SlideShare a Scribd company logo
1 of 75
MRS Advanced Analytics
Innovation Symposium
30th April 2015
#MRSlive
Brand Share & Industry Size: Will
the twain ever meet?
Using a portfolio of techniques to improve
accuracy of market volume evolution in
price change scenarios
April 2015
Sreeram Srinivasan, IMRB International
Ranjana Gupta, IMRB International
30
40
Marketer’s pricing dilemma
30 Market
volum
e 100
bn
Marke
t
volum
e
70 bn
With new prices, my share will
grow…
… but will my volume
also grow?
Market share grows….. ….but, market size shrinks
Both the consumer and the macro variables needed to
answer the questions..
C
o
n
s
u
m
e
r
M a c r
o
Brand choice
Same brand, switching etc.
In context of price
Category choice
Frequency, consumption,
substitutes etc.
Industry
Past volumes,
substitutes, prices etc.
Economy
Income, affordability,
Inflation etc.
Relies on past trends
Future may be different
Future oriented.
Past learnings not
fully leveraged
Hence for accurate volume forecasts, no single
methodology can provide complete answer
Consumer research and macro-economic models should
be integrated
Choice Model
Consumer Research
• Brand shares
• Switching
Econometric
Macro Modelling
• Market size
An approach to integration of
choice and econometric model
However, before integration, the individual
tools need to be refined, adapted…
…to account for the nature of the category
How to account for occasion?
Is it an impulse or
considered purchase?
Is it a repertoire or non-
repertoire category?
Do number of units matter?
What about frequency of
purchase?
How can we ensure that the
respondent reacts to only
relevant offers?
Adapting the Choice Model
The questioning technique uses FORCE principle
to make the consumer response more realistic…
Familiar
Only brands that the respondent interacts with / likely to interact with shown –
customized for each respondent real time
Evoked set created using respondent’s current repertoire, past usage and future
disposition
Occasio
n
Respondent’s answers using an occasion as a context in occasion led categories
like CSDs, Snacks etc. There can be other household purchase categories where
occasion is irrelevant
Repertoir
e
Respondents allowed to select multiple brands, SKUs and units, as they might in
real life
Channel
Primary channel of purchase identified and specified in the questioning
Event
Recurren
ce
Frequency of purchase
Here’s an example…
around 20 such choice tasks shown
Imagine you are doing your monthly grocery shopping from the supermarket and you
have to buy bathing soaps. On the shelf you see the following brands with the given
prices? Which brands are you likely to buy?
You can choose as many brands as you like. Or you can walk out of the shop without
buying any.
Do state the number of packs that you would be buying.
I will not buy any
anything
2
Johnson’s Baby
Soap
75 gms
Rs. 30
Dove
75 gms
Rs. 40
Santoor – pack of 4
100gms X 4
Rs. 50
Now at Rs. 40
Pears
100 gms + 25 gms
extra free
Rs. 25
Rexona
200 gms
Rs. 25
I would buy a
shower gel
Lux
100 gms
Rs. 10
1
Consumer choices are converted into utilities – two levels
of calculations
Main Effect
For every level within each attribute at a
respondent level
Example: Utility or preference for Brands like
Dove, Pears etc., for SKUs like 100 gms,, 75
gms and for various price levels
Cross Effect
Interaction between attributes
Example: Utility for Dove by itself and for Dove
at a particular price may be different
The utilities are transformed into
share of preference and
weighted to give the shares in
various scenarios
For all existing brands & SKUs,
the current levels of distribution
are built into the model – to
ensure current scenario shares
are in line with actual market
shares
Respondent level estimation of preference helps in calculation of
gains and losses from one brand to another
The output
Current
Scenario
New price
scenario
(Client’s
brands cut
prices by
10%)
Company
Brand X 20.0% 21.8%
Brand Y 17.0% 17.2%
Brand A 10.0% 9.9%
Brand B 5.0% 4.4%
Brand P 8.0% 7.6%
Brand Q 15.0% 14.8%
Brand R 10.0% 9.7%
Brand S 15.0% 14.6%
39%37%
Brand X
Gaining From
Brand B, Brand
S
Net Gain/Loss 1.8%
Econometric Model: Inputs and
Outputs
The input
Past volumes Population
Substitute
categories
(Real) Price Purchase frequency
No. of packs(Real) Income Basket size
The statistics
Options
Simple Regression
Volume = fn (Price)
Easy but can lead to situations like
increase price to increase
volumes!
