we use Euclidean distance) Conduct hierarchical cluster analysis Decide on the number of clusters Conduct K-means cluster analysis Interpret and profile clusters

Dr. Songting Dong UNSW Business School

Adjust N of clusters if necessary

Decide N of Clusters • The distances at which clusters are combined – Normal RMS distance in SAS hierarchical results

• SSB/(SSW+SSB) – The ratio of between-group variance to total variance – Similar to R2 in regression – “Approximate Expected Over-All R-Squared”

• Theoretical, conceptual or practical considerations • The relative sizes of the clusters should be meaningful

Dr. Songting Dong UNSW Business School

K-means Cluster Analysis • A method aims to partition  observations into  clusters in which each observation belongs to the cluster with the nearest mean. • K-means clustering algorithm from “Visual Statistics Using SAS”

Dr. Songting Dong UNSW Business School

k-Means Clustering Algorithm Training Data

1. Select inputs. 2. Select k cluster centers. 3. Assign cases to closest center. 4. Update cluster centers. 5. Reassign cases. 6. Repeat steps 4 and 5 until convergence.

Dr. Songting Dong UNSW Business School

k-Means Clustering Algorithm Training Data

1. Select inputs. 2. Select k cluster centers. 3. Assign cases to closest center. 4. Update cluster centers. 5. Reassign cases. 6. Repeat steps 4 and 5 until convergence.

Dr. Songting Dong UNSW Business School

k-Means Clustering Algorithm Training Data

1. Select inputs. 2. Select k cluster centers. 3. Assign cases to closest center. 4. Update cluster centers. 5. Reassign cases. 6. Repeat steps 4 and 5 until convergence.

Dr. Songting Dong UNSW Business School

k-Means Clustering Algorithm Training Data

1. Select inputs. 2. Select k cluster centers. 3. Assign cases to closest center. 4. Update cluster centers. 5. Reassign cases. 6. Repeat steps 4 and 5 until convergence.

Dr. Songting Dong UNSW Business School

k-Means Clustering Algorithm Training Data

1. Select inputs. 2. Select k cluster centers. 3. Assign cases to closest center. 4. Update cluster centers. 5. Reassign cases. 6. Repeat steps 4 and 5 until convergence.

Dr. Songting Dong UNSW Business School

k-Means Clustering Algorithm Training Data

1. Select inputs. 2. Select k cluster centers. 3. Assign cases to closest center. 4. Update cluster centers. 5. Reassign cases. 6. Repeat steps 4 and 5 until convergence.

Dr. Songting Dong UNSW Business School

k-Means Clustering Algorithm Training Data

1. Select inputs. 2. Select k cluster centers. 3. Assign cases to closest center. 4. Update cluster centers. 5. Reassign cases. 6. Repeat steps 4 and 5 until convergence.

Dr. Songting Dong UNSW Business School

k-Means Clustering Algorithm Training Data

1. Select inputs. 2. Select k cluster centers. 3. Assign cases to closest center. 4. Update cluster centers. 5. Reassign cases. 6. Repeat steps 4 and 5 until convergence.

Dr. Songting Dong UNSW Business School

k-Means Clustering Algorithm Training Data

1. Select inputs. 2. Select k cluster centers. 3. Assign cases to closest center. 4. Update cluster centers. 5. Reassign cases. 6. Repeat steps 4 and 5 until convergence.

Dr. Songting Dong UNSW Business School

k-Means Clustering Algorithm Training Data

1. Select inputs. 2. Select k cluster centers. 3. Assign cases to closest center. 4. Update cluster centers. 5. Reassign cases. 6. Repeat steps 4 and 5 until convergence.

Dr. Songting Dong UNSW Business School

k-Means Clustering Algorithm Training Data

1. Select inputs. 2. Select k cluster centers. 3. Assign cases to closest center. 4. Update cluster centers. 5. Reassign cases. 6. Repeat steps 4 and 5 until convergence.

Dr. Songting Dong UNSW Business School

k-Means Clustering Algorithm Training Data

1. Select inputs. 2. Select k cluster centers. 3. Assign cases to closest center. 4. Update cluster centers. 5. Reassign cases. 6. Repeat steps 4 and 5 until convergence.

