Print Ghosh Amit

Designing Effective & Efficient Retention & Acquisition Strategies for Retailers of CPG Amit K. Ghosh Cleveland State University Copyright © 2007, SAS Institute Inc. All rights reserved.

Challenges facing Retailers ƒ Decline in profits & contribution • Increased competition from low-cost, low-price retailers • Increased reliance on price cuts, temporary promotions, and markdowns • Increased “temporary price reductions” by manufacturers Copyright © 2007, SAS Institute Inc. All rights reserved.

Copyright © 2007, SAS Institute Inc. All rights reserved.

1

Strategic Initiatives Adopted ƒ Mergers & acquisitions ƒ New & competitive sourcing opportunities ƒ Leveraging the use of information technology & decision support tools • Innovative models – e.g., merchandise price and promotion optimization software • Continual manipulation of pricing & promotional strategies • Increased “temporary price reductions” by manufacturers Copyright © 2007, SAS Institute Inc. All rights reserved.

Category Management (CM) ƒ Data from Retail audits • Data at the store level • Analyze drivers of sales e.g., price sensitivity • Different models across time, region, etc • Across all customers (MACRO LEVEL) • Independently for each product category

ƒ Promising results Copyright © 2007, SAS Institute Inc. All rights reserved.

Copyright © 2007, SAS Institute Inc. All rights reserved.

2

CM – “striving to reach potential” ƒ merely “stripping waste from system” ƒ “shifting volume … rather than driving incremental consumption” ƒ “individuality in spite of volumes … magic formula” ƒ “future success …how they can retain customers by means of individual proposition” Copyright © 2007, SAS Institute Inc. All rights reserved.

Potential Limitations of CM ƒ Acquisition & retention strategies? • Drivers of loyalty / switching not analyzed

ƒ Incremental ROI implications? • Sensitivity of individual consumers’ to price reductions, coupons, etc. not analyzed

ƒ Cross-selling & Up-selling opportunities? • Each product category treated independently Copyright © 2007, SAS Institute Inc. All rights reserved.

Copyright © 2007, SAS Institute Inc. All rights reserved.

3

Customer Relationship Management for CPG ƒ “continuously create, enhance, & manage, customer equity by interacting with customers” ƒ Applications • “High customer interaction” industries −Service industries −Direct marketers −Business-to-Business • Not by retailers of Consumer Packaged Goods Copyright © 2007, SAS Institute Inc. All rights reserved.

Available database in CPG ƒ Loyalty or Affinity cards (Chain Level) • Response data (each purchase occasion) • Marketing initiative (price, promotion, etc.) • Demographic data

ƒ Supplemental Data (more complete picture) • Competitive activity (e.g., through surveys) • Customer data (purchased from other stores) • Category statistics (from IRI & others) • Demographic data overlay Copyright © 2007, SAS Institute Inc. All rights reserved.

Copyright © 2007, SAS Institute Inc. All rights reserved.

4

Potential benefits: CRM approach ƒ Existing customers • Identifying best customers and their potential impact on retailer ROI • Setting customer-level objectives to enhance incremental ROI ƒ Systematic acquisition of new customers by understanding profile of existing customers

Copyright © 2007, SAS Institute Inc. All rights reserved.

Potential benefits: CRM approach ƒ Strategies for existing customers • Understanding drivers of customer response at the individual level • Designing marketing strategies for each based on effectiveness of drivers • Understanding market-basket of customers to identify cross-selling opportunities

ƒ Designing “boarding” strategies for new customers

Copyright © 2007, SAS Institute Inc. All rights reserved.

Copyright © 2007, SAS Institute Inc. All rights reserved.

5

Uniqueness of our approach ƒ Focus on “individual” relationship −Analysis at MICRO (household) level

ƒ More complete understanding of consumer −Analysis of competitive information

ƒ Manage incremental ROI & consumption • Analysis NOT at the profitability level but at the component level

ƒ Utilize “standard” data-mining techniques Copyright © 2007, SAS Institute Inc. All rights reserved.

Basis for Today’s Illustration ƒ Scanner panel data (IRI) • One product category • Multiple regions / markets • Multiple years • All brands • All major stores • Limited demographics

Copyright © 2007, SAS Institute Inc. All rights reserved.

Copyright © 2007, SAS Institute Inc. All rights reserved.

6

CRM Business Process ƒ Managing existing customer base • Focus on existing customers

ƒ Changing existing customer base to achieve organizational goals • Focus on new customer acquisition

Copyright © 2007, SAS Institute Inc. All rights reserved.

Managing Existing Customer Base 1. Set Individual Objectives based on Metrics − Examples: total purchase potential, price premium paid, wallet share

2. Formulate Customer-Oriented Marketing Strategies based on Identification of Response Drivers − Examples: price, feature, display, coupon, temporary price reduction

Copyright © 2007, SAS Institute Inc. All rights reserved.

