1. Brand Positioning – Attributes
February 1, 2005
Cola Wars: Pepsi versus Coca Cola
1. What attributes and benefits do soft-drinks
offer? Which are physical and which are
psychological? Which are "essential" and
which are consumers willing to be flexible on?
2. Why did people react so strongly when New
Coke was introduced?
3. Comment on the design of the New Coke
study – what would you have changed?
Case Discussion
1
2. STRATEGIC MARKET MARKET BRAND
QUESTIONS DEFINITION SEGMENTATION POSITIONING
What geographic area e.g., domestic US
will we operate in?
What timeframe have e.g., calendar 2003, 4
we in mind?
Brand with attributes
What will we offer? Some version of product responsive to DCCg
category X Identify kinds of
To whom will we offer Define prospects demand-creating condition Prospects who experience
it? e.g., people who buy/use (DCCa - n); DCCg, i.e., targets
product category X or Assess state of want-
engage in corresponding satisfaction for each kind;
activity Y. Order DCC with a view to Specific price (e.g., $1.79)
roi and choosing DCCx to
What price will we State broad price band target
charge? (e.g., discount, premium)
How will we let them State tentative ideas for Media vehicles,
know about our media type Attentional strategy,
offering? Brand claims
How/where will we State tentative ideas for Specific exchange venues
engage in exchange exchange venue type
with them?
Whom will we compete State competitors implicated Competitors
with? by choices on other Implicated by DCCg
dimensions
Conceptualizing User Wants
Personal Systems Brands As
Perceived
(Brand Beliefs);
Motivating Desired Brand
Action,
Conditions Attributes Consideration
Including
Set;
Brand Use
Brand
Environmental Preference
Systems Ordering
Ex-Ante
Analysis
Ex-Post
Analysis
Theory
2
3. How to Learn
What Customers Want?
Ask Direct Questions about preference:
What brand do you prefer?
What Interest Rate would you like?
What Annual Fee would you like?
What Credit Limit would you like?
Answers often trivial and unenlightening (e.g.
respondents prefer low fees to high fees, higher
credit limits to low credit limits)
Conjoint Overview
Introduction
Analysis for Brand Positioning
Conjoint analysis is a regression based
technique where:
– you measure consumers’ preferences,
– by confronting them with different product
profiles that contain tradeoffs and
– letting them rate their relative preferences.
Dependent variable: Relative Preference
Statements
Independent variables: Attributes
Conjoint Analysis
3
4. Basic Assumption
Product is a bundle of attributes who importance is
driven by customer needs.
– Should we offer our business travelers more room
space or a fax machine in their room?
– Should we offer more leisure-time activities (sauna,
exercise room, tennis courts) or more food related
services (several dining options, vending machines,
in-room kitchen facilities)?
– Should we use a steel or aluminium casing to
increase customer preference for our new extrusion
equipment?
Conjoint Analysis
Outboard Engine Attributes
Brand Name Durability
Evinrude, Force, Vibration and Noise
Johnson, Mariner,
Mercury, Yamaha
Acceleration
Price Speed
Fuel Economy Emissions
Reliability Technology
Conjoint Analysis
4
5. Credit Card Attributes
Bank Annual Fee
current, local, out-of- $50, $25, $0
state Rebates
Interest Rates 1%, .5%, none
8%, 7%, 6% , variable Credit Line
Reward Programs $2000, $5000
free checking, greater Grace Period
interest on savings
45 days, none
Conjoint Analysis
Mathematical Representation
y = x1β1 + x2β2 + x3β3 + ... + ε
or
y = Xβ + ε
where
y = the overall value of the product
X = product attributes
β = the value placed on the attributes
Conjoint Analysis
5
6. How Does Conjoint Analysis
Work?
We vary the product features (independent
variables) to build many (usually 12 or more)
product concepts
We ask respondents to rate/rank those product
concepts (dependent variable)
Based on the respondents’ evaluations of the
product concepts, we figure out how much
unique value (utility) each of the features added
(Regress dependent variable on independent
variables; betas equal part worth utilities.)