Simple Time
Series
Volume = fn (Past volumes)
Builds in past volume trends but assumes that
history will definitely repeat itself
Price
Volume
The statistics
Adopted method
ARIMA
(Auto Regressive Integrated Moving
Average)
Volume = fn (Real Price, Past Volume, Real Income, etc….)
Moving average included – accounts for any possible prediction error in
previous time periods
Accounts for autocorrelation
Better prediction accuracy
The statistics
Adopted method
ARIMA
(Auto Regressive Integrated Moving
Average)
Deseasonalized data - predict organic change in volumes
Model by major sub-groups to account for different trends – break the
market
Price gap between sub-groups used – inter-movement built in
The output
Market volumes
125000 tonnes
Current scenario
126250 tonnes
New price
scenario
1%
Bringing the twain together
It’s quite simple actually…
Shares X Market Volumes = Company Volumes
37
Current
Scenario
125000 tonnes
New Scenario
Share
Market Volumes
46250 tonnesCompany volumes
39 126250 tonnes
Share
Market Volumes
49238 tonneCompany volumes
Share change: 5%
Volume change: 6%
Proof of the pudding
Results validated across markets
0
20
40
60
80
100
Predicted volume accuracy by brands in various markets
(74 data points in this graph)
Average :
88%
The trick: improve accuracy of the individual models
Identifying the relative impact of
touchpoints: A tailored statistical
technique for real-time data
Shane Baxendale, Cranfield School of Management
Heval Ceylan-Gilchrist, MESH
MRS ADVANCED ANALYTICS NETWORK
30th April 2015
Why are we here today?
 Real-time Experience Tracking Methodology
 Analysis - using linear mixed-effects
regression
24
Our thinking
 Consumers experience brands through multiple channels
(not just advertising!)
 Brand experiences influence a consumer’s attitude toward
brands
 The majority of existing literature focusses on the impact of
one or two types of experience
What impact are different encounters having on
consumer attitudes toward the brand?
25
*Baxendale S., Macdonald E.K., Wilson H.N., (2015), The impact of different
touchpoints on brand consideration, Journal of Retailing, 37(2), 203.
Real-time Experience
Tracking (RET)
ONLINE REAL-TIME ONLINE
Day 9Day 2 - 8Day 1
Text us whenever you
see, hear or experience
anything to do with the
following brands…
Text framework
27
BRAND: A)Brand A B)Brand B C)Brand C
D)Brand D E)Brand E F) Other
OCCASION: A)TV B)Poster/Billboard
C)Radio D)In store E)Cinema F)Newspaper
G)Magazine H)Conversation
I)Online/Mobile J)Mailing/leaflet K)Me
Purchasing L)Me using M)Someone else
using N)Sponsorship O)Other
FEELING: 5)Very positive 4)Fairly positive
3)Neutral 2)Fairly negative 1)Very Negative
CHOICE: 5)Much more likely to choose
4)Slightly more likely to choose 3)No
difference 2)Slightly less likely to choose 1)
Much less likely to choose
Imagine you experienced Brand A Online…
…you would text:
28
a 5i 5
CHOICE:
5) Much more likely to
choose
4) Slightly more likely to
choose
3) No change
2) Slightly less likely to
choose
1) Much less likely to
choose
ENGAGEMENT:
5) Very positive
4) Fairly positive
3) Neutral
2) Fairly negative
1) Very negative
BRAND:
a) Brand A
b) Brand B
c) Brand C
d) Brand D
e) Brand E
f) Other
OCCASION:
a) TV
b) Poster/Billboard
c) Radio
d) In store
e) Cinema
f) Newspaper
g) Magazine
h) Conversation
i) Online/Mobile
j) Mailing/Leaflet
k) Me purchasing
l) Me using
m) Someone else using
n) Sponsorship
o) Other
Which brand was it? Where did you
experience it?
How likely did it make you
to choose the brand next
time?
How did it make you feel?
