Marketing

Engineering
1. Introduction
2. Fact & Data Based Decision
Material designed to accompany the book Principles of Marketing
Engineering and Analytics by Lilien, Rangaswamy and De Bruyn, and the Making
marketing analytics software Enginius available at http://www.enginius.biz
3. Models
4. Economic Concepts
5. A Look Ahead
© Copyright DecisionPro, Inc. 2018. Commercial distribution and any
publication is not authorized. Copying and distribution for non-commercial
educational purposes only is authorized if this notice appears on every copy.
Adaption for non-commercial educational purposes only is authorized if
© Copyright
attribution appears on every copy. DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 1
 Learn Specific Analytical
Learn Concepts and
Learn About Different Decision Tools and their Value
Analytical Frameworks
Areas in Marketing for Improving Marketing
Used in Marketing Decisions

OBJECTIVES
Learn to Interpret Analytical Learn How to Address
Become a Better Consumer of
Results, and Link them to Organizational Issues in
Business Performance Implementing Analytical Results Marketing Analytics

Market Response
Marketing Mix and Opportunity Costs Return on Investment
Models and Elasticity of of Decisions (ROI)
Resource Allocation
Response

Fixed Costs and Variable Using Excel and

EXPERIENCES Break Even
Costs and their Impact on Using Solver Tool Break-Even Analysis
and Safety Margin
(Session 2) Business Performance Within Excel

Sales and Market Share Response Developing/Using Organizational Understand the Contributions
totoDifferent
DifferentPrices
Pricesand
andLevels
Levelsof Chart to Anticipate Issues in of Judgment (Intuition or
ofAdvertising
Advertisingand
andSales
SalesEffort
Effort Implementing Analytical Results Experience) and Analytics

Installing and Using Running Different Computing Elasticity Reflecting on the

Marketing Engineering Decision Scenarios for And Cross-Price Results from Allegro
Software (Exercise) Allegro (Exercise) Elasticity (Exercise) Smart Spreadsheet
SUPPORTING
EVIDENCE
How Analytics Helps Decision Developing an Organizational Thinking About How to Present
Making in High Fixed Cost and Chart for Their own Resource Planning Results to Peers
High Variable Cost Industries Role/Position (Exercise) and to Higher Ups (Exercise)

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 2
Typical Decisions Made by
Marketing Managers

Segmentation Budgets

Campaign Effectiveness Marketing Mix

Targeting Market Size

Sales channels Market Share

Positioning Pricing Policy

Portfolio Management Advertising Design

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 3
Typical Consequences of
Marketing Decisions

• Website • Clickthroughs?
• CRM • Satisfaction?
? • New Alliances • Sales?
• Brand Advertising • Loyalty?
• Sales Promotions • Inventory turns?

• Sales?
ROI? •
•
Profits?
Share?
?

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 4
Usual Approach

Seat-of-the-Pants
Marketing Decisions

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 5
Data Mining Approach???

Marketing Data
Overload

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 6
Data  Insights for Action

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 7
Develop the Right Balance Through
Marketing Engineering

Marketing Engineering

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 8
Trends in Software Supported
Decision Making

• Marketing managers have high-powered personal

computers connected to networks, 7/24, everywhere.
• There are over 400 million installed copies of
Microsoft Excel (Business Week, July 13, 2006).
• Volume of marketing data is exploding (e.g., 500
Terabytes of transaction data at Wal-Mart).
• Firms are reengineering marketing for the
information age (e.g., Using Customer Relationship
Management systems).
• Faster Faster Faster !!!!

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 9
21st Century Marketing Decisions

• We have too much data of the wrong kind, not

enough of the right kind (data and information have
no value by themselves, but generate value through
their use).
• Humans are inconsistent, but “creative” information
processors (in both analyzing and synthesizing
information).
• Computers/mathematical models are consistent
information processors.
• Managers ➔ (Possibility of)
+ Models Better Decisions

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 10
Something to Think About

“In the end, sustained white-collar productivity

enhancement is less about breakthrough technologies
and more about newfound efficiencies in the cerebral
production function of the high value-added knowledge
worker”
-- Roach 2002 (Chief Economist, Morgan Stanley)

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 11
Marketing Engineering:
Marketing Analytics for the Manager

Marketing Engineering involves developing and using

interactive, customizable, computer-decision models
for analyzing, planning, and implementing marketing
tactics and strategies……or
Concepts, frameworks and tools to the rescue!
Even people who don’t particularly care for computers,
software, or math can learn some systematic ways to
think about marketing problems, and ask the right
questions.