Conjoint Overview
Conjoint Analysis
First Step: Create Attribute List
Attributes assumed to be independent (Brand,
Speed, Color, Price, etc.)
Each attribute has varying degrees, or “levels”
Brand: Coke, Pepsi, Sprite
Speed: 5 pages per minute, 10 pages per minute
Color: Red, Blue, Green, Black
Each level is assumed to be mutually exclusive of
the others (a product has one and only one level
level of that attribute)
Conjoint Attributes
6
7. Rules for Formulating
Attribute Levels
Levels are assumed to be mutually exclusive
Attribute: Add-on features
level 1: Sunroof
level 2: GPS System
level 3: Video Screen
If define levels in this way, you cannot determine the
value of providing two or three of these features at the
same time
Conjoint Attributes
Traditional Conjoint:
Full Profile Analysis
Using a 100-pt scale where 0 means definitely
would NOT and 100 means definitely WOULD…
How likely are you to purchase…
1997 Honda Accord
Automatic transmission
No antilock brakes
Driver and passenger airbag
Blue exterior/Black interior
$18,900
Your Answer:___________
Traditional Conjoint
7
8. Strengths of Traditional Conjoint
• Good for both product design and pricing
issues
• Can be administered on paper,
computer/internet
• Shows products in full-profile, which many
argue mimics real-world
• Can be used even with small sample sizes
Traditional Conjoint
Credit Card Attributes
Brand
MasterCard, Visa, Discover
Interest Rate
18%, 15%, 12%
Annual Fee
None, $10, $20
Credit Limit
$1000, $2500, $5000
Conjoint Data
8
9. How much do you like this credit card
offering? 0 = Terrible, 10 = Excellent
MasterCard
18% interest
No annual fee
$1000 credit limit
Score= ___________
(1)
Conjoint Data
How much do you like this credit card
offering? 0 = Terrible, 10 = Excellent
MasterCard
12% interest
$20 annual fee
$5000 credit limit
Score= ___________
(2)
Conjoint Data
9
10. How much do you like this credit card
offering? 0 = Terrible, 10 = Excellent
Visa
18% interest
$10 annual fee
$5000 credit limit
Score= ___________
(3)
Conjoint Data
How much do you like this credit card
offering? 0 = Terrible, 10 = Excellent
Discover
18% interest
$20 annual fee
$2500 credit limit
Score= ___________
(4)
Conjoint Data
10
11. How much do you like this credit card
offering? 0 = Terrible, 10 = Excellent
Discover
12% interest
$10 annual fee
$1000 credit limit
Score= ___________
(5)
Conjoint Data
How much do you like this credit card
offering? 0 = Terrible, 10 = Excellent
Discover
15% interest
No annual fee
$5000 credit limit
Score= ___________
(6)
Conjoint Data
11
12. How much do you like this credit card
offering? 0 = Terrible, 10 = Excellent
Visa
12% interest
No annual fee
$2500 credit limit
Score= ___________
(7)
Conjoint Data
How much do you like this credit card
offering? 0 = Terrible, 10 = Excellent
MasterCard
15% interest
$10 annual fee
$2500 credit limit
Score= ___________
(8)
Conjoint Data
12
13. How much do you like this credit card
offering? 0 = Terrible, 10 = Excellent
Visa
15% interest
$20 annual fee
$1000 credit limit
Score= ___________
(9)
Conjoint Data
Part-Worths of Attribute-Levels
Some product attributes are “defining”
All credit cards have a brand name, interest rate and
credit limit.
Not possible to estimate importance of a defining
attribute, only importance of changes in the attribute-
levels (e.g., 12% vs. 15% interest rate).
Other product attributes are “optional”
Apartments need not have balconies, so a natural zero
“0” can occur for the attribute.
Implication: One level of each attribute is not
estimable
Conjoint Data
13
15. Multiple Regression
y = β0 + β1x1+ … + βkxk + ε
βi = the expected change in the dependent
variable (y) for a unit change in xi, holding fixed
the other independent variables (x’s).
