Dominique M. Hanssens - Massachusetts Institute of...
Transcript of Dominique M. Hanssens - Massachusetts Institute of...
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