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 12
The Opportunity for Marketing
Analytics
• The Global 1,000 companies spend about $1 trillion on
Marketing (Source: Accenture study 2001).

• 68% of the participants indicate they have problems even

articulating, much less measuring, the ROI of marketing
(Source: Accenture study 2001).

• “The mathematical modeling of humanity promises to be

one of the great undertakings of the 21 st century.”
(Business Week, January 23, 2006).

• Systematic marketing decision making can improve

marketing productivity by 5 – 10% with minimal additional
costs (i.e., it has very high ROI). (Source: Several studies
documented in Principles of Marketing Engineering).

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 13
Different Types of Analytics and
Their Competitive Implications
Competitive Advantage

Optimization What’s the best that can happen?

Analytic Capabilities

Predictive modeling What will happen next?

Forecasting/extrapolation What if these trends continue?

Statistical analysis Why is this happening?

Alerts What actions are needed?

Query/drill down Where exactly is the problem?

Ad hoc reports How many, how often, where?

Standard reports What happened?

Insights/Intelligence
Source: Adapted from Tom Davenport

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 14
 Climbing the Ladder of Marketing
Analytic Capabilities

Real-time
analysis
Develop flexible and dynamic offers and prices

Predictive
modeling
Learn to anticipate and prepare for the future

Campaign
management
Become efficient and effective in marketing spend

Event
triggers
Develop process and response capabilities

Segmentation
Treat different customers differently

Customer
database Get enterprise customer data into one place

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 15
What Wal-Mart did on 9/11

• Within a few hours of the 9/11 attacks,

sales of flags and other patriotic items
started skyrocketing.
• On Sept 11, the 2,700 Wal-Mart stores
sold over 100,000 flags (compared to
6,400 the previous year on that day), and
over 200,000 on Sept 12th.
• Detecting these increases, Wal-Mart
locked up all the supplies it could find
before its competitors (like Kmart) could
react.
• Real-time tracking and analysis helped
cope with a demand surge.

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 16
Top Performing Companies Use
Analytics
High Low
Performers Performers
Have significant decision-support/analytical
65% 23%
capabilities

36% Value analytical insights to a very large extent 08%

Have above average analytical capability within

77% 33%
industry

77% Have BI/Data Warehouse modules installed 62%

73% Make decisions based on ES data and analysis 51%

40% Use analytics across their entire organization 23%

* Based on an Accenture study with a sample of 400 companies worldwide (2005).

Source: Tom Davenport.

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 17
Analytics Begins at the Top in
Companies that Navigate via
Analytics
• Berry Berach at Sara Lee
“In God we trust; all others bring data” (quoting W. Edwards
Deming).
• Gary Loveman at Harrah’s
“Do we think, or do we know?”
“There are three ways to get fired at Harrah’s – for stealing, sexual
harassment, or instituting a program without first running an
experiment.”

• Jeff Bezos at Amazon

“We never throw away data”
• David Kearns at Xerox
“Do you have evidence to support that hypothesis?”

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 18
Today’s trends

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 19
Today’s trends

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 20
Today’s trends

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 21
Today’s trends

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 22
How do managers make decisions?

• Facts
“I know for a fact it will work…”

• Intuition
“I have the feeling it would work…”

• Reasoning
“In theory, it should work…”

• Experience / Practice
“It has worked before…”

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 23
Decision-making process

Experience Practices
Individual mental models Collective mental models
(self, colleagues, others) (common rules, ratios)

Error/Biases
in decision-making?

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 24
Decision-making process

• Common rule, promotional spending: budget

allocated in proportion of market shares
• Example:

Market Share Spending

Region A 30% 43%

Region B 40% . 57%
100%
• But what if?

Market Share Spending

Region A 0% ?
Region B 100% ?
• Is the rule good enough? If not, why not
change it?
© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 25
Benefits of response models
• Issue: Is company getting the best
efficiency and effectiveness on its
promotional spending across US markets?

• Approach: Used market response analysis

and resource allocation to conclude that
Heinz:
− was misallocating spending
− could substantially reduce overall spending
without sacrificing national market share.

• Results: Reduced promotional spending

40% AND increased market share from
34% to 37%.
© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 26
Benefits of response models

• Issue: Running out of good sites for

typical full-service Marriott hotels.

• Approach: Conjoint analysis to

determine customer preferences, critical
information for hotel design.

