How to Measure and GrowHow to Measure and GrowReturn on Marketing Return on Marketing

InvestmentInvestmentacross Marketsacross Markets

and Marketing Actionsand Marketing Actions

Professor dr. Koen PauwelsProfessor dr. Koen Pauwels

Ozyegin University, IstanbulOzyegin University, Istanbul

Tuck School of Business at DartmouthTuck School of Business at Dartmouth

Why measure Return on Why measure Return on Marketing Investment ?Marketing Investment ?

• In services and manufacturing industries, In services and manufacturing industries, marketing should create profitable growthmarketing should create profitable growth

• Operations and finance are accountable Operations and finance are accountable for return on investments, why not for return on investments, why not marketing?marketing?

• Return on Marketing Investment badly Return on Marketing Investment badly needed:needed:

– 80 % of new products fail80 % of new products fail

– 50 % of advertising spending has no 50 % of advertising spending has no effecteffect

– 85 % of sales promotions result in net 85 % of sales promotions result in net losseslosses

How to Measure ROMI ?How to Measure ROMI ?

•Structured Executive OpinionStructured Executive Opinion

•Market ExperimentationMarket Experimentation

•Examining long-term data Examining long-term data patternspatterns

Limitations of Limitations of ExperimentationExperimentation

• Experiments assess short-run impact, but are Experiments assess short-run impact, but are costly and don’t capture competitive reactioncostly and don’t capture competitive reaction

• Short-term and long-term effects over time?Short-term and long-term effects over time?

• Impact on stages customer buying process Impact on stages customer buying process (awareness, intention, satisfaction) and firm (awareness, intention, satisfaction) and firm value process (sales, income, valuation)?value process (sales, income, valuation)?

• Media and marketing budget allocation?Media and marketing budget allocation?

Long-term Response Long-term Response ModelingModeling

•Econometric representation of Econometric representation of customer, competitor and customer, competitor and environmental driversenvironmental drivers

•Yield performance response elasticity Yield performance response elasticity (ROI), in the short-run and the long-run(ROI), in the short-run and the long-run

•Translate into actionable client advice Translate into actionable client advice for media allocation and marketing for media allocation and marketing strategystrategy

Long-term modeling Long-term modeling challengeschallenges• How long is long-term ?How long is long-term ?

• How to model wear-in and wear-out ?How to model wear-in and wear-out ?

• How to incorporate long-term reactions How to incorporate long-term reactions of retailers, competitors, company of retailers, competitors, company action (future decision making) ?action (future decision making) ?

Vector-Autoregressive Vector-Autoregressive ModelsModels• Flexible model of complex dynamic interactions Flexible model of complex dynamic interactions

between a set of endogenous variablesbetween a set of endogenous variables

• Unique perspective on marketing-performance :Unique perspective on marketing-performance :

1) consumer response to marketing is dynamic 1) consumer response to marketing is dynamic

2) competitive response to marketing actions2) competitive response to marketing actions

3) company response: performance feedback3) company response: performance feedback

Long-term sales response Long-term sales response modelmodel• Conventional sales-price response model:Conventional sales-price response model:

SS11 = a = a11 + + 11SS1,t-11,t-1 + b + b1111pp1 1 + b+ b12 12 pp22 + + 11

SS22 = a = a22 + + 22SS2,t-12,t-1 + b + b2121pp1 1 + b+ b22 22 pp22 + + 22

• But consumers forward buy and stockpile :But consumers forward buy and stockpile :

SS1 1 = a= a11++11SS1,t-11,t-1+b+b1111pp11++1111pp1,t-11,t-1++1212 p p1,t-21,t-2+..+..

Long-term Marketing ModelLong-term Marketing Model

• And competitors react with price change:And competitors react with price change:

PP22 = a = a44++4 4 pp1,t-11,t-1++4141pp1,t-11,t-1++4242 p p2,t-2 2,t-2 + … ++ … +44

• And firms decide based on sales And firms decide based on sales feedback:feedback:

PP11 = a = a33++3 3 pp1,t-11,t-1+b+b3131SS1,t-11,t-1+b+b3232 S S2,t-1 2,t-1 + … ++ … +33

Vector Auto-Regression Vector Auto-Regression (VAR)(VAR)

• ““long-term” can be few weeks, months or forever:long-term” can be few weeks, months or forever:

decided by data: forecast precision vs. complexitydecided by data: forecast precision vs. complexity

• Compute long-term ROMI as the net impact of Compute long-term ROMI as the net impact of consumer, competitor and company responseconsumer, competitor and company response

1 11 12 13 14

2 21 22 24

3 31 32 33 34

4 41 42 43 44

1, 13 14 1, 1,

2, 23 24 2, 2,23

1, 1,

2, 2,

0 0

0 0

0 0 0 0

0 0 0 0

i i i it t t i

i i i it t

i i i it t

i i i it t

S b b S S

S b b S S

p p

p p

1,

2,

1, 3,1

2, 4,

t

k t i t

t i ti

t i t

p

p

Price deal yields positive Price deal yields positive consumer response, but lowers consumer response, but lowers future sales, especially if future sales, especially if competitors reactcompetitors react

