Market Response Models Market Response Models and and

Demand Creation Demand Creation

Dominique M. HanssensDominique M. Hanssens

MIT Data Center ColloquiumMIT Data Center ColloquiumDecember 2005December 2005

UCLA Anderson School of Management and Marketing Science Institute

2

OverviewOverview

What is marketing ? What is marketing ?

Research traditionsResearch traditions

Building brandsBuilding brands

Growing customer equityGrowing customer equity

3

Natural Evolution of BusinessNatural Evolution of Business

Production orientation Production orientation

Sales orientationSales orientation

Market orientation: new challengesMarket orientation: new challenges

4

STPSTP–– segmentationsegmentation–– targetingtargeting–– positioningpositioning

Marketing MixMarketing Mix–– productproduct–– pricingpricing–– distributiondistribution–– communicationscommunications

Building customer equity & brand equityBuilding customer equity & brand equity

5

Academic Research Support Academic Research Support Customer Behavior Customer Behavior –– Cognitive, social psychologyCognitive, social psychology–– Behavioral decision theoryBehavioral decision theory

Marketing Science Marketing Science –– Analytical models (economics, mgmt science)Analytical models (economics, mgmt science)–– Empirical models (econometrics, stochastic models)Empirical models (econometrics, stochastic models)

6

How important is marketing? How important is marketing?

PIMS studies:PIMS studies:–– ROI = f(marketing, cost factors)ROI = f(marketing, cost factors)–– RR22 = 85 %= 85 %

From US and UK focus to global focusFrom US and UK focus to global focus

7

Performance Improvement is called Performance Improvement is called forfor

Half of advertising does not workHalf of advertising does not work85% of promotions lose money85% of promotions lose moneyUp to 80% of new products failUp to 80% of new products failClio award winners do not perform betterClio award winners do not perform betterSpending escalations prevail Spending escalations prevail Cost plus pricing still rulesCost plus pricing still rules

8

The New Marketing StrategyThe New Marketing Strategy

Importance of Marketing InvestmentsImportance of Marketing InvestmentsNeed for a Market Response focusNeed for a Market Response focusDigital data enriched, 1Digital data enriched, 1--onon--1 marketing1 marketingMarketing Science approach Marketing Science approach –– statistics, econometrics, data miningstatistics, econometrics, data mining–– resource allocation prescriptions resource allocation prescriptions –– longlong--term strategic viewterm strategic view

9

Serving a dual purposeServing a dual purpose

Building brand equityBuilding brand equity–– difference between products and brands difference between products and brands –– give customers a reason give customers a reason notnot to shop to shop

Growing customer equityGrowing customer equity–– difference between firstdifference between first--time and repeat buyingtime and repeat buying–– asymmetry between acquisition and retention costsasymmetry between acquisition and retention costs

In both cases, longIn both cases, long--term effects are essentialterm effects are essential

10

I. I. Focus on the brandFocus on the brand

11

Market Response PrinciplesMarket Response Principles

Sales response curves are concave or SSales response curves are concave or S--shaped shaped ElasticitiesElasticities are good response metricsare good response metricsMost models are short run or crossMost models are short run or cross--sectionalsectionalVAR models offer systems approach, with VAR models offer systems approach, with longlong--term effects term effects

12

ShortShort--term Empirical term Empirical GeneralizationsGeneralizations

Price elasticity is Price elasticity is --2.52.5Distribution response is SDistribution response is S--shaped, shaped, elasticitieselasticities are highare highSales force elasticity is 0.5Sales force elasticity is 0.5Quality elasticity is 0.4 Quality elasticity is 0.4 Promotion elasticity is 4 to 12 Promotion elasticity is 4 to 12 Advertising elasticity is 0.1Advertising elasticity is 0.1Life cycles: p=0.01, q=0.5Life cycles: p=0.01, q=0.5

13

Sources of longSources of long--term marketing term marketing effectseffects

.. Immediate effectImmediate effect•• CarryCarry--over effectsover effects•• Feedback effectsFeedback effects•• Purchase reinforcement effectsPurchase reinforcement effects•• Decision rules in the firmDecision rules in the firm•• Competitive reactionsCompetitive reactions

14

LongLong--term Effects ? term Effects ?