• Results – Courtyard by Marriott:

− Fastest growing moderately priced hotel
chain in the United States
− Sales over $1 billion with occupancy rate
above industry average
− Market share improved by +4%
− Created new segment with 5 clone chains
© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 27
Benefits of response models

• Issue: Sales force size expanding

approximately 10% per year: was size
and allocation the best?

• Approach: Used resource allocation

to determine the optimum sales force
size and allocation across products and
markets.

• Results: Implementation increased

profits $24M/year: more than 12% over
plan.
© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 28
Benefits of response models
• Issue: Forecasting the time path of
sales of DirecTV, before introduction.

• Approach: Used Bass diffusion model.

− Market size estimate from customer
survey.
− Diffusion parameters estimated from
managerial judgments and analogous
products (cable TV).

• Results:
− Five year forecasts made 3 years before
launch were, on average, +16% above old
forecasts.
(growth was going to be larger than
anticipated)
− Forecast justified earlier launch of a
satellite for expanded transmission
© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
capability.
for non-commercial educational purposes only is authorized if this notice appears on every copy. 29
Benefits of response models

• Issue: Exelon wanted to differentiate itself

from other second-tier companies.

• Approach: Used positioning analysis to

determine that:
− Customer preferences were more associated
with the characteristics and offerings of second
tier companies, especially price.
− Analysis identified opportunity for re-position
towards more customer focus, reliability, and
value.

• Results: New advertising campaign with the

revised theme increased customer
awareness by 45% in 3 months.
© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 30
Benefits of response models
• Issue: Soliciting donors is expensive,
and donors are more and more
sensitive to wise spending of donation
money.

• Approach: Used customer choice

model to identify best potential donors,
based on past data, and solicit only
those donors who are likely to be
profitable.

• Results: In the first year of

implementation, number of solicitations
decreased by -40%, while donation
amounts increased overall by +9%.
© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 31
Marketing Analytics Increased
Marketing Campaign Effectiveness at
Grainger

• Issue: Small business customer revenue growth flat after

several years of double digit growth.
• Approach: New segmentation and targeting models with “link”
to full universe of customers.
• Results:
− Integrated data base program added over $100 million to
sales, reduced costs by well over $100 million.
− Net profit increase: over $200 million.

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 32
Use of Marketing Analytics Improved
Marketing Productivity at Pfizer

• Issue: Pfizer was considering splitting its sales force

into two product-specialized sales forces to target
effort on 15 products in 80 target segments.

• Approach: Used ReAllocator to determine optimal

size of sales force and optimal deployment.

• Results: 6% profit gain by maintaining current sales

force size / structure and re-deployed effort to
products and markets based on model
recommendations.

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 33
Small Models Example:
Trial/Repeat Model

Share = % Aware x

% Available | Aware x

% Try | Aware, Available x

% Repeat | Try, Aware, Available x Usage Rate

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 34
Trial/Repeat Model

Target Population

Aware?
50%

Available?
80%

Try?
40%

Repeat?
50%

Market Share = ?

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 35
Model Diagnostics

Trial

hi low

hi

Repeat

low

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 36
Trial Dynamics

100% You never get

10 everyone to try
9

6
% Population
5
Trying (Trial)
4

0
0 1 2 3 4 5
Time 6 7 8 9 10

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 37
× Repeat Dynamics

100%
10

7
Note—late triers
6 often do not become
% Repeaters regular users
Among Triers
5
(Repeat)
4

0
0 1 2 3 4 Time
5 6 7 8 9 10

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 38
= Share Dynamics!

10
100%
Fiona ‘the brand
9
manager’ gets promoted
8

7
Steve, her
6 replacement, gets fired
Share =
5
(Trial ×Repeat)
4 John, ‘the caretaker’,
takes over
3

0
0 1 2 3 4 Time
5 6 7 8 9 10

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 39
Model Benefits

• Small models can offer insight – they can change

your goals and priorities, even if they don’t influence
your decisions.

• Even simple models can align management beliefs

with marketing policy.

• You don’t need hard data to get value from models--

judgments and intuition is often enough.

• Digital data capture enables large model ROI.