Now tell us more in an online diary…
29
This is an individual’s experience log By clicking on each entry, the experience can
be expanded upon in detail
Wednesday 13th February 2012,11:54
Wednesday13th February2012,11:54
Wednesday13th February2012,10:22
Tuesday512h February2012,18:46
Tuesday12th February2012,13:05
Tuesday12th February2012,08:38
Brand A,Online, Very Positive,
MuchMoreLikely toChoose
Brand C,Conversation, Very Negative,
MuchLess Likely toChoose
Brand E, TV, Fairly Positive,
Slightly More Likely to Choose
Other, Instore, Fairly Positive,
Slightly morelikely toChoose
Brand B, Mailing/Leaflet, Slightly Negative,
Nochange
Brand A
Brand’s website
Very positive
Much more likely to choose
I was looking on the brand website to find out
more information about the product range. Looks
like there are some good options.
Look for product info
13/02/2012, 11:54
Please tell us exactly what yousaw? :
What was the purpose ofyour online activity? :
Brand’s website
Ad from brand
In the news
Social networking site
Price comparison site
Other
For each level of data captured
in real-time we can tailor extra
questions to get more granular
information in near-time
Data
For one individual
Brand A Brand B … Brand N
Consideration Wk0
Consideration Wk1
Consideration Wk0
Consideration Wk1
Consideration Wk0
Consideration Wk1
Freq. & Pos.
Brand Ad
Retailer Ad
In Store
WOM
…
Freq. & Pos.
Brand Ad
Retailer Ad
In Store
WOM
…
Freq. & Pos.
Brand Ad
Retailer Ad
In Store
WOM
…
30
+ve -ve -ve
Model
© Cranfield University 31
Change in Consideration = ???
Data Rationale Considerations Implications
Demographics / Participant
information
Certain consumer groups
may be more / less likely to
change their opinion
towards brands over time
Multiple responses per
participant means that we
can learn more about
individual tendencies
Need to account for the
homogeneity in repeated
responses, therefore
random effects modelling
Frequency of experience More experiences can be a
positive impact (via
reinforcement of
messaging) or negative
(over-exposure)
There could be many
potential ways of including
this in the model;
Constant effect
Diminishing returns
Check the validity of the
model by testing multiple
approaches
Positivity of experience A consumers’ perception of
an experience can
determine the impact it
has on them
How do we account for
positivity over multiple
encounters?
Average?
Check the validity of the
model by testing multiple
approaches
Parameter operation
Frequency
© Cranfield University 32
0 1 2 3 4 5
0 1 2 3 4 5
0 1 2 3 4 5
Exposure
Impact = x if Freq.>0
Increasing Impact
Impact = x*Freq.
Diminishing Returns
Impact = x*ln(Freq.+1)
Impact = x*Freq. - y*Freq.^2
Parameter operation
Positivity
1. Average positivity across experiences
2. Average positivity and variance of positivity
3. Positivity of last experience
4. Freq. of positive and Freq. of negative
experiences
© Cranfield University 33
Results
Focal Frequency Positivity
In-store
communications
=1 1
Brand advertising =1 =2
Retailer advertising =1 =2
Peer observation =1 =4
Traditional earned =5 =4
WOM =5 6
© Cranfield University 34
Competitor Frequency Positivity
In-store
communications
1 =1
WOM =2 =1
Peer observation =2 =1
Retailer advertising =4 =1
Brand advertising =4 =1
Traditional earned 6 =1
Thank You!
The contents of this document are the sole and confidential property of Lieberman Research
Worldwide, and may not be reproduced or distributed without the express written permission
of Lieberman Research Worldwide.
Prepared for CLIENT
TITLE
LRW Europe
BAYESIAN ANALYSIS
FOR MARKETING IMPACT
April
2015
LRW Europe
1, Heathcock
Court, 415, Strand
London
WC2R 0NT
Prepared by:
Adele Gritten &
Graham Williams
for MRS Advanced
Analytics
Conference
‹#›
© 2015 Lieberman Research Worldwide.
All rights reserved. CONFIDENTIAL.
 Why should our
industry care about
BayesNets?
 What is BayesNets?
 What are the
Advantages of
BayesNets?
 Illustrative outputs
 Live UK Case Study
 Summary
TODAY’S AGENDA
‹#›
© 2015 Lieberman Research Worldwide.