Dr. Songting Dong UNSW Business School

k-Means Clustering Algorithm Training Data

1. Select inputs. 2. Select k cluster centers. 3. Assign cases to closest center. 4. Update cluster centers. 5. Reassign cases. 6. Repeat steps 4 and 5 until convergence.

Dr. Songting Dong UNSW Business School

k-Means Clustering Algorithm Training Data

1. Select inputs. 2. Select k cluster centers. 3. Assign cases to closest center. 4. Update cluster centers. 5. Reassign cases. 6. Repeat steps 4 and 5 until convergence.

Dr. Songting Dong UNSW Business School

k-Means Clustering Algorithm Training Data

1. Select inputs. 2. Select k cluster centers. 3. Assign cases to closest center. 4. Update cluster centers. 5. Reassign cases. 6. Repeat steps 4 and 5 until convergence.

Dr. Songting Dong UNSW Business School

K-means Cluster Analysis 3-cluser result for new PDA data:

Dr. Songting Dong UNSW Business School

Conducting Cluster Analysis Formulate the problem Select a distance measure (we use Euclidean distance) Conduct hierarchical cluster analysis Decide on the number of clusters Conduct K-means cluster analysis Interpret and profile clusters

Dr. Songting Dong UNSW Business School

Adjust N of clusters if necessary

Profiling Clusters Segment 1 (type I users) • Needs for innovation & ergonomic low • Important to get work done (message, PIM, remote access, monitor) • Low WTP (maybe their company would pay for them??)

• Sales pros Dr. Songting Dong UNSW Business School

Profiling Clusters Segment 2 (type II users) • Important to access info remotely and demonstrate on monitor • Moderate WTP

• Service pros Dr. Songting Dong UNSW Business School

Profiling Clusters Segment 3 (innovators) • Needs for innovation, multi-media, and ergonomic is high • High WTP

• Die hard tech lovers Dr. Songting Dong UNSW Business School

Conducting Cluster Analysis Formulate the problem Select a distance measure (we use Euclidean distance) Conduct hierarchical cluster analysis Decide on the number of clusters Conduct K-means cluster analysis Interpret and profile clusters

Dr. Songting Dong UNSW Business School

Adjust N of clusters if necessary

Export Cluster Membership

Dr. Songting Dong UNSW Business School

Export Cluster Membership

• In the cluster data, you will find the “CLUSTER” column, which tells you which cluster a record belongs to.

Dr. Songting Dong UNSW Business School

[Optional] Link Clusters to Descriptors • Multinomial logistic regression: http://www.realstatistics.com/multinomial-ordinal-logistic-regression/  ℎ =  ~ +   +   +   +   +   +   +  ℎ +   +   +  ℎℎ +  ℎ +   + ⋯ • How to utilize the results?

Dr. Songting Dong UNSW Business School

New PDA: Managers’ Question • Managerial question: – To whom shall we sell this new PDA?

• Research question – What segments exist in the market? (Features of each segment?) – Who are the consumers in each segment? – Which segment(s) is most attractive?

• Decision – Select segments and allocate resources. Dr. Songting Dong UNSW Business School

Determining Which Segments to Serve Criterion

Examples of Considerations

I. Size and Growth 1. Size 2. Growth

• Market potential, current market penetration • Past growth forecasts of technology change

II. Structural Characteristics 3. Competition • Barriers to entry, barriers to exit, position of competitors, ability to retaliate 4. Segment saturation • Gaps in the market 5. Protectability • Patentability of products, barriers to entry 6. Environmental risk • Economic, political, and technological change III. Product-Market Fit 7. Fit 8. Relationships with existing segments 9. Profitability

Dr. Songting Dong UNSW Business School

• Coherence with company’s strengths and image • Synergy, cost interactions, image transfers, cannibalization • Entry costs, margin levels, return on investment

Segment Selection: GE Portfolio Matrix Competitive Strength Low 1.0

Average

High

3.0

5.0

Dr. Songting Dong UNSW Business School

Neutral Unattractive

Market Attractiveness

Attractive

5.0

3.0

1.0 Market Segment 1

Market Segment 4

Market Segment 2

Market Segment 5

Market Segment 3

Market Segment 6

Example Criteria Relative Market Attractiveness Analysis 1 = compl etel y di s agree ; 2 = s omewha t dis a gree; 3 = neutra l; 4 = s omewha t a gree; 5 = comple tely a gree