Allows for statistical adjustments and couter-
factual analysis.
Provides a measure of the effect of xj relative to xk.
Regression Analysis
Regression Results
Rating = 4.00 + 2.00 Visa + 0.00 Discover
+1.00 15% + 1.00 12%
- 1.00 $10 - 1.00 $20
+ 2.00 $2,500 + 3.00 $5,000
Default Card: MasterCard; 18% interest;
No annual fee; $1000 credit limit.
All importances (part-worths) measured relative to these levels.
Part-Worth Analysis
15
16. Analysis of Part-Worths
Brand Name:
MasterCard = 0
Visa = 2
Discover = 0
Implication: MasterCard and Discover
valued equally. Visa valued more.
Part-Worth Analysis
Analysis of Part-Worths
Interest Rate:
18% = 0
15% = 1
12% = 1
Implication: Consumers are indifferent to a
15% or 12% rate, but both are preferred to
an 18% rate.
Part-Worth Analysis
16
17. Analysis of Part-Worths
Annual Fee:
None = 0
$10 = -1
$20 = -1
Implication: Consumers are indifferent to a
$10 or $20 annual fee, but prefer no
annual fee to either amount.
Part-Worth Analysis
Analysis of Part-Worths
Credit Limit:
$1000 = 0
$2500 = 2
$5000 = 3
Implication: Consumers prefer a $5000 to either a
$2500 or $1000 limited. The change from
$1000 to $2500 is twice as important as the
chance from $2500 to $5000.
Part-Worth Analysis
17
18. Analysis Across Attributes
Brand Name: Interest Rate: Annual Fee: Credit Limit:
MasterCard = 0 18% = 0 None = 0 $1000 = 0
Visa = 2 15% = 1 $10 = -1 $2500 = 2
Discover = 0 12% = 1 $20 = -1 $5000 = 3
Value of Visa vs. MasterCard = Value of $2500 vs. $1000 Credit Limit
Value of 15% vs. 18% Interest = Value of $0 vs. $10 Annual Fee
Value of $5000 vs. $2500 Credit Limit = Value of 15 vs. 18% Interest
Part-Worth Analysis
Attributes Play Two Roles
Desired attributes are the basis of consideration
set formation and preference orderings.
Market Defining
Attributes that form consideration sets are market
defining – without them there is no hope of attracting
a prospect.
Brand Positioning
Other attributes play a role in brand positioning –
consumers are willing to tradeoff one attribute level
for another.
Theory
18
19. Screening Rules
“A Choice Model with Conjunctive, Disjunctive, and
Compensatory Screening Rules”
Conjunctive rule: alternative must be acceptable
on all attributes.
Disjunctive rule: alternative must be acceptable on
a least one attribute.
Compensatory rule: really good attributes can
make up for really bad attributes.