All rights reserved. CONFIDENTIAL.
Why should
our Industry
Care about
BayesNets?
‹#›
© 2015 Lieberman Research Worldwide.
All rights reserved. CONFIDENTIAL.
Why Should Our Industry Care About BayesNets?
 BayesNets is a unique and more comprehensive driver analysis to
assist Marketers
 Overcomes the shortcomings of traditional [drivers analysis] methods
 Allows the integration of profiling, behavioural and other metrics with
attitudinal/preference ratings to create a more holistic view of what
drives the dependent variable
“Bayesnets has played a major role in several
recent wins. It’s especially helpful with brand
positioning research where the complex
relationships between brand attributes
demands a more nuanced and flexible
approach to analysis”. LRW Account Director
‹#›
© 2015 Lieberman Research Worldwide.
All rights reserved. CONFIDENTIAL.
What is
BayesNets?
‹#›
© 2015 Lieberman Research Worldwide.
All rights reserved. CONFIDENTIAL.
What is BayesNets?
Think of BayesNets as “drivers analysis on steroids”
Let’s review the logic and goals of
“key drivers analysis”:
 There is a market attitude or behaviour – the
“target outcome” – which:
 A client needs to favourably influence in the
marketplace, but which…
 They cannot influence directly
 So we need to find the best “levers to pull”
through which to indirectly influence that attitude
or behaviour; e.g. customer attitudes or
perceptions:
 Which we can influence through
product/service design or marketing, and…
 Which have strong “derived importance” in
driving the “target outcome”
 Ultimately, the purpose of “key drivers analysis” is
to empirically identify the best “levers to pull” for
maximum in market impact.
BayesNets:
 Can be thought of
generally as a more
powerful key drivers
analysis methodology
 Offers a number of
significant advantages
compared to hitherto
commonly used key
drivers analysis
approaches
Where Did BayesNets Come From?
| Thomas Bayes
“Bayesian” refers to Reverend Thomas Bayes’
Theorem from the 18th century that paved the way
for data to be used in prediction. Bayes’ Theorem
basically allows us to look at multi-directional
probabilities.
 18th Century English statistician,
philosopher, and minister
 Formulated Bayes’ Theorem: a
mathematical expression of probabilities
from observed data
 Hotly debated and contested by
Frequentists until recent years
42
© 2015 Lieberman Research Worldwide.
All rights reserved. CONFIDENTIAL.
‹#›
© 2015 Lieberman Research Worldwide.
All rights reserved. CONFIDENTIAL.
What are BayesNets? | A Model of Relationships
Bayesian networks (BayesNets) are a type of “path analysis” model that simultaneously describes the
relationships between variables in a network system based on joint probabilities between the variables.
BayesNets improves upon models that use advanced correlation and regression techniques (such as most
regression analyses and Structural Equation Modeling). Basically, it’s a better way of understanding
interactions between independent variables as they drive dependent variables.
WET
PAVEMENT
SPRINKLER
RAIN
SLIPPERY
Probabilistic
Relationship
Probabilistic
Relationship
Variables
or Factors
Variables
or Factors
High-Dimensional Probability Hypercube
‹#›
© 2015 Lieberman Research Worldwide.
All rights reserved. CONFIDENTIAL.
What are the
Advantages
of
BayesNets?
‹#›
© 2015 Lieberman Research Worldwide.
All rights reserved. CONFIDENTIAL.
What are Advantages
of BayesNets?
BayesNets’ advantages over
more commonly used
techniques are “technical” but
nonetheless significant.
BayesNets’ advantages include:
 Does fully interactive, “multivariate” modeling
 Not “confused” by multicollinearity
 No implicit assumption of “linear” relationships
 Siloed Regression Models don’t capture indirect or interaction effects
BayesNets helps us find the best model:
 We don’t have to “hypothesise” the structure of the multivariate network
 Rather, BayesNets’ “machine learning” algorithms seek out the best network structures
quickly & cost effectively
 From there we bring in the “art” that mixes with the “science” to yield a highly actionable
understanding of what drives the target outcome – the dependent variable – in the
marketplace.
Traditional Approach
Approach with
BayesNets
Variables or
Factors
Independent
Variables or
Factors
‹#›
© 2015 Lieberman Research Worldwide.