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 40
The Connected Marketing Analytics
Process
• Profit
• Effectiveness
• Action guidelines/Reports Improve
• Competitive advantage
• What if capabilities Company
• Integration with company Performance
processes

Guide Support
Opportunity • New revenue
Implementation
Identification • Cost reduction
• Productivity gains
Opportunities
for Marketing
Analytics

• Structured process
• Real-time
• Interactive Promote Build an • Models
• Distributed across Reasoned Analytic • Data
the organization Decisions Foundation • Digital infrastructure

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 41
 The Opportunity for Marketing
Analytics
Zone of Death Zone of Exceptional
Wish Marketing Marketing

The Bell Curve for (Below (Above

Marketing Performance Average) (Average Average)
marketing
(Well program) (Well Above
Below Average)
Average)

Marketing Performance Critical Troubling Average Pleasing Amazing

Marketing Share Growth Precipitous Significant Modest Decline Increase Dramatic
Decline Decline Increase

New Product Success Rate 0% 5% 10% 25% 40%+

Advertising ROI Negative 0% 1-4% 5-10% 20%

Promotional Programs Disaster Un-profitable Marginally Unprofitable Profitable Very Profitable

Customer Satisfaction 0-50% 51-65% 66-75% 76-82% 83-90%

Customer Retention (Annual) 0-40% 41-60% 61-80% 81-90% 91-97%

• The objective is very simple.

• To improve marketing performance.

Source: Adapted from Copernicus Marketing Consulting

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 42
 Key Concepts
Behind
Marketing
Engineering

Material designed to accompany the book Principles of Marketing

Engineering and Analytics by Lilien, Rangaswamy and De Bruyn, and the
marketing analytics software Enginius available at http://www.enginius.biz

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any

publication is not authorized. Copying and distribution for non-commercial
educational purposes only is authorized if this notice appears on every copy.
Adaption for non-commercial educational purposes only is authorized if
© Copyright
attribution appears on every copy. DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 43
Return on Investment (ROI)

• ROI is a frequently used metric to evaluate

marketing expenditures (i.e. investments)
• It is given by a simple formula:
($𝑮𝒂𝒊𝒏𝒔 𝒇𝒓𝒐𝒎 𝒊𝒏𝒗𝒆𝒔𝒕𝒎𝒆𝒏𝒕 − $𝒄𝒐𝒔𝒕 𝒐𝒇 𝒊𝒏𝒗𝒆𝒔𝒕𝒎𝒆𝒏𝒕)
𝑹𝑶𝑰 % ∗ 𝟏𝟎𝟎
$𝒄𝒐𝒔𝒕 𝒐𝒇 𝒊𝒏𝒗𝒆𝒔𝒕𝒎𝒆𝒏𝒕
• Example:
(𝟏𝟐𝟎𝟎 − 𝟏𝟎𝟎𝟎)
𝑹𝑶𝑰 = ∗ 𝟏𝟎𝟎 = 𝟐𝟎%
𝟏𝟎𝟎𝟎

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 44
Interpreting ROI numbers

• ROI is a single “dimensionless” metric that can be

applied to any investment. It is simply a metric of
how much money your money earned.

• ROI does not have a time reference. Typical time

frames in marketing are “campaigns”(a few weeks)
and annual.

• The long-term return in the stock market is about 8%

per year. If your marketing investment does better,
you are using your money well!
(http://www.stern.nyu.edu/~adamodar/pc/datasets/histretSP.xls).

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 45
Applying ROI

• Helps you make one-off decisions which require

investments (of time, money, and equivalents): (1)
Does a display ad campaign at Yahoo! provide
adequate ROI? (2) Does getting a liberal arts degree
provide a good ROI?

• Allows you to rank alternative decisions: Should I

advertise in newspapers or Craigslist?

• Companies set ROI hurdles for investment (say 15 –

20%), but ROI metric has been notoriously difficult to
compute for marketing spend.

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 46
Breakeven Analysis: Definitions

• Fixed cost (FC ): Costs that remain the same

regardless of the sales level. Often, these costs are
incurred before the company makes any sale.
• Unit variable cost (c ): the cost incurred to produce
and sell one unit of a product/service.
• Quantity sold (q ): the number or amount of the
product/service sold.
• Variable cost (VC(q)): Total variable cost at a given
sales level (q). This is equal to qc.
• Breakeven quantity (BE ): The sales (units) at which
total revenue equals total costs, i.e. profit is zero.

