Upload
ngoquynh
View
220
Download
1
Embed Size (px)
Citation preview
E. Craig Stacey, Ph.D.
Measuring the Return on Marketing Investment
The Center for Measurable Marketing at NYU Stern
Measuring the Return on Marketing Investment
Outline
The Need for Marketing Accountability 1
A Framework for Marketing Productivity 2
Marketing Mix Modeling and Marketing Planning 3
MMM: Econometric Time-Series Analysis 4
Model Estimation 5
MMM for Today’s Media Landscape 6
Measuring the Return on Marketing Investment
Outline
The Need for Marketing Accountability 1
A Framework for Marketing Productivity 2
Marketing Mix Modeling and Marketing Planning 3
MMM: Econometric Time-Series Analysis 4
Model Estimation 5
MMM for Today’s Media Landscape 6
✓
“Half the money I spend on advertising is wasted; the trouble is I don’t know which half.”
- John Wanamaker
Corporate marketing is the last bastion of unaccountable spending in corporate America.”
- Eric Schmidt
2008 – 2010 Research Priorities
1. Accountability and ROI of Marketing Expenditures 2. Understanding Consumer/Customer Behavior 3. New Approaches to Generating Customer Insights 4. Innovation 5. Marketing Strategy 6. New Media
1. The Need for Marketing Accountability
The Center for Measurable Marketing at NYU Stern
www.msi.org
2010 – 2012 Research Priorities
“Marketing Management and Practice in an Evolving Landscape” 1. Delivering Value through Enhanced Media and Channels 2. Managing Brands in a Transformed Marketplace 3. Allocating Resources to Marketing Activities 4. Leveraging Research Tools and New Sources of Data
1. The Need for Marketing Accountability
The Center for Measurable Marketing at NYU Stern
www.msi.org
The Center for Measurable Marketing
• Measure marketing performance and integrate into business process
• Develop metrics and methods for existing and emerging communication channels
• Measure and manage brand equity
1. The Need for Marketing Accountability
The Center for Measurable Marketing at NYU Stern
Research Priorities
Measuring the Return on Marketing Investment
Outline
The Need for Marketing Accountability 1
A Framework for Marketing Productivity 2
Marketing Mix Modeling and Marketing Planning 3
MMM: Econometric Time-Series Analysis 4
Model Estimation 5
MMM for Today’s Media Landscape 6
✓
✓
“Metrics for Making Marketing Matter” 2. A Framework for Marketing Accountability
The Center for Measurable Marketing at NYU Stern Lehmann (2004)
Marketing Metrics Hierarchy 2. A Framework for Marketing Accountability
The Center for Measurable Marketing at NYU Stern
Marketing Capabilities
Marketing Activities (Inputs)
Individual (Customer) level attitudes and behaviors
(Aggregate) Product-market outcomes
Financial results (Performance)
Expectations of future performance (Valuation)
Outputs
Donald R. Lehmann and Dominique M. Hanssens, Marketing Metrics, Wiley International Encyclopedia of Marketing, 2010.
Framework for Marketing Performance Measurement 2. A Framework for Marketing Accountability
The Center for Measurable Marketing at NYU Stern Adapted from Lehmann and Reibstein (2006)
What marketers and competitors do • Advertising • Price • Promotions • Distribution
What customers think & feel
Awareness Consideration Liking
What customers do
Brand purchases
Direct Effect Feedback Effect
Advertising Campaign Testing 2. A Framework for Marketing Accountability
The Center for Measurable Marketing at NYU Stern Adapted from Lehmann and Reibstein (2006)
What marketers and competitors do • Advertising • Price • Promotions • Distribution
What customers think & feel
Awareness Consideration Liking
What customers do
Brand purchases
Direct Effect Feedback Effect
Brand Health Tracking 2. A Framework for Marketing Accountability
The Center for Measurable Marketing at NYU Stern Adapted from Lehmann and Reibstein (2006)
What marketers and competitors do • Advertising • Price • Promotions • Distribution
What customers think & feel
Awareness Consideration Liking
What customers do
Brand purchases
Direct Effect Feedback Effect
Marketing Mix Modeling 2. A Framework for Marketing Accountability
The Center for Measurable Marketing at NYU Stern Adapted from Lehmann and Reibstein (2006)
What marketers and competitors do • Advertising • Price • Promotions • Distribution
What customers think & feel
Awareness Consideration Liking
What customers do
Brand purchases
Direct Effect Feedback Effect
Mindset Metrics & Marketing Mix Modeling 2. A Framework for Marketing Accountability
The Center for Measurable Marketing at NYU Stern Adapted from Lehmann and Reibstein (2006)
What marketers and competitors do • Advertising • Price • Promotions • Distribution
What customers think & feel
Awareness Consideration Liking
What customers do
Brand purchases
Direct Effect Feedback Effect
Measuring the Return on Marketing Investment
Outline
The Need for Marketing Accountability 1
A Framework for Marketing Productivity 2
Marketing Mix Modeling and Marketing Planning 3
MMM: Econometric Time-Series Analysis 4
Model Estimation 5
MMM for Today’s Media Landscape 6
✓
✓
✓
Marketing Mix Models…
• … are representations of marketing systems that are useful for understanding the behavior of markets and for predicting the impact of marketing actions.