All rights reserved. CONFIDENTIAL.
Siloed Regression Models Don’t Capture Indirect or Interaction Effects
Key Satisfaction/Loyalty Metric
ProductsAtmosphere
Store
Experience
Minutes in
line
Seconds
Ordering
Speed of
Ordering
Ordering
Process
Volume
Customer
Service
Greeted by
Employee
Friendly
Employees
EXPERIENCE
DOMAINS
PERCEPTIONS
OF BEHAVIORS
QUANTIFIABLE
MICRO-
BEHAVIORS
SUPPORTING
IMPRESSIONS
Metrics can impact other domains, not just those up the ladder in
our hierarchy
‹#›
© 2015 Lieberman Research Worldwide.
All rights reserved. CONFIDENTIAL.
The Former Best Approach: SEM
Structural Equation Modeling
 Tests Complex Structures
 Interactive & Indirect Effects
 Explanatory & Prescriptive
 Multicollinearity may still be a
problem
 Can only test the “fit” of
specifically hypothesized
networks
Siloed Regression Tree
 Forces Simple Structure
 Direct Effects Only
 Diagnostic & Descriptive
‹#›
© 2015 Lieberman Research Worldwide.
All rights reserved. CONFIDENTIAL.
Illustrative
Outputs
‹#›
© 2015 Lieberman Research Worldwide.
All rights reserved. CONFIDENTIAL.
Variables or Factors
Colours identify
“nodes” that belong
to different factors
Probabilistic
Relationship
“Arcs” connect the
various “nodes” in
the network
What does it look like? | It Starts With Networks
‹#›
© 2015 Lieberman Research Worldwide.
All rights reserved. CONFIDENTIAL.
BayesNets Satisfaction Key Drivers Analysis:
A Case Study
 Customer satisfaction surveys with >650,000 retail customers
 Surveys conducted throughout 2013
 Dependent variable: Overall quality of in-store experience rating
 Independent variables:
 Is a place for someone like me
 Clothing was neatly displayed and well organized
 The wait time in the checkout line was acceptable
 Service you received in the fitting room met your needs
 The cashier worked quickly and efficiently to check out all customers in line
 Your experience in the store was more fun and engaging than other stores you typically
shop
 The signs clearly indicated what was on sale
 Employees were easily accessible
 Employees were willing to find style, color, size
 Employees acknowledged and made you feel welcome
 Employees seemed genuinely glad you were there
 Overall clothing quality
‹#›
© 2015 Lieberman Research Worldwide.
All rights reserved. CONFIDENTIAL.
BayesNets Customer Satisfaction Drivers Analysis - Retail Example
Wait time at checkout
Feel welcomed
For someone like me
Overall clothing quality
Neat displays /organised
Sale signs clear
More fun and engaging
than other stores
Accessible employees
Employees glad
you were there
Cashier
worked
quickly and
efficiently
Employees willing to
find style, colour, size
Service received
in fitting room
met needs
Overall experience
‹#›
© 2015 Lieberman Research Worldwide.
All rights reserved. CONFIDENTIAL.