-4

-3

-2

-1

0

1

2

3

4

5

1 2 3 4 5 6 7 8 9 10 11 12 13

Weeks

Sal

es E

last

icit

y

Consumer

Competitive

Discount ‘works’, so firm Discount ‘works’, so firm repeatsrepeats

-4

-3

-2

-1

0

1

2

3

4

5

1 2 3 4 5 6 7 8 9 10 11 12 13

Weeks

Sal

es E

last

icit

y

Consumer

Competitive

Company

Worst case: each time you Worst case: each time you promote promote it hurts it hurts your bottom lineyour bottom line

CHEESE

-140000

-120000

-100000

-80000

-60000

-40000

-20000

0

1 3 5 7 9 11 13 15 17 19 21 23 25

Weeks

Reta

iler

Reven

ue r

esp

on

se

($)

Immediate effect

Adjustmenteffects

Permanent effect

Best case: permanent benefits Best case: permanent benefits through consumer trial and through consumer trial and learninglearning

-4

-3

-2

-1

0

1

2

3

4

5

1 2 3 4 5 6 7 8 9 10 11 12 13

Weeks

Sal

es E

last

icit

y

Consumer

Competitive

New product

Profitable Growth Profitable Growth ConditionsConditions

Mature Mature MarketsMarkets

Changing Changing MarketsMarkets

CampaigCampaignn

MarketinMarketingg

Business-as-Business-as-UsualUsual

FMCG FMCG promotions promotions

Pharma Pharma advertisingadvertising

Permanent Permanent BenefitBenefit

Automobiles Automobiles Product Product

innovationinnovation

ChanginChangingg

MarketinMarketingg

EscalationEscalation

Automobiles Automobiles Price RebatesPrice Rebates

Evolving Evolving BusinessBusiness

Medical Services Medical Services Detailing Detailing

Journal Advertising Works Journal Advertising Works Right Away, then dies outRight Away, then dies out

0

1

2

3

4

5

6

7

8

9

10

1 3 5 7 9 11 13 15 17 19 21 23

Months

JA

D E

lasti

cit

y

Direct-to-Consumer ads Direct-to-Consumer ads show Wear-in and Wear-show Wear-in and Wear-

outout

0

0.02

0.04

0.06

0.08

0.1

0.12

0.14

0.16

0.18

1 3 5 7 9 11 13 15 17 19 21 23

Months

DT

C E

lasti

cit

y

What is the Best that What is the Best that could Happen? could Happen?

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

Oct-97 Feb-98 Jun-98 Oct-98 Feb-99 Jun-99 Oct-99 Feb-00 Jun-00 Oct-00 Feb-01 Jun-01 Oct-01 Feb-02

Months

Gro

ss

Ma

rgin

$

Honda launches 1999 Odyssey

One-shot innovation One-shot innovation increases Honda’s Firm increases Honda’s Firm

Value ForeverValue Forever

$0

$5

$10

$15

$20

$25

$30

$35

$40

$45

$50

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26

Weeks

Fir

m V

alu

e R

OI

in M

illi

on

s

Long-Term ROI Varies by Long-Term ROI Varies by Segment and FirmSegment and Firm

0

50

100

150

200

250

Cars Trucks Minivans & SUVs

RO

I in

$ M

illio

ns

Toyota Daimler-Chrysler General Motors Ford

And by Innovation LevelAnd by Innovation Level

-4

-2

0

2

4

6

8

10

12

1 2 3 4 5

Innovation Level

RO

I Ela

sti

cit

y

Top Line Bottom Line

In Contrast, Rebates In Contrast, Rebates Decrease Firm Value in the Decrease Firm Value in the

long runlong run

-$60

-$40

-$20

$0

$20

$40

$60

$80

$100

$120

$140

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26

Weeks

Fir

m V

alu

e R

OI i

n M

illio

ns

As They lead to Competitive As They lead to Competitive EscalationEscalation

$0

$500

$1,000

$1,500

$2,000

$2,500

1/3/99 5/3/99 9/3/99 1/3/00 5/3/00 9/3/00 1/3/01 5/3/01

Avera

ge r

eb

ate

Chevy Dodge

Finally, are some Actions Finally, are some Actions Needed to Maintain Sales Needed to Maintain Sales

??

0.00

1000.00

2000.00

3000.00

4000.00

5000.00

6000.00

7000.00

8000.00

9000.00

Jan-93

Jul-93

Jan-94

Jul-94

Jan-95

Jul-95

Jan-96

Jul-96

Jan-97

Jul-97

Jan-98

Jul-98

Jan-99

Jul-99

Month

De

taili

ng

an

d S

ale

s

How to Grow ROMI ?How to Grow ROMI ?

Mature Mature MarketsMarkets

Changing Changing MarketsMarkets

MarketinMarketingg

CampaignCampaigns s

Continue to Continue to create better create better campaignscampaigns

Focus and be Focus and be ready to handle ready to handle

growthgrowth

Changing Changing marketinmarketingg

Allocate $ to Allocate $ to other other

marketing marketing actionsactions

Ensure Ensure marketing marketing

spending keeps spending keeps up with salesup with sales

Help Managers Drive Help Managers Drive GrowthGrowth

• Managerial GoalManagerial Goal remains the same : remains the same : Build and Defend Profitable Growth Build and Defend Profitable Growth

• 2121th th C Reality C Reality : Need to demonstrate : Need to demonstrate ROMI faced with rising costs and channel ROMI faced with rising costs and channel powerpower

• ApproachApproach: long-term response models : long-term response models yield measurable long-term benefits yield measurable long-term benefits

• Translate into actionable managerial Translate into actionable managerial insights and need for future monitoringinsights and need for future monitoring