• … describe the relationship between performance measures, such as sales or market share, and controllable variables, such as price and distribution.
• ... are designed to aid managers in planning and forecasting.
3. Marketing Mix Modeling and Marketing Planning
The Center for Measurable Marketing at NYU Stern
Maximize Return on Marketing 3. Marketing Mix Modeling and Marketing Planning
The Center for Measurable Marketing at NYU Stern
Marketing ROI
Marketing Expenditure
Marketing Support
Incremental Contribution
Incremental Volume
Costs Margins
Model
Dimensions of Marketing Planning 3. Marketing Mix Modeling and Marketing Planning
The Center for Measurable Marketing at NYU Stern
Product
Time
Geography
Types of Marketing Planning 3. Marketing Mix Modeling and Marketing Planning
The Center for Measurable Marketing at NYU Stern
Product
Time
Geography
BRAND PLANNING ASSORTMENT
PLANNING
PORTFOLIO MANAGEMENT
Marketing Mix Planning 3. Marketing Mix Modeling and Marketing Planning
The Center for Measurable Marketing at NYU Stern
TRADEMARK
BRAND
PACK/SIZE
ACCOUNT DMA MARKET
Measuring the Return on Marketing Investment
Outline
The Need for Marketing Accountability 1
A Framework for Marketing Productivity 2
Marketing Mix Modeling and Marketing Planning 3
MMM: Econometric Time-Series Analysis 4
Model Estimation 5
MMM for Today’s Media Landscape 6
✓
✓
✓
✓
Model Design
• Model objectives • Selection of variables • Relations among variables • Data sources • Functional form
4. MMM: Econometric Time-Series Analysis
The Center for Measurable Marketing at NYU Stern
Objectives Specified in Model
• Sales • Market Share • Traffic, Average check size, … • Awareness, purchase intent, …
4. MMM: Econometric Time-Series Analysis
The Center for Measurable Marketing at NYU Stern
Selection of Variables
Examples of dependent variables: sales, market share, brand choice, awareness, attitudes, trial rates ... Examples of explanatory variables: marketing mix activities, macroeconomic factors ...
4. MMM: Econometric Time-Series Analysis
The Center for Measurable Marketing at NYU Stern
Variables
A Simple Model 4. MMM: Econometric Time-Series Analysis
The Center for Measurable Marketing at NYU Stern
Y (Sales Level)
} b (slope of the salesline) }
1
X (Advertising)
a (sales level when advertising = 0)
Response Function 4. MMM: Econometric Time-Series Analysis
The Center for Measurable Marketing at NYU Stern
Current Sales
Response Function
Min
Max
Current Effort
Sales Response
Effort Level
Data Sources
• Dependent and Independent Variables
• Internal vs. Syndicated
• Levels of Aggregation
4. MMM: Econometric Time-Series Analysis
The Center for Measurable Marketing at NYU Stern
Multiplicative Models
4. MMM: Econometric Time-Series Analysis
The Center for Measurable Marketing at NYU Stern
ihi
H
hii
hxs εα βΠ=
=1
)exp(
4. MMM: Econometric Time-Series Analysis
The Center for Measurable Marketing at NYU Stern
Q = eB0 XB1
(0 < B1 < 1)
0
2
4
6
8
10
12
0 20 40 60 80 100 120
Series1
Multiplicative model
• Only highest order interaction • Constant elasticity model • Log-log model
Most widely used sales response function
• Cobb-Douglas Production Function (1928)
4. MMM: Econometric Time-Series Analysis
The Center for Measurable Marketing at NYU Stern
Q = eB0 X1B1 X2
B2 ….. XjBj
4. MMM: Econometric Time-Series Analysis
The Center for Measurable Marketing at NYU Stern
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
0 1 2 3 4 5
Series1
lnQ = B0 + B1lnX1 (Linear in logs)
Measuring the Return on Marketing Investment
Outline
The Need for Marketing Accountability 1
A Framework for Marketing Productivity 2
Marketing Mix Modeling and Marketing Planning 3
MMM: Econometric Time-Series Analysis 4
Model Estimation 5
MMM for Today’s Media Landscape 6
✓
✓
✓
✓
✓
How Modeling Works 5. Model Estimation
The Center for Measurable Marketing at NYU Stern
Fit a model…
sale
s
to minimize the errors…
sale
s
-5
0
5
10
Jan Feb Mar Apr May Jun Jul Aug Sept Oct Nov Dec
error
0
10
20
30
40
50
60
70
80
90
100
Jan Feb Mar Apr May Jun Jul Aug Sept Oct Nov Dec
Advertising
Promotion
Seasonality
Distribution
Price
Base
Sales
Calculate contributions…
…over the year
0
20
40
60
80
100
120
Jan