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 47
Breakeven Analysis: Computation

• 𝑻𝒐𝒕𝒂𝒍 𝒓𝒆𝒗𝒆𝒏𝒖𝒆 = 𝒒𝒖𝒂𝒏𝒕𝒊𝒕𝒚 𝒔𝒐𝒍𝒅 ∗ 𝒑𝒓𝒊𝒄𝒆 = 𝒒𝑷

• 𝑻𝒐𝒕𝒂𝒍 𝒄𝒐𝒔𝒕 = 𝒇𝒊𝒙𝒆𝒅 𝒄𝒐𝒔𝒕 + 𝒗𝒂𝒓𝒊𝒂𝒃𝒍𝒆 𝒄𝒐𝒔𝒕 = 𝑭𝑪 +
𝑽𝑪 𝒒 = 𝑭𝑪 + 𝒄𝒒
• 𝑷𝒓𝒐𝒇𝒊𝒕 = 𝑻𝒐𝒕𝒂𝒍 𝒓𝒆𝒗𝒆𝒏𝒖𝒆 − 𝑻𝒐𝒕𝒂𝒍 𝒄𝒐𝒔𝒕 = 𝒒𝒑 − 𝑭𝑪 −
𝒄𝒒
At Breakeven quantity, profit = 0. We will solve the
profit equation to get BE:
𝟎 = 𝑩𝑬 ∗ 𝒑 − 𝑭𝑪 − 𝑩𝑬 ∗ 𝒄
𝑭𝑪 𝑭𝒊𝒙𝒆𝒅 𝒄𝒐𝒔𝒕
➔ 𝑩𝑬 = =
𝒑−𝒄 𝒖𝒏𝒊𝒕 𝒎𝒂𝒓𝒈𝒊𝒏

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 48
Breakeven Analysis: Example

Fixed Cost[FC] 6000

Unit Variable 10
Cost[C]

Price(Unit 25
revenue) [p]

Sales(Unit) Variable Cost Fixed Cost Total Cost Total Revenue Profit
10 100 6000 6100 250 -5850
20 200 6000 6200 500 -5700
30 300 6000 6300 750 -5550
40 400 6000 6400 1000 -5400
50 500 6000 6500 1250 -5250
60 600 6000 6600 1500 -5100
70 700 6000 6700 1750 -4950
80 800 6000 6800 2000 -4800
90 900 6000 6900 2250 -4650
100 1000 6000 7000 2500 -4500
110 1100 6000 7100 2750 -4350

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 49
Breakeven Analysis: Example

14000

12000

10000
BE = 300 Units
8000
Variable Cost (𝐕𝐂(𝐪))

6000 Fixed Cost (FC)

Total Cost

4000 Total Revenue

2000

0
0 200 BE 400 600

Units Sold

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 50
Breakeven Analysis:
Example: Increase in Fixed Cost
14000
Profit
12000

10000
BE = 400 Units
8000
Variable Cost (𝐕𝐂(𝐪))

6000 Fixed Cost (FC)

Total Cost

4000 Total Revenue

2000

0
0 200 BE 400 600

Units Sold

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 51
Breakeven Analysis:
Interpretation and Application
𝑭𝑪
• Recall 𝑩𝑬 =
𝑼𝒏𝒊𝒕 𝒎𝒂𝒓𝒈𝒊𝒏

Industries in which
“scale of operations” Industries like
Pharmaceuticals and
determine succcess
Aircraft
(e.g. Amazon, United
Hi

Airlines) manufacturing
Fixed cost

Small-scale industries Great industry!

(e.g. local restaurant, (e.g. Management
Lo

taxi service) consulting)

Lo Hi
Unit Margin
© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 52
Opportunity Cost of an Action or
Decision

Opportunity cost: It is the cost you incur by not

pursuing the next best alternative to an action or
decision that you do decide to pursue.
Example: What is the opportunity cost of our social
media campaign (the action we decide to pursue)?
• Profit from social media campaign: $10,000
If the next best alternative is a newspaper campaign
with a profit potential of $12,000, then the opportunity
cost is $2,000 (the amount you forgo). If potential
profit from the newspaper campaign is $8,000, then
you did not forgo any gains – you have already made
the best decision.

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 53
Marketing Engineering

Marketing Environment
Automatic scanning, data entry,
subjective interpretation

Marketing Data
Engineering Database management, e.g..,
selection, sorting, summarization,
report generation
Information
Decision model; mental model

Insights
Judgment under uncertainty,
e.g.., modeling, communication,
introspection
Decisions
Financial, human, and other
organizational resources

Implementation

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 54
Model-Based Analytics are the core
of ME: What is a Model?

• A model is a stylized representation of reality that is

easier to deal with and explore for a specific purpose
than reality itself.

• There are many types of models:

− Verbal

− Box and Arrow

− Graphical

− Mathematical

− Spreadsheets

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 55
Verbal Model

Sales of a new product often start slowly as

“innovators” in the population adopt the product. The
innovators influence “imitators,” leading to accelerated
sales growth. As more people in the population
purchase the product, sales continue to increase but
sales growth slows down.