Copyright © 2007, SAS Institute Inc. All rights reserved.

7

Setting Objectives (Step 1) Identify Business Pain Points or Opportunities Select Success Metrics

Determine Customer Objectives Copyright © 2007, SAS Institute Inc. All rights reserved.

Analysis of Metrics for Selected Customers Customer Number

Total across entire product category

Proportion bought at the chosen retailer

Mean price per ounce

# of visits

Dollar value of purchases

Volume bought (in ounces)

Dollars

Sales (in ounces)

# of visits

Compared to price across all retailers

30

40

352.98

6534

0.72285

0.70921

0.65

1.03065

1346

34

203.5

3518

0.08821

0.1137

0.0882

0.65225

3410

92

503.35

8434

0.03008

0.01945

0.0326

1.5116

3552

57

642.95

14862

0.17109

0.16498

0.2456

0.97838

3908

31

205.88

3688

0.02905

0.06941

0.0322

0.39639

4579

73

398.59

7350

0.97554

0.97279

0.9726

1.00313

4875

33

201.66

3144

0.12987

0.09701

0.1212

1.42147

Copyright © 2007, SAS Institute Inc. All rights reserved.

Copyright © 2007, SAS Institute Inc. All rights reserved.

8

Managerial Implications ƒ Identification of Best Customers • What is the success metric? • Potential impact on ROI?

ƒ Changing Objectives (Metrics) at the customer level • New value of success metric? • Incremental ROI?

Copyright © 2007, SAS Institute Inc. All rights reserved.

Formulating Customer-Oriented Strategies (Step 2) Identifying Drivers Of Customer Response

Analysis

Customer-Oriented Marketing Strategy Copyright © 2007, SAS Institute Inc. All rights reserved.

Copyright © 2007, SAS Institute Inc. All rights reserved.

9

Response Drivers in CPG Markets P R O M O T I O N S

Features (%age of times bought when brand is featured)

Displays (%age of times bought when brand is displayed)

Temporary price reductions

Presence of TPR (%age of times bought During TPR)

Depth of TPR

C U S T O M E R

R E S P O N S E

C U S T O M E R

R E S P O N S E

(Mean & median of TPR) Copyright © 2007, SAS Institute Inc. All rights reserved.

Response Drivers in CPG Markets Store Loyalty Store Consideration Set Store Switching Brand Loyalty Brand Switching

(Number in last two years)

Brand Consideration Set (Number in last two years)

Copyright © 2007, SAS Institute Inc. All rights reserved.

Copyright © 2007, SAS Institute Inc. All rights reserved.

10

Response Drivers in CPG Markets Preferred Product Form C U S T O M E R

(Mode over last two years)

Average Packaging Size (Median over last two years)

Average Brand Price (Mean over last two years)

R E S P O N S E

Copyright © 2007, SAS Institute Inc. All rights reserved.

Response Drivers for Selected Customers Customer Number

Store Consid eration Set

Brand Consideration Set

Preferred Product Form

Average Product Size (ounces)

Average Price (Per Ounce)

Effective -ness Of TPR

Depth Of TPR Mean (Median)

30

2

8

liquid

100

0.055

12.5%

1335

3

7

liquid

100

0.085

25.8%

2853

7

4

liquid

100

0.045

13.9%

2952

2

7

liquid

50

0.082

6.1%

3590

4

7

liquid

64

0.041

8.6%

11.4 (10) 19 (14) 26.2 (22) 25 (25) 21.2 (20)

Effective -ness Of Features

Effective -ness Of Displays

80%

67.5%

19.4%

0%

5.6%

2.8%

6.1%

6.1%

4.9%

0%

Copyright © 2007, SAS Institute Inc. All rights reserved.

Copyright © 2007, SAS Institute Inc. All rights reserved.

11

Managerial Implications ƒ Strategy to retain Best Customers ƒ Strategy to change Baselines Metrics ƒ Evaluate cost and profitability at customer level ƒ Easily Assess Success of Marketing Campaigns

Copyright © 2007, SAS Institute Inc. All rights reserved.

Cross-Selling (Current Customers) ƒ Best Next Offer – from a customer perspective ƒ Goals • Make them more profitable : increase wallet share • Make them more loyal: form stronger relationships • Recognize inter-relationships across products & divisions

ƒ Methods • Cluster analysis • Market Basket analysis

Copyright © 2007, SAS Institute Inc. All rights reserved.

Copyright © 2007, SAS Institute Inc. All rights reserved.

12

Beyond the Existing Customer Base ƒ Analyze existing customer base to determine whether mix should be changed ƒ Use Customer Intelligence from existing Customer Base to • Acquire new customers • Develop a relationship with these customers Copyright © 2007, SAS Institute Inc. All rights reserved.