Unlikely for price and sales force due to Unlikely for price and sales force due to competitive matching competitive matching Distribution : yes Distribution : yes Quality: yes, except in highQuality: yes, except in high--technology technology LR Promotion elasticity is 0 LR Promotion elasticity is 0 LR Advertising elasticity = 2 * SR LR Advertising elasticity = 2 * SR

15

Illustration: Illustration: Short and LongShort and Long--Term Sales Effects of Price PromotionsTerm Sales Effects of Price Promotions

(VAR methodology)(VAR methodology)

average SR elasticity: 3.944average SR elasticity: 3.944average LR elasticity: 0.046average LR elasticity: 0.046

SR LR

Positive 64% 4%

Negative 5% 1%

Zero 31% 95%

16

Conclusions Conclusions

Most marketingMost marketing--mix effects are realized in mix effects are realized in the short run the short run Little evidence of Little evidence of hysteresishysteresisBrand building opportunity resides mostly Brand building opportunity resides mostly in sustained quality, innovation, distribution in sustained quality, innovation, distribution and advertising (weaker) and advertising (weaker) Thus brand building is expensiveThus brand building is expensive

17

Are brands worth the sustained Are brands worth the sustained marketing investment?marketing investment?

Research opportunity: compare branded and Research opportunity: compare branded and generic products across various categories generic products across various categories ((AilawadiAilawadi et al., JM, October 2003)et al., JM, October 2003)Branded products enjoy a positive revenue Branded products enjoy a positive revenue premium = premium = volvolbb**pricepricebb--volvolgg**pricepricegg

Source of premium: higher market share, Source of premium: higher market share, lower price elasticity lower price elasticity

18

II. II. Focus on the customerFocus on the customer

based on joint research with Shijin based on joint research with Shijin YooYoo, ,

Singapore Management UniversitySingapore Management University

19

Contrasting Domains and MetricsContrasting Domains and Metrics

Product marketing (PM): sales volume and Product marketing (PM): sales volume and revenue, market share, brand equity, product revenue, market share, brand equity, product prices and margins, marketing mix spending, prices and margins, marketing mix spending, profitability profitability Comes from the world of mass marketing Comes from the world of mass marketing Relationship marketing (RM) : customer Relationship marketing (RM) : customer acquisition, retention rates, crossacquisition, retention rates, cross--selling, lifetime selling, lifetime value, customer equity, offervalue, customer equity, offerComes from the world of direct marketingComes from the world of direct marketing

20

PM/RM focus differs by industryPM/RM focus differs by industry

Marketing

individual

mass

RelationshipContractual Non-Contractual

Time magazine subscription renewals from promotion calls

Dell computer purchases from email

promotions

Netflix.com new membership

purchases from banner ads

Toyota car purchases from TV ads

Relatio

nship

Mar

keting

Product

Mar

keting

t( , ) (marketing mix )t t tCE a r f=

21

CE: why hard to measure in PM CE: why hard to measure in PM environment?environment?

1R rCE am A a mr rδ

= − + − + − (Blattberg and Deighton 1996)

# prospects acquiredAcquisition rate ( ) # prospects

a ≡Acquisition spending (A)Retention spending (R)Contribution margin (m) Discount factor (δ) # customers retainedRetention rate ( )

# customersr ≡

22

Research questionsResearch questions

How do marketing mix efforts influence How do marketing mix efforts influence aacquisition cquisition aand nd rretention ratesetention rates, and thereby CE in , and thereby CE in product marketing environment?product marketing environment?Is marketing mix impact on CE different from its Is marketing mix impact on CE different from its impact on sales ? impact on sales ? Is there any difference between these effects in the Is there any difference between these effects in the shortshort--run vs. the run vs. the longlong--run?run?

23

What is CE for a product What is CE for a product marketer ? marketer ?

The trick is to distinguish sales to existing The trick is to distinguish sales to existing customers vs. sales to new customers customers vs. sales to new customers Also, competitor sales come from Also, competitor sales come from either lost either lost customers or lost prospects customers or lost prospects These distinctions map product sales into These distinctions map product sales into acquisition and retention acquisition and retention

24

Acquisition rateAcquisition rate vs. Rvs. Retention rateetention rate

{ { {1 1 1purchasesales purchasequantity incidence rate

/ /

AP AP PRO AP AP PRO AP PRO PROt t t t t t t

t P PRO P PRO PRO P PRO AP Pt t t t t t t

N N N S q N S q NaN N N S q N S q N− − −

= = ⋅ = ⋅ = ⋅ ⋅

{ { {1 1 1purchasesales purchasequantity incidence rate

/ /

RC RC CUS RC RC CUS RC CUS CUSt t t t t t t

t C CUS C CUS CUS C CUS RC Ct t t t t t t

N N N S q N S q NrN N N S q N S q N− − −

= = ⋅ = ⋅ = ⋅ ⋅

25

Switching MatrixSwitching Matrix

tCustomers Prospects

Retained customers Lost customers

Acquired prospects(= New customers)