Weight

1. Large Served Available Market Size 2. Strong Market Segment Growth Potential 3. Profitable Market Segment Customers 4. Weak Supply Base Negotiating Position 5. Weak Customer Negotiating Position 6. Absence of Strong Competition 7. Hard for Customers to Use Alternate Products 8. Barriers to New Competitors Entering Market

Totals

Dr. Songting Dong UNSW Business School

0.00

Score (1 - 5)

Weighted Score 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

Example Criteria Relative Competitive Strength Analysis 1 = completely dis a gree; 2 = s omewha t dis a gree ; 3 = neutra l; 4 = s omewha t a gree; 5 = comple tely a gree

1. Strongest Global Presence 2. Largest Market Share 3. Best-in-Class Product Quality 4. Powerful Brand Image 5. Premier Distribution Network 6. Highly Effective MARCOM 7. Right-Sized Production Capacity 8. Competitive Unit Cost 9. Superior Technical Depth and Breadth 10. Strongest Customer Relationships Totals Dr. Songting Dong UNSW Business School

Weight

0.00

Score (1 - 5)

Weighted Score 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

GE Portfolio Analysis

Market Segment Attractiveness (adaptation of Porter Five Forces Model)

Capitalize on Attractiveness

Challenge for Top Position

Build Preemptive Position

• Selectively position to improve base business

• Invest for growth

• Invest for growth

• Build on major strengths

• Diversify globally

• Create well-defined and rigidly enforced cut-off criteria • Monitor closely

• Protect areas of vulnerability

• Consolidate position

Harvest

• Focus on long-term cash flow

Preserve Position

• Reduce fixed costs

Expand Selectively

• Minimize capital expenditures

• Make selective investments to maintain position

• Seek most desirable areas of segment

• Reduce product line

• Seek opportunistic growth

• Seek opportunistic sale

• Apply strengths in more attractive areas

• Trade share for earnings as appropriate • Monitor closely for further decline

Exit • Exit market or prune product line • Determine timing so as to present maximum value • No further commitment of resources

Strict Cash Flow Management • Minimum commitment of resources • Protect position in attractive areas

Protect/Refocus • Selectively invest for earnings • Defend strengths • Refocus to attractive segments • Monitor for harvest or divestment timing

Competitive Strength (Available Levers) Dr. Songting Dong UNSW Business School

Most Segmentation Projects Typically Provide… General Insights

Not Action Plans Dr. Songting Dong UNSW Business School

In Practice… • Many segmentation studies are one-time projects that drain resources, because of how we think about segmentation. • Instinctively, firms think about target market segments that are: – Easily defined – Clear-cut – And reachable . . .

Dr. Songting Dong UNSW Business School

In Reality… • In reality, market segments are . . . – Hard to define – Fuzzy, and – Overlapping

Seg A

Overlap Seg B

• And, customer needs evolve over time. Dr. Songting Dong UNSW Business School

Seg C

A Good Segmentation Study … • Identifies segments of customers with differentiated needs. – How many different segments? – How do their needs differ?

• Enables segments to be separately targeted/reached (this can be problematic even if segments have distinctly different needs). • Finds one or more attractive segments (i.e. a profitable and separate marketing program can be designed for selected segments). • Facilitates the implementation of the segmentation scheme as an ongoing process, not a discrete project.

Dr. Songting Dong UNSW Business School

What about this “segmentation”? Ad in London Newspapers, 1913* Men wanted for hazardous journey. Small wages, bitter cold, long months of complete darkness, constant danger, safe return doubtful. Honor and recognition in case of success. — Ernest Shackleton,

Did it work? Dr. Songting Dong UNSW Business School

Lastly

Dr. Songting Dong UNSW Business School

Homework • Exercise – Examples in the slides – Tutorial preparation

• Team project – Revise your research plan and conduct the analyses

Dr. Songting Dong UNSW Business School

Next Week • Lecture topic: – Demand curve and revenue management

• Tutorial topic: – Exercise hierarchical clustering – Exercise K-means clustering

Dr. Songting Dong UNSW Business School...