Customer Mix Optimal? ƒ Analyze Current Customer-base on Success Metrics ƒ Distribution of current metrics: acceptable and consistent with organizational objectives? ƒ If not satisfactory, what should the new distribution of metrics be? ƒ How can the new distribution be achieved? • New customer acquisition • Forced churn Copyright © 2007, SAS Institute Inc. All rights reserved.

Copyright © 2007, SAS Institute Inc. All rights reserved.

13

Explore distribution of Success Metric Category Spending / Dollars spent per visits

Copyright © 2007, SAS Institute Inc. All rights reserved.

Customer Acquisition Strategy ƒ Select Metric • Example, Category spending

ƒ Profile Demographics of Customers with various levels of Category Spending ƒ Design Campaign to Acquire High Category Spenders ƒ Assess Campaign Success Copyright © 2007, SAS Institute Inc. All rights reserved.

Copyright © 2007, SAS Institute Inc. All rights reserved.

14

Total Category Spending versus Household size

Copyright © 2007, SAS Institute Inc. All rights reserved.

Total Category Spending versus Income

Copyright © 2007, SAS Institute Inc. All rights reserved.

Copyright © 2007, SAS Institute Inc. All rights reserved.

15

Modeling & Prediction Is total category spending associated with demographics?

Copyright © 2007, SAS Institute Inc. All rights reserved.

Boarding Strategy for New Customers ƒ Determine the “drivers” of Positive Response by High Category Spenders ƒ Design Marketing Campaign based on these drivers ƒ Assess Campaign Success

Copyright © 2007, SAS Institute Inc. All rights reserved.

Copyright © 2007, SAS Institute Inc. All rights reserved.

16

Total Category Spending versus Feature

Copyright © 2007, SAS Institute Inc. All rights reserved.

Total Category Spending versus Display

Copyright © 2007, SAS Institute Inc. All rights reserved.

Copyright © 2007, SAS Institute Inc. All rights reserved.

17

Total Category Spending Rate versus TPR

Copyright © 2007, SAS Institute Inc. All rights reserved.

Modeling & Prediction Total category spending associated with marketing mix? ƒ

Analysis of Variance

ƒ

Sum of

ƒ

Source

DF

ƒ

Model

9

ƒ

Error

1872

ƒ

Corrected Total

ƒ

Mean

Squares

Square F Value

1798271

199808

5402238 2885.810800

1881

7200508

Pr>F

69.24 <.0001 .

.

.

.

.

Model Fitting Information

ƒ

R-square

ƒ

AIC

15004.9235

BIC

15007.0303

ƒ

SBC

15060.3244

C(p)

10.0000

ƒ

0.2497

Adj R-sq

0.2461

Analysis of Parameter Estimates

ƒ

Standard

ƒ

Parameter

ƒ

Intercept

ƒ

coup_pr_sum

ƒ

display_sum

1

8.1372

1.0718

7.59

<.0001

ƒ

feature_sum

1

4.6464

0.6650

6.99

<.0001

ƒ

prop_dollars_br_max

1

20.5472

5.4605

3.76 0.0002

ƒ

prop_dollars_ch_max

1 -32.1034

5.9514

-5.39

ƒ

tpr_pr_sum

3.7587

ƒ

tpr_dep_mean

ƒ

coup_val_mean

1

0.8807

5.7119

0.15 0.8775

ƒ

price_oz_mean

1

-5.7125

12.0518

-0.47 0.6356

DF 1

Estimate 59.5097

1

1 1

Error 5.9003

3.7212

-45.5749

t Value

Pr>|t|

10.09 <.0001

0.7998

0.7037 11.4139

4.65

5.34

<.0001

<.0001

<.0001

-3.99

<.0001

Copyright © 2007, SAS Institute Inc. All rights reserved.

Copyright © 2007, SAS Institute Inc. All rights reserved.

18

Conclusion ƒ Utility of CRM in the marketing of CPG by retailers ƒ Natural fit between CRM and data mining ƒ Manage incremental ROI & consumption • Analysis NOT at the profitability level but at the component level

ƒ Distinctive advantage and increased ROI • Efficient strategies • Effective strategies Copyright © 2007, SAS Institute Inc. All rights reserved.

Issues not addressed completely ƒ Cross-selling opportunities? ƒ Up-selling opportunities? ƒ Improve and expand metrics? ƒ Include other response drivers?

Copyright © 2007, SAS Institute Inc. All rights reserved.

Copyright © 2007, SAS Institute Inc. All rights reserved.

19

Copyright © 2007, SAS Institute Inc. All rights reserved.

Copyright © 2007, SAS Institute Inc. All rights reserved.

20