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 56
Box and Arrow Model

Fixed
Population Size

Innovators Imitators

Innovators
Influence
Imitators
Timing of Purchases by Timing of Purchases by
Innovators Imitators

Pattern of Sales Growth

of New Product

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 57
Graphical Model

Cumulative Sales
of a Product
10

9
Fixed
8 Population Size
7

0
0 1 2 3 4 5 6 7 8 9 Time
10

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 58
New York City’s Weather

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 59
What Do You See?

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 60
Mathematical Model

𝑑𝑥𝑡
= (𝑎 + 𝑏𝑥𝑡 )(𝑁 − 𝑥𝑡 )
𝑑𝑡

xt = Total number of people who have adopted product

by time t

N = Population size

a,b = Constants to be determined. The actual path of the

curve will depend on these constants

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 61
Are Models Valuable?
Models vs Intuition/Judgments
Types of Judgments Experts Had to Make Mental Subjective Objective
Model Decision Decision
Model Model

Academic performance of graduate students 0.19 0.25 0.54

Life expectancy of cancer patients -0.01 0.13 0.35

Changes in stock prices 0.23 0.29 0.80

Mental illness using personality tests 0.28 0.31 0.46

Grades and attitudes in psychology course 0.48 0.56 0.62

Business failures using financial ratios 0.50 0.53 0.67

Students’ rating of teaching effectiveness 0.35 0.56 0.91

Performance of life insurance salesman 0.13 0.14 0.43

IQ scores using Rorschach tests 0.47 0.51 0.54

Mean (across many studies) 0.33 0.39 0.64

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 62
Applicant Profile
(Academic performance of graduate students)

Applican Personal Selectivity of Undergra- CollegeGrade Work GMAT GMAT

t Essay Undergra- duate Major Avg. Experien Verbal Quantitati
duate Institution ce ve

1 poor highest science 2.50 10 98% 60%

2 excellent above avg business 3.82 0 70% 80%

3 average below avg. other 2.96 15 90% 80%

• • • • • • • •

• • • • • • • •

117 weak least business 3.10 100 98% 99%

118 strong above avg other 3.44 60 68% 67%

119 excellent highest science 2.16 5 85% 25%

120 strong not very business 3.98 12 30% 58%

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 63
Some Takeways
• Seat-of-the-pants decision making is not enough any
more. Marketing Analytics is becoming increasingly
important for all types of businesses, especially for
supporting core decisions.

• Analytics generate results and insights for action.

• For important issues, both analytics and judgment

(Humans + Computers) are needed for translating results
and insights into decisions and actions, especially in
marketing.

• The process and discipline associated with analytics, by

themselves, offer valuable benefits.

• Merely quantifying judgment can improve the quality of

marketing decisions.
© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 64
 OTHER SLIDES

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 65
What is a model?

A model is a stylized representation of reality that is easier

to deal with and explore for a specific purpose than reality
itself.

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 66
What is a model?

A model is a stylized representation of reality that is easier

to deal with and explore for a specific purpose than reality
itself.

• “Models are not to be trusted, they are to be used.”

• “No model is true, but some models are useful.”

• “Models do not make decisions. Managers do.”

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 67
What does a model look like?

Verbal Box and Arrow Graphical Mathematical

“Sales is a function Sales

Advertising
of customers’ 10

S = W  D  A1/2
awareness,
8

distribution, and 4

advertising”
2

Sales
Advertising
0 1 2 3 4 5 6 7 8 910

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 68
What kinds of models?

Descriptive Predictive/Normative
E.g., how are my products and my competitors’ E.g., which customers should I target?
products perceived by the market? Gaining (given objectives & constraints)
insight in the process…

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 69
What benefits?

• Gain additional insight

− Even small models can offer insight – they can change your goals, priorities
and mental models, even if you stop using them

• Explore more options

− De-anchoring, simulations, “what if” scenarios, etc.

• Help reach group consensus, support group decisions

− Avoiding the battle of faiths

• Make more consistent, better decisions

− Humans are inconsistent, but “creative”. Models are consistent information
processors. Managers + Models = (possibility of) better decisions
− Models do not make decisions. Managers do

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 70
How to build a model

• Specify
− Variables (which ones to include)
− Relationships, interactions, dynamics (how they are linked)

• Calibrate
− Statistical estimation with real data (econometric approach)
− Judgmental calibration (tribal wisdom approach)

• Validate
− Global fit (R², model fit)
− Variable significance (correct signs, t-tests)
− Face validity (does it make sense?)

• Apply
− Unique vs. multiple objectives?
− Short term vs. long term?

