View
535
Download
0
Category
Preview:
Citation preview
Brand AnalyticsCase Study on Brand Positioning
D3M
Understand your Industry
o Who are the key players Ownership structure (Illusion of Choice!)
Market share of brands
Price/quality tiers Generics/store brands
Marketing Mix (Promotion/AD)
o Information sources Depends on the Industry (e.g. Comscore, IRI, Nielsen, IMS Health)
What we want Insights into consumer decision making
Example: Why do we form loyalties? How do we decode the black box? Elicit
preferences/decision rules? Simply ask people (Stated preferences) Observe what people do and reverse engineer to derive
underlying preferences or mechanisms (Revealed preference)
Experimentation
Understand your BrandCustomers+ Competitors
o Some form of 80/20 analysiso Who are the top customers?
Demographics Location Behavior/Life style (what else do they buy, what
Magazine/Sports/TV shows)o Competition
o Elasticity (Own & Cross)o Brand perceptions
Approach 1: Competitive Analysis using Revealed Preference Data
log𝑞𝐴=𝛽0 𝐴+𝛽𝐴𝐴 log 𝑃 𝐴+𝛽𝐴𝐵 log 𝑃𝐵+𝛽𝐴𝐶 log 𝑃𝐶+𝛽𝐴𝐷 log 𝑃𝐷+¿ 𝜀𝐴¿
Own price elasticity Cross price elasticities
Approach 2: Elicit Brand Perceptions by asking questions
Stated Preference Data
Example: Brand Perceptions
Overall
S3
S2
S1
Poor_value
Avant_Garde
Successful
Economical
Common
Hi_prestige
Easy_ServiceRoomy
Uncomfortable
Sporty
Interesting
Poorly_built
Unreliable
Quiet
Attractive
Mercury Capri
BMW 318i
Pontiac Firebird
Saab 900
Honda PreludeEagle Talon
Toyota Supra
Audi 90
Ford T-BirdG20
8
Brand StrategiesPositioning
9
.
.
.
What is a Product?
10
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
....
Engineering versus Perceptual Attributes
MPG Fuel Efficiency
Air Bags Safety
Conjoint MDS
11
Differentiation and Positioning
• Differentiation: “The creation of tangible or intangible differences on one or two key dimensions between a focal product and its main competitors”– How do retailers differentiate?– How do airlines differentiate?
• Positioning: “The set of strategies that firms develop and implement to ensure that the differences occupy a distinct and important position in the minds of consumers”
– Example: Auto Rentals– Positioning an issue: CO2
http://www.youtube.com/watch?v=7sGKvDNdJNA
12
Cola Wars
Cub/Omni Jewel TI7 UP 10.9% 11.8% 9.1%
COKE 30.5% 35.3% 51.2%PEPSI 28.9% 25.9% 15.4%
R C 10.9% 4.3% 2.2%CRUSH 1.2% 2.7% 1.8%
DR PEPPER 3.5% 4.0% 3.5%MOUNTAIN DEW 1.5% 2.0% 1.1%
DIET RITE 6.1% 5.6% 10.0%SLICE 1.1% 1.6% 1.3%
SPRITE 2.6% 3.2% 3.7%SUNKIST 2.7% 3.8% 0.7%
17
Product Line
18
Other Issues in Positioning
Me Too Positioning
Strong Positioning: Activity
Managing Your Image
19
A good positioning strategy requires …An understanding of the
dimensions along which the consumer perceives the product
Knowing how competitors’ products are perceived along these dimensions
Identifying the gaps that your product can fill
20
Creating Perceptual Maps in R
Overall
S3
S2
S1Poor_value
Avant_Garde
Successful
Economical
Common
Hi_prestige
Easy_Service
Roomy
UncomfortableSporty
Interesting
Poorly_built
Unreliable
Quiet
Attractive
Mercury Capri
BMW 318i
Pontiac Firebird
Saab 900
Honda PreludeEagle TalonToyota Supra
Audi 90
Ford T-Bird G20
22
Older Techniques for Brand SimilarityPlease rate the following pairs of toothpaste brands on the basis of their similarity (1 = very similar, 9 = very dissimilar).