Feb
Mar
Apr
May Jun Jul
Aug
Sept
Oct
Nov
Dec
Sales Model
Base20%
Price17%
Distribution21%
Seasonality15%
Promotion12%
Advertising15%
Example Model 5. Model Estimation
The Center for Measurable Marketing at NYU Stern
Sales Volume Fitted Model
R2 – 0.92 MAPE – 4.4%
Sal
es V
olum
e
Measuring the Return on Marketing Investment
Outline
The Need for Marketing Accountability 1
A Framework for Marketing Productivity 2
Marketing Mix Modeling and Marketing Planning 3
MMM: Econometric Time-Series Analysis 4
Model Estimation 5
MMM for Today’s Media Landscape 6
✓
✓
✓
✓
✓
✓
Traditional mix modeling does not work for integrated online measurement
• Treating online marketing (especially digital media) as just another direct driver of sales leads to misattribution of sales to drivers and ultimately bad mix decisions
• Even the best of today’s commonly used approaches have failed to keep up with today’s challenges
• The following picture simply no longer holds true:
6. MMM for Today’s Media Landscape
The Center for Measurable Marketing at NYU Stern
PRINT x COUPONS TRADE Online SALES X TV x
A far more complex marketplace exists in which online activity acts as an outcome of offline marketing, as well as a driver of sales.
A new approach is needed to manage complexity of the new media paradigm 6. MMM for Today’s Media Landscape
The Center for Measurable Marketing at NYU Stern
Earned Media
Coupon
Paid Media
Owned Media
Trade
SALES
Organic Search
Earned Media
Coupon
Paid Media
Owned Media
Trade
SALES
Organic Search
How Marketing Influences Sales 6. MMM for Today’s Media Landscape
The Center for Measurable Marketing at NYU Stern
As well as feedback effects
Direct Impact Sales
In-Store
TV
Online Display
Radio
It’s amazing what soup can do
How Marketing Influences Sales 6. MMM for Today’s Media Landscape
The Center for Measurable Marketing at NYU Stern
Marketing drives customers both directly and indirectly
Direct Impact Sales
In-Store
TV
Online Display
Radio
It’s amazing what soup can do
Drivers to Sales Immediate, Persistent, and Indirect 6. MMM for Today’s Media Landscape
The Center for Measurable Marketing at NYU Stern
Level of conversation activity (and sentiment) is typically monitored 6. MMM for Today’s Media Landscape
The Center for Measurable Marketing at NYU Stern
7/4/
2010
7/18
/201
0
8/1/
2010
8/15
/201
0
8/29
/201
0
9/12
/201
0
9/26
/201
0
10/1
0/20
10
10/2
4/20
10
11/7
/201
0
11/2
1/20
10
12/5
/201
0
12/1
9/20
10
1/2/
2011
1/16
/201
1
1/30
/201
1
2/13
/201
1
2/27
/201
1
3/13
/201
1
3/27
/201
1
4/10
/201
1
4/24
/201
1
5/8/
2011
5/22
/201
1
6/5/
2011
6/19
/201
1
7/3/
2011
7/17
/201
1
7/31
/201
1
Negative Neutral Postitive
Black Friday Store Promo Online Coupon Deal
PR Event
Context of conversations matters: Went to store/made a purchase 6. MMM for Today’s Media Landscape
The Center for Measurable Marketing at NYU Stern
• Went to store\Made a purchase
Context of conversations matters: Store/brand advertisements 6. MMM for Today’s Media Landscape
The Center for Measurable Marketing at NYU Stern
• Store\Brand Advertisements
Context of conversations matters: I love store/brand 6. MMM for Today’s Media Landscape
The Center for Measurable Marketing at NYU Stern
• I love store\brand
Drivers of offline store traffic 6. MMM for Today’s Media Landscape
The Center for Measurable Marketing at NYU Stern
• I love store\brand
So what can businesses do? 6. MMM for Today’s Media Landscape
The Center for Measurable Marketing at NYU Stern
• I love store\brand
• Track the content, not just volume and valence of social interaction, and relate to marketing and sales
• Move beyond observer role by targeting paid marketing to
stimulate specific social interactions • Encourage virtuous circle of social interactions leading to
purchase, which further induces social interactions about purchase (e.g. Kohls Facebook)
The Center for Measurable Marketing NYU Stern School of Business 40 West Fourth Street, Suite 915 New York, NY 10012 [email protected] MeasurableMktg
References • Lehmann, Donald R. and David J. Reibstein (2006),
Marketing Metrics and Financial Performance. • Lehmann, Donald (2004), “Metrics for Making
Marketing Matter,” Journal of Marketing.