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 71
Response Models

• Aggregate response models

• Individual response models

• Shared-experience models

• Qualitative response models

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 72
The Concept of a Response Model

Idea: Marketing Inputs:

• Selling effort
• Advertising spending
• Promotional spending

Market

Marketing Outputs:
• Sales
• Share
• Profit
• Awareness, etc.

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 73
Input-Output Model

Marketing Actions Competitive Actions Observed Market

Inputs Outputs
02

Product design
Price Market Awareness level
Advertising Response Preference level
Model Sales Level
Selling effort etc. 01 04

03

Environmental
Conditions

Control Adaption Evaluation

Objectives
06 05

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 74
Objectives Specified in Models

• Profit
(= Sales x Margin – Costs)

• Sales

• Market share

• Time horizon

• Uncertainty

• Multiple goals

• Multiple points of view

• Others ??

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 75
Response Function

10

9 Max
8
Sales 7
Response Response
6
Function
5
Current
Sales
4

2 Min
1

0
0 1 2 3 Current
4 Effort
5 6 7 8 9 10

Effort Level

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 76
A Simple Model

10

Y (Sales Level)
6

5
b (slope of the
4 salesline)
a
(sales level when3
1
advertising = 0)
2

0
0 1 2 3
X 4(Advertising)
5 6 7 8 9 10

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 77
Phenomena

P1: Through Origin P2: Linear

10 10

Y5 Y5

0 0
0 1 2 3 4 X5 6 7 8 9 10 0 1 2 3 4 X5 6 7 8 9 10

P3:Decreasing Returns P4: Saturation

(concave) 10

10 8
6
Y4
Y5
2
0 0
0 1 2 3 4 X5 6 7 8 9 10 0 1 2 3 4 X5 6 7 8 9 10

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 78
Phenomena

P5:Increasing Returns (convex) P6: S-shape

10 10
8 8
6 6
Y4 Y4
2 2
0 0
0 1 2 3 4 X5 6 7 8 9 10 0 1 2 3 4 X5 6 7 8 9 10

P7: Threshold P8: Super-saturation

10
8
6
Y
Y4
2
0
X 0 1 2 3 4 X5 6 7 8 9 10

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 79
Aggregate Response Models:
Linear Model

Y = a + bX

• Linear/through origin

• Saturation and threshold (in ranges)

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 80
Aggregate Response Models:
Fractional Root Model

Y = a + bXc

c can be interpreted as elasticity when a = 0.

Linear, increasing or decreasing returns (depends on c ).

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 81
Aggregate Response Models:
Exponential Model

Y = aebx; x > 0

Increasing or decreasing returns (depends on whether

b is positive or negative).

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 82
Aggregate Response Models:
Modified Exponential Model

Y = a (1 – e–bx) + c

Decreasing returns and saturation.

Widely used in marketing.

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 83
Aggregate Response Models:
Adbudg Function

Y = b + (a–b)

S-shaped and concave; saturation effect.

Widely used.

Amenable to judgmental calibration.

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 84
Aggregate Response Models:
Multiple Instruments

Additive model for handling multiple marketing

instruments

Y = af(X1) + bg(X2)

Easy to estimate using linear regression.

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 85
Aggregate Response Models:
Multiple Instruments cont’d
• Multiplicative model for handling multiple marketing
instruments

𝑏 𝑐
𝑌= 𝑎𝑋1 𝑋2

b and c are elasticities.

Widely used in marketing.

Can be estimated by linear regression (by taking

logarithms on both sides of the equation).
© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 86
Dynamic Effects

1. Marketing Effort
e.g., sales promotion

Spending Level

Time

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 87
Dynamic Effects

2. Conventional “delayed response” and “customer

holdout” effects

Sales Response

Time

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 88
Dynamic Effects

3. “Hysteresis” effect

Sales Response

Time

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 89
Dynamic Effects

4. “New trier”
“wear out” effect

Sales Response

Time

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 90
Dynamic Effects

5. “Stocking” effect

Sales Response

Time

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 91
Aggregate Response Models:
Dynamics

• Dynamic response model

Yt = a0 + a1 Xt +  Yt–1
carry-over
effect
current
effect

Easy to estimate.

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 92
Aggregate Response Models:
Market Share
• Market share (attraction) models

Ai
Mi = ––––––––––––––––––
A1 + A2 + . . . + An

Ai = attractiveness of brand i.

Satisfies sum (market shares sum to 1.0) and range constraints

(brand share is between 0.0 and 1.0)

Has “proportional draw” property.