Challenge Paper
A partial listing of offerings and features of digital cameras
from Nov. 2002 Consumer Reports
A selection of tested products (Within
types, in performance order)
Brand & model
3- TO 5-MEGAPIXEL CAMERAS
Sony DSC-F707 MS 5x · · · ·
Canon PowerShot G2 CF 3x · · · ·
Olympus Camedia C-3040 Zoom SM 3x · · ·
Olympus Camedia D-40 Zoom SM 2.8x
· ·
Fujifilm FinePix F601 Z SM 3x · · ·
Sony Cyber-shot DSC-S75 MS 3x · · · ·
HP PhotoSmart 812 SD 3x
·
Kodak EasyShare DX4900 CF 2x ·
Olympus Camedia E-10 CF, SM 4x · · ·
Canon PowerShot S40 CF 3x · ·
Casio QV-4000 CF 3x · · · ·
Kyocera Finecam S4 SD 3x · · ·
Panasonic Lumix DMC-LC5 SD 3x · · · ·
Sony Cyber-shot DSC-P71 MS 3x
· · ·
Sony CD Mavica MVC-CD400 CD-R/RW 3x
· · ·
Minolta Dimage S404 CF 4x
· ·
Toshiba PDR-3310 SD 3x · · ·
Kyocera Finecam S3 SD 2x · · ·
Challenge Paper
Source: http://www.consumerreports.org/main/detailv2.jsp?CONTENT%3C%3Ecnt_id=163145&FOLDER%3C%3Efolder_id=162691&bmUID=1034956296813
19
20. Empirical Application
Alternatives are described on 8 attributes
Attribute Levels Attribute Levels
Basic body style Low Zoom None
and standard Medium 2x
features High 4x
Mid-Roll None Viewfinder Regular
Change* Manual Large
Automatic Camera Settings None
Annotation* None Feedback LCD
Pre-Set Viewfinder
Customized List Both
Custom List 1 Price $41 - $499
Custom List 2
Custom List 3
*Attributes unique to APS cameras at
Camera Operation No the time of the study
Feedback* Yes
Challenge Paper
Alternative Approach
Traditional choice model:
Pr(i) = Pr(xi’β + εi > xk’β + εk for all k)
Proposed choice model:
Pr(i) = Pr(xi’β + εi > xk’β + εk for all k
such that I(xk,γ) = 1)
Threshold levels
Challenge Paper
20
21. Empirical Application – Conjunctive Model
Proportion Screening on Each Attribute
45%
40% 41%
40%
35%
30%
25% 22%
19%
20%
15%
10% 8%
6% 5%
4% 4% 4% 4%
5%
0%
om
e*
*
*
*
*
*
r
e
s
e
de
n1
n2
n3
n4
ck
ng
yl
ic
ng
St
Pr
ba
Zo
fin
tio
tio
tio
tio
tti
ha
Se
dy
ew
ed
ta
ta
ta
ta
C
no
no
no
no
Bo
Fe
Vi
l
ol
An
An
An
An
p.
-R
O
id
M
*Attributes unique to APS offerings at time of study. Challenge Paper
Empirical Application – Conjunctive Model
Distribution of Price Threshold
70%
59%
60%
50%
41%
40%
30%
30%
20%
20%
10%
10%
0%
<$276 <$356 <$425 <$499 >$499
• Respondents who “almost always” take their camera when attending “social
gatherings away from” home have higher price thresholds.
Challenge Paper
21
22. Empirical Application
Findings
• Conjunctive model fits data best
• 92% of respondents are forming consideration sets
• Results show which attributes, which levels, and who is
screening
• Respondents tend to screen alternatives using familiar
attributes
• After controlling for the their effect in forming consideration
sets, screening attributes are “less important” than indicated by
the standard choice model
Challenge Paper
Market Simulations
Make competitive market scenarios and
predict which products respondents would
choose
Accumulate (aggregate) respondent
predictions to make “Shares of
Preference” (some refer to them as
“market shares”)
Market Simulators
22
23. Which Color is Preferred?
Consider the following utilities:
Blue Red Yellow
Respondent #1 50 40 10
Respondent #2 0 65 75
Respondent #3 40 30 20
---- ---- ----
Average: 30 45 35
Red has the highest average preference
But, does any one respondent prefer red?
Market Simulators
Which Color is Chosen?
Assume each respondent “chooses” preferred color:
Blue Red Yellow Choice
Respondent #1 50 40 10 Blue
Respondent #2 0 65 75 Yellow
Respondent #3 40 30 20 Blue
---- ---- ----
Average: 30 45 35
Blue “chosen” twice, Yellow once
Market Simulators
23
24. Why Conduct Market
Simulations?
• Simulations better reflect real-world behavior
– Represent idiosyncratic preferences of segments and
individuals (remember, you don’t have to appeal to
the “fat” part of the market to carve out a profitable
business)
• A “choice laboratory” for testing multitude of real-
world possibilities
• Results expressed in terms that make sense to
management and are actionable
Market Simulators
24