Lost Prospects

RCtS LC

tS

APtS LP

tS

RC

RC LC

SS S+

Customers → Retention

t-1AP

AP LP

SS S+

→ AcquisitionProspects

26

Now include marketing spending Now include marketing spending

Sales from existing

customers

Sales from new customers

Sales from lost customers

Sales from lost prospects

Marketing mix impacts on…

acquisition

retention

Own Sales

Competitors’ Sales

27

ModelModel

11 12 13 14

21 22 23 24

1 0131 32 33 34

41 42 43 44 1

o o

c c

aa l l l l tt t lrr l l l lp qtt t l

o ol l l lx xl kt t t

l l l lc cx xt t t

ca acr r

x xcx xc

εφ φ φ φεφ φ φ φ

φ φ φ φ εφ φ φ φ ε

−

−

= =−

−

= + + +

∑ ∑ qt-kΘ Z

Direct effectCompetitive effect

Feedback effect Competitive reaction & Company’s decision rules

WOM effect

( | ) | ( ) | ( )t j t t t j t tIRF j x y E y x E x E y x E xδ+ + → = = + − =

28

Automobile Data ExampleAutomobile Data Example

Weekly transaction and marketing mix data of luxury Weekly transaction and marketing mix data of luxury segment in automobile industry from JDPAsegment in automobile industry from JDPA–– Observation period: 1/10/99~6/30/02 (182 weeks)Observation period: 1/10/99~6/30/02 (182 weeks)–– 26 PIN markets26 PIN markets–– 9 of 12 brands (93% of 9 of 12 brands (93% of 500K 500K transactions): Acura, Audi, Benz, transactions): Acura, Audi, Benz,

BMW, Cadillac, Infiniti, Lexus, Lincoln, VolvoBMW, Cadillac, Infiniti, Lexus, Lincoln, Volvo–– TradeTrade--in modelin model informationinformation–– Vehicle price, consumer rebates, and APRVehicle price, consumer rebates, and APR

Monthly advertising data from CMRMonthly advertising data from CMR–– Virtually all media including print, TV, radio, and outdoorVirtually all media including print, TV, radio, and outdoor–– Manufacturer ad + dealer adManufacturer ad + dealer ad

29

Sample DataSample DataAcura …wk1 wk2 wk3 wk182

Acura

Audi

BMW

:

Ford

20 25 30 … 40

Trad

e-in

Bra

nds

2 3 0 … 2

4 1 3 … 4

Price

Rebate

APR

AD spending

…

1 1 2 … 4

35,100 35,100 36,200 … 38,400

2,000 0 1,000 … 3,000

7.5% 4.6% 8.5% … 7.1%

3,500K 3,500K 3,800K … 4,000K

30

VariablesVariablesEndogenous VariablesEndogenous Variables–– Retention rate and acquisition rateRetention rate and acquisition rate–– Discount index = Discount index = f f (price, rebate, APR)(price, rebate, APR)–– Advertising spendingAdvertising spending–– Competitive discountCompetitive discountinging / advertising/ advertising

Exogenous VariablesExogenous Variables–– Constant and TrendConstant and Trend–– Seasonal Dummy: Labor Day, Memorial Day, End of each quarter Seasonal Dummy: Labor Day, Memorial Day, End of each quarter

(Pauwels et al. 2003)(Pauwels et al. 2003)–– New product introductionNew product introduction: step dummy: step dummy–– Product quality and customer satisfactionProduct quality and customer satisfaction

»» APEAL (Automotive Performance Execution And Layout)APEAL (Automotive Performance Execution And Layout)»» IQS (Initial Quality Study)IQS (Initial Quality Study)»» VDI (Vehicle Durability Index)VDI (Vehicle Durability Index)