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 93
Individual-Level Response Models:
Requirements

• Satisfies sum and range constraints.

• Is consistent with the “random utility” model.

• Has the “proportional draw” property.

• Widely used in marketing.

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 94
Individual-Level Response Models
MNL
• The multinomial logit model can be used to represent
“probability of choice.” The individual’s probability of
choosing brand 1 is given by (similar equations can be
developed for other brands in the consideration set of
consumers):

eA1
Pi 1 = ––––
e Aj
j
where Aj =  bk Xijk and
k
bk are parameters representing importance weights that
are to be estimated from data (i represents consumer, j
represents brand, and k represents marketing variable).

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 95
An Important Implication of the Logit
Model
10
High
9

7
Marginal 6
Impact of a
5
Marketing
4
Action
3

1
Low0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

Probability of Choosing an Alternative

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 96
Attribute Ratings per Store

Store Performance Quality Variety Value

1 0.7 0.5 0.7 0.7

2 0.3 0.4 0.2 0.

3 0.6 0.8 0.7 0.4

4 (new) 0.6 0.4 0.8 0.5

Importance
2.0 1.7 1.3 2.2
Weight

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 97
Shares per Store
(Illustrates Proportional Draw
Property)
(a) (b) (c) (d) (e)

Share Share
Draw
eAj estimate estimate
Store Aj = bk Xjk (c)–(d)
without with new
new store store

1 4.70 109.9 0.512 0.407 0.105

2 3.30 27.1 0.126 0.100 0.026
3 4.35 77.5 0.362 0.287 0.075
4 4.02 55.7 0.206

The new store draws share from each existing store proportional to that
store’s market share (subscript i to represent individual is omitted).

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 98
Response models in the decision
loop…

Marketing Actions Competitive Actions Observations

Inputs (outputs)

Product design,
Price, Awareness
Response
Advertising Preferences
Model
Sales
Selling effort

Environmental
Conditions

Control Adaption Evaluation

Objectives

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 99
A (simple) example of response
model
10

9 Max
8
Sales 7
Response Response
6
Function
5
Current
Sales
4

2 Min
1

0
0 1 2 3 Current
4 Effort
5 6 7 8 9 10

Effort Level

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 100
Shared Experience Models

• Base the response model on behavior observed at

other leading firms (i.e., this results in a “benchmark”
response function).

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 101
Qualitative Response Models

• Rules to capture qualitative response:

The retailer will accept the trade deal, but what he does
with it is based on coop advertising dollars. If the deal
includes coop money, the retailer will accept the deal
and pass on all of the discount to the consumer. If the
discount is greater than 30 percent, he will put up a big
display. Otherwise, the retailer leaves the item at regular
price and does not use an ad feature or a display

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 102
Model Benefits

• Small models can offer insight – they can change your

goals and priorities, even if they don’t influence your
decisions.

• Even simple models can align management beliefs

with marketing policy.

• You don’t need hard data to get value from models--

judgments and intuition is often enough.
• Digital data capture enables large model ROI.

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 103
We Focus on End-User Models

End-User Models High-End Systems

Scale of problem Small/Medium Small/Large

Time Availability
Short Long
(for setting up model)

Costs/Benefits Low/Medium High

User Training Moderate/High Low/Moderate

Technical Skills Low/Moderate High

Recurrence of
Low Low or High*
problem

*Low for one-time studies. High for models in continuous use

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 104
Are Models Valuable?
Belief: ‘No mechanical prediction method can
possibly capture the complicated cues and
patterns humans use for prediction.’

Hard Fact: A host of studies in medical diagnosis, loan

granting, auditing and production
scheduling have shown that even simple
models out-perform expert judgement.

Example: Bowman and Kunreuther showed that

simple models based on managers’ past
behavior, (in terms of production
scheduling and inventory decisions) out-
perform the managers themselves in the
future.
© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 105
Are ‘Models’ the Whole Answer? No!
The widespread availability of statistical packages has put
mathematical bazookas in the hands of those who would be
dangerous with an abacus.
—Barnett

To evaluate any decision aid, you need a proper baseline.

1. Intuitive judgement does not have an impressive track

record.
2. When driving at night with your headlights on you do not
necessarily see too well. But turning them off will not
improve the situation.
3. ‘Decision aids do not guarantee perfect decisions but when
appropriately used they will yield better decisions on
average than intuition.’
—Hogarth, p.199

© Copyright DecisionPro, Inc. 2018. Commercial distribution and any publication is not authorized. Copying and distribution
for non-commercial educational purposes only is authorized if this notice appears on every copy. 106