Very Very
Similar Dissimilar1. Aqua-Fresh vs Crest 1 2 3 4 5 6
72. Aqua-Fresh vs Colgate 1 2 3 4 5 6 7 … 45. Pepsodent vs Dentagard 1 2 3 4 5 6 7
Aqua-Fresh Crest Colgate Aim Gleem Macleans Ultra Brite Close-Up Pepsodent DentagardAqua-FreshCrest 3Colgate 2 1Aim 4 2 2Gleem 6 5 4 3Macleans 5 5 4 4 3Ultra Brite 6 6 6 5 3 3Close-Up 6 6 6 6 2 3 2Pepsodent 6 6 6 6 2 2 1 2Dentagard 7 6 4 6 4 5 5 4 5
Average of which brand pairs are considered most (dis)similar?
23
Data on Attributes & PreferencePopular
with men
Popular with
womenGood Value Heavy Full
BodiedSpecial Occasio
nOn a
Budget
Bud 4 6 7 2 2 3 7Beck’s 7 3 4 3 5 5 3. . . . . . . .. . . . . . . .. . . . . . . .Stroh’s 3 2 3 6 5 5 2
Respondent 1
Overall RatingBud 6Beck’s 9
.
.
.Stroh’s
3
Your overall rating for each Beer: 1 2 3 4 5 6 7 8 9
Rating of Brands on different attributes
24
Input to Factor Analysis
Vectors of attributes can be plotted based on factor loadings. Individual brand’s location on the perceptual map is based on factor scores.
Heavy Pop/Men Pop/Women Full Bodied Blue Collar Good Value Spec Occ
Beck'sBudweiser
Coors Ratings of the brands on each attributes
averaged across All RespondentsCoors lightHeineken
Meister BrauMichelob
MillerMiller Lite
Stroh's
BAV Database
Brand “Personality”(Click on Article to see the paper)
Comprehensive Data on Top 700 Brands in the US(Click on Article to see the paper & download data)
Perceptual AttributesHigh Correlation but Difficult to see/analyze
Conduct Factor analysis
V1 V2 V3 V4 V5 V20…..
Cluster Analysis
(Group Subjects)
Factor Analysis
(Group Variables)
Data
Interpreting the Output
We are not capturing several attributes well. These are somewhat unique, not correlated with other
If we use 9 Factors rather than 52 attributes we capture about 72% of total information
Factors are arranged in terms of proportion of variance explained
Factor Analyze the Data to Understand the Correlation Structure
Notice that some of the variables that had high “uniqueness” are not correlated with the Factors. If these were important in our context, we will keep them as individual variables.
Labels of Factors is SubjectiveFactor 1: “Best Brand”Factor 2: Innovative/VisionaryFactor 3: PrestigiousFactor 4: Fun/FriendlyFactor 5: CaringFactor 6: StylishFactor 7: DifferentFactor 8: EnergeticFactor 9: ??
Interpret The factors
It is our job to interpret what these underlying “factors/themes” are
Go down each column and look for large positive or negative numbers These are correlations between original variables and
the “Factors” Large numbers help us interpret what these underlying
Factors are Note that R has created 9 new variables “Scores”
The new Variables (Scores) are(1) Standardized: They have mean of 0 and std. deviation of 1(2) Uncorrelated with each other
Using New Variables
• Run a regression of “Brand Asset” on the 9 FactorModel 1
(Intercept) 51.13 (0.24)***
Factor1 22.37 (0.24)***
Factor2 8.16 (0.25)***
Factor3 -1.82 (0.25)***
Factor4 5.38 (0.25)***
Factor5 2.87 (0.26)***
Factor6 -0.86 (0.25)***
Factor7 -4.90 (0.26)***
Factor8 1.98 (0.26)***
Factor9 -4.52 (0.27)***
R2 0.76Adj. R2 0.76Num. obs. 3669***p < 0.001, **p < 0.01, *p < 0.05
Clustering Brands on Factor Scores
Segments 3 & 1 are composed of Best Brands
Perceptual Maps are Usually Made on Factor 1 & 2
Segmentation of Brands in BAV (2012Q1) data
Recommended