31

Case study: Acura vs. LincolnCase study: Acura vs. Lincoln

0100200300400500600700800900

1 11 21 31 41 51 61 71 81 91 101

111

121

131

141

151

161

171

181

AcuraLincoln

Total Sales

32

Marketing MixMarketing Mix

Discounting Advertising

0

10002000

30004000

5000

60007000

8000

1 15 29 43 57 71 85 99 113

127

141

155

169

AcuraLincoln

1.0

1.1

1.2

1.3

1.4

1.5

1.6

1 14 27 40 53 66 79 92 105

118

131

144

157

170

AcuraLincoln

33

Sales DecompositionSales Decomposition

Sales from existing customers Sales from prospects

0

50100

150200

250

300350

400

1 15 29 43 57 71 85 99 113

127

141

155

169

AcuraLincoln

050

100150200250300350400450500

1 15 29 43 57 71 85 99 113

127

141

155

169

AcuraLincoln

34

CE MetricsCE Metrics

Retention rate Acquisition rate

00.10.20.30.40.50.60.70.80.9

1 14 27 40 53 66 79 92 105

118

131

144

157

170

AcuraLincoln

0.00

0.05

0.10

0.15

0.20

0.25

1 15 29 43 57 71 85 99 113

127

141

155

169

AcuraLincoln

35

Impact of Marketing Mix on CEImpact of Marketing Mix on CE

Discounting

AcuraST LT ST LTRetention Rate Acquisition Rate

-0.91 0.00 0.00 0.27Lincoln 0.00 0.00 0.00 -0.48

Acura 0.00 0.00 0.00 0.08Lincoln 0.00 0.00 0.00 0.21

AdvertisingST LT ST LTRetention Rate Acquisition Rate

ST LTSales

0.00 0.002.10 0.00

ST LTSales

0.17 0.000.00 0.00

36

Numerical IllustrationNumerical Illustration

Acura (Hysteresis)

Status QuoCurrent 4 Weeks later 1 Year later

Number of Customers 87,000 87,209 89,554 Profit 1,265,162 1,266,776 1,281,190 Customer Equity 561,735,680 562,207,732 567,483,558

Marketing Intervention ($2,000 extra discount)Current 4 Weeks later 1 Year later

Number of Customers 86,982 87,169 89,755 % of Status Quo 99.98% 99.95% 100.22%

Profit 277,732 1,280,712 1,296,524 % of Status Quo 21.95% 101.10% 101.20%

Customer Equity 567,916,404 568,982,982 574,470,488 % of Status Quo 101.10% 101.21% 101.23%

37

Numerical IllustrationNumerical Illustration

Lincoln (Escalation)

Status QuoCurrent 4 Weeks later 1 Year later

Number of Customers 50,762 50,452 47,081 Profit 473,532 472,119 459,842 Customer Equity 193,131,746 192,833,691 189,591,654

Marketing Intervention ($2,000 extra discount)Current 4 Weeks later 1 Year later

Number of Customers 50,762 50,512 47,087 % of Status Quo 100.00% 100.12% 100.01%

Profit 121,454 272,211 218,799 % of Status Quo 25.65% 57.66% 47.58%

Customer Equity 75,835,987 75,565,550 87,516,407 % of Status Quo 39.27% 39.19% 46.16%

38

ConclusionsConclusions

Marketing mix effects are different in the sales Marketing mix effects are different in the sales domain and the CE domain, both in the short run domain and the CE domain, both in the short run and in the long runand in the long runOn average,On average,–– Advertising and discounting affect salesAdvertising and discounting affect sales–– Discounting does not increase customer equity in the Discounting does not increase customer equity in the

longlong--runrun–– Advertising does not affect retention rate, but it has a Advertising does not affect retention rate, but it has a

positive impact on acquisition rate in the longpositive impact on acquisition rate in the long--runrunHigherHigher--quality brands are more acquisition quality brands are more acquisition effective in their marketing, and lowereffective in their marketing, and lower--quality quality brands are more retention effective brands are more retention effective

39

III. Overall ConclusionsIII. Overall Conclusions

40

As economies develop, a market orientation As economies develop, a market orientation becomes more important for firmsbecomes more important for firmsMarketing is about building brand equity and Marketing is about building brand equity and customer equity customer equity A good academic resource base is already in place, A good academic resource base is already in place, and growingand growingMarketing practice is being challengedMarketing practice is being challengedRigorous research on good data can be Rigorous research on good data can be managerially relevant managerially relevant

and, of course, and, of course,

The internet changes everythingThe internet changes everything

41

ReferencesReferences

Gupta, Sunil and Donald Lehmann (2005). Gupta, Sunil and Donald Lehmann (2005). Managing Managing Customers as InvestmentsCustomers as Investments. Wharton School Publishing. . Wharton School Publishing.

Hanssens, Dominique M., Leonard J. Parsons and Randall Hanssens, Dominique M., Leonard J. Parsons and Randall L. Schultz (2001). L. Schultz (2001). Market Response Models, 2nd EditionMarket Response Models, 2nd Edition. . Boston: Boston: KluwerKluwer Academic Publishers. Academic Publishers.

www.anderson.ucla.edu/faculty/dominique.hanssenswww.anderson.ucla.edu/faculty/dominique.hanssens