BayesNets “total effects” analysis looks & feels like standard drivers analysis
All Brands Retailer 1 Retailer 2 Retailer 3 Retailer 4 Retailer 5
Total Effects on Target Promoter_Rev Indexed Indexed Indexed Indexed Indexed Indexed
Standardized Standardized Standardized Standardized Standardized Standardized
Total Effects Total Effects Total Effects Total Effects Total Effects Total Effects
F1: MERCHANDISE FOR ME 144.3 134.3 159.5 137.4 150.9 133.1
F0: EMOTIONALLY ENGAGED 131.9 114.6 127.8 110.5 115.4 111.7
F17: GREAT VALUE 125.8 121.6 126.8 112.7 129.7 118.6
F4: EXCITING AND STYLISH MERCHANDISE 125.0 111.6 147.3 115.9 102.7 116.7
F5: QUALITY BRANDS 120.1 111.1 132.7 152.7 111.7 115.6
F3: GREAT FIT & SIZES 112.8 114.9 112.0 132.3 119.1 115.3
F7: GREAT PRICES AND SAVINGS 110.0 120.6 141.9 117.6 95.1 114.7
F14: MERCHANDISE FOR MY HOME AND FAMILY 109.8 107.1 146.2 75.6 128.3 105.7
F11: GREAT SALES 105.1 111.6 102.2 95.8 117.5 106.3
F2: ENJOYABLE SHOPPING 103.9 94.0 56.8 101.3 105.5 100.5
I can always count on STORE to have what I want on sale 102.9 103.2 104.2 69.5 109.1 104.4
F12: BETTER DEALS 101.0 108.3 126.9 99.9 97.2 102.0
F13: EASY RETURN POLICY 93.4 93.4 104.9 107.1 91.4 101.5
F9: PRICES I TRUST 80.6 78.5 98.0 61.9 90.8 74.7
F6: LOYALTY PROGRAM 79.0 78.7 56.3 59.3 52.2 75.2
F8: COUPONS 72.2 93.5 39.0 85.7 75.1 89.1
F16: INSPIRING DISPLAYS 70.0 78.7 26.6 120.5 74.3 95.7
F18: SUPPORTS MY COMMUNITY 61.9 61.5 68.8 71.7 72.7 73.0
F10: EASY PROMOTION 50.5 62.7 22.1 72.5 61.4 46.3
Illustrative
‹#›
© 2015 Lieberman Research Worldwide.
All rights reserved. CONFIDENTIAL.
The output here gives specific advice on which factors to affect first, and when it is optimal to focus on the
next factor.
Bayesian Analysis can also provide clear recommendations on
where businesses should focus
Initial Mean
Rating
Mean Rating
After
Improving
Preceding
Factors
Target
Mean
Overall Opinion
Mean
Initial Value 4.45
HEALTHY 9.07 9.27 4.79
SELECTION/VARIETY 8.64 8.84 9.05 4.88
QUALITY 8.25 8.40 8.59 4.91
EASY/SIMPLE 8.05 8.56 8.86 4.92
First, the goal is to
move the mean on the
Healthy factor from
9.07 to 9.27
This would increase
the Overall Opinion
0.34 points, from 4.45
to 4.79
Moving the Healthy
mean from 9.07 to 9.27
also affects
Selection/Variety,
moving it from 8.64 to
8.84.
Illustrative
Moving the Selection
mean from 8.84 to 9.05
similarly impacts both
the Overall Opinion
mean (up to 4.88) and
the Quality mean
(moving it to 8.40) and
so on for each
successive factor
‹#›
© 2015 Lieberman Research Worldwide.
All rights reserved. CONFIDENTIAL.
UK Media Owner Client: Live Case Study
Current data sources include:
• Brand Tracking with image metrics
• Industry audience measurement
• Content audit v competitors
• Content appreciation
• Social Media tracking
They tend to do a rough & ready comparison and they are doing some KDA in the tracking
data, but so far no joined up stuff
With so much data they worry about how much stakeholders trust or care about the data
‘I have an overload of data and metrics and I
want to see what combination of factors drive
audience growth (or decline).... At the moment I
can’t see the wood for the trees – I’m hoping
Bayesnet will help’
LRW are working with the client to
initially conduct a Bayesnet
analysis on the monthly tracker
(which goes back over 2 years) to
identify relationship between
behaviour and the metrics and
which ones are the ones that they
really need to look at
Ideal solution would be to
stream line the cumbersome
tracker – strip out metrics that
don’t add value and then look to
widen the Bayesnet analysis to
other data sources and conduct
a wider analysis
‹#›
© 2015 Lieberman Research Worldwide.
All rights reserved. CONFIDENTIAL.
Summary
‹#›
© 2015 Lieberman Research Worldwide.
All rights reserved. CONFIDENTIAL.
In Summary | Why BayesNet Modeling?
 BayesNets modeling is often more effective
than more traditional advanced modeling
of derived importance
analysis
 BayesNet measures both direct impact on the dependent
variables and indirect impacts through other independent
variables in the model
 BayesNets overcomes multicollinearity and makes no
assumptions of either normal distributions of data or linear
correlations between variables.
 BayesNets mathematics and software allow for quicker creation
of the model, optimizations and “what if” scenarios.
 BayesNets offers more effective optimization modeling with
target means to guide activation and appropriate levels of effort
and investment.
‹#›
© 2015 Lieberman Research Worldwide.
All rights reserved. CONFIDENTIAL.
If you’d like more info: We can set up a time for you to talk to one of our
genuine experts!
Mick
McWilliams
PhD, Sr. VP,
Marketing
Science
Marketing scientist specializing in
segmentation, brand engagement, database
scoring, SEM, KDA and BayesNets
25+ years of MR experience with specialties
in neuroscience studies & evolutionary
psychology
PhD, Sociology, Virginia Polytechnic Institute
& State University
Thank you!
‹#›
© 2015 Lieberman Research Worldwide.
All rights reserved. CONFIDENTIAL.
Graham Williams
Research Director, Europe
gwilliams@LRWonline.com
Lieberman Research
Worldwide
1, Heathcock Court, 415
Strand
London
WC2R ONT
www.lrwonline.com
Direct Line: 0203 551 7075
Contact Information
Copyright © Nepa All Rights Reserved
Space Optimisation
MRS Advanced Analytics
30 April 2015
Copyright © Nepa All Rights Reserved
60
Space Optimisation is the process of maximising profit
by allocating the appropriate amount of store shelf
space to each product category
Copyright © Nepa All Rights Reserved
Typical client – a retail chain with a wealth of sales and loyalty club data
61
...of different locations
and demographic
profiles
Many hundred stores of
different sizes
Nearly 100
product
categories
Copyright © Nepa All Rights Reserved
Many aspects determine the profit that a product category will yield.
First, the most important ones are identified
Significant factors, in selection :
• Affluence in neighbourhood
• Gender profile
• Age profile
• Proximity to low-cost competition
• ...
• ...
• ...
62
Demography
Location / competition
Sales details
Customer
Satisfaction
Copyright © Nepa All Rights Reserved
Linear regression is used to isolate the relationship between space and
profit, per product category
63
Space allocated
...
...
Affluence in neighbourhood
Gender profile
Age profile
Proximity to low-cost
competition
...
...
β1
+ β2
+ β3
+ β4
+ β5
+ βs
Copyright © Nepa All Rights Reserved
”Space elasticity” – not the same for all product types, illustrated by an
intuitive example from a pharmacy
64
1 shelf
Margin: £500 per day
 £500 /shelf
4 shelves
Margin: £1000 per day
 £250 /shelf
+?
Margin
Space
Margin
Space
Due to its higher space elasticity,
it is likely more profitable to add
another beauty shelf than one for
pain killers. This despite the fact
that painkillers presntly give
more profit per shelf unit.
Copyright © Nepa All Rights Reserved
Store-specific linear regression gives accurate space elasticity curves in
steps, for each category
65
Shelf space allocated
...
...
Affluence in neighbourhood
Gender profile
Age profile
Proximity to low-cost competition
...
...
βs
Copyright © Nepa All Rights Reserved
Store-specific linear regression gives accurate space elasticity curves in steps,
for each category
66
0
20
40
60
80
100
120
0 20 40 60
Margin(£)
Space (Shelf sections)
Shelf space allocated
...
...
Affluence in neighbourhood
Gender profile
Age profile
Proximity to low-cost competition
...
...
βs
Copyright © Nepa All Rights Reserved
Store-specific linear regression gives accurate space elasticity curves in steps,
for each category
67
0
20
40
60
80
100
120
0 50 100 150 200
Margin(£) Space (Shelf sections)
Shelf space allocated
...
...
Affluence in neighbourhood
Gender profile
Age profile
Proximity to low-cost competition
...
...
βs
Copyright © Nepa All Rights Reserved
Store-specific linear regression gives accurate space elasticity curves in steps,
for each category
68
0
10
20
30
40
50
60
70
80
90
0 50 100 150 200
Margin(£) Space (shelf sections)
Shelf space allocated
...
...
Affluence in neighbourhood
Gender profile
Age profile
Proximity to low-cost competition
...
...
βs
Copyright © Nepa All Rights Reserved
Store-specific linear regression gives accurate space elasticity curves in steps,
for each category
69
0
10
20
30
40
50
60
70
80
90
0 50 100 150 200
Margin(£) Space (shelf sections)
Shelf space allocated
...
...
Affluence in neighbourhood
Gender profile
Age profile
Proximity to low-cost competition
...
...
βs
Copyright © Nepa All Rights Reserved
Store-specific linear regression gives accurate space elasticity curves in steps,
for each category
70
0
10
20
30
40
50
60
70
80
0 10 20 30 40 50
Margin(£) Space (shelf sections)
Shelf space allocated
...
...
Affluence in neighbourhood
Gender profile
Age profile
Proximity to low-cost competition
...
...
Copyright © Nepa All Rights Reserved
We will never start
adding delicassy
cheeses, since the start
of the curve is so flat.
Combining curves for the optimal space allocation – stepwise incremental
assignment doesn’t always find the best solution available
71
0
10
20
30
40
50
60
70
80
90
100
0 1 2 3 4 5 6 7 8 9 10
Margin(£)
ABBBC
C
D
D
E
F
GG/HIJK
Store plan, 15 shelves
8
7
Stepwise adding products to shelves
using the highest incremental value at
each step will result in assigning 8 to
vegetables and 7 to sauces
Copyright © Nepa All Rights Reserved
The optimal distribution includes many shelves of delicassy cheeses, giving
a large profit at substantial space assignment
72
0
10
20
30
40
50
60
70
80
90
100
0 1 2 3 4 5 6 7 8 9 10
Margin(£)
ABBBC
C
D
D
E
F
GG/HIJK
Store plan, 15 shelves
96
Copyright © Nepa All Rights Reserved
We search through all possible combinations to find the best one – an
enormous optimisation problem which we use logic to reduce
100 shelf units to allocate
73
30 categories
...
...
...
...
...
...
...
...
6 x 1028 combinations!
...
... Even this rather moderate number of shelf units and categories
presents an enormous number of potential combinations.
We need to use logic to reduce the computational complexity,
and find the best solution available.
Copyright © Nepa All Rights Reserved
An online tool is used for
space allocation, bespoke
to each individual store
74
Copyright © Nepa All Rights Reserved
Thank you!
kalle.backlund@nepa.com
0785-19 49 379
75

More Related Content

What's hot

Win-Win – Give telecoms customers the freedom they want and they’ll reward yo...
Win-Win – Give telecoms customers the freedom they want and they’ll reward yo...Win-Win – Give telecoms customers the freedom they want and they’ll reward yo...
Win-Win – Give telecoms customers the freedom they want and they’ll reward yo...SKIM
 
Pricing research
Pricing researchPricing research
Pricing researchDeepak Gaur
 
Marketing Research - Perceptual Map
Marketing Research - Perceptual MapMarketing Research - Perceptual Map
Marketing Research - Perceptual MapMinha Hwang
 
[Project] Retail Management Report Brands Versus Private Labels- Fighting to Win
[Project] Retail Management Report Brands Versus Private Labels- Fighting to Win[Project] Retail Management Report Brands Versus Private Labels- Fighting to Win
[Project] Retail Management Report Brands Versus Private Labels- Fighting to WinBiswadeep Ghosh Hazra
 
Tweeter Electronics: Marketing Case Analysis
Tweeter Electronics: Marketing Case AnalysisTweeter Electronics: Marketing Case Analysis
Tweeter Electronics: Marketing Case AnalysisDipak Senapati
 
From Online to Mobile - Impact of Consumers' Online Purchase Behaviors on Mob...
From Online to Mobile - Impact of Consumers' Online Purchase Behaviors on Mob...From Online to Mobile - Impact of Consumers' Online Purchase Behaviors on Mob...
From Online to Mobile - Impact of Consumers' Online Purchase Behaviors on Mob...Minha Hwang
 
BCG matrix with example
BCG matrix with exampleBCG matrix with example
BCG matrix with exampleMayur Narole
 
Repositioning of Wendy's
Repositioning of Wendy'sRepositioning of Wendy's
Repositioning of Wendy'sTeri Grossheim
 
[foresight research] Introduction to Brand Health Tracking
[foresight research] Introduction to Brand Health Tracking[foresight research] Introduction to Brand Health Tracking
[foresight research] Introduction to Brand Health TrackingDuy, Vo Hoang
 
Best Buy International Strategic Update
Best Buy International Strategic UpdateBest Buy International Strategic Update
Best Buy International Strategic Updatefinance7
 
Rm project best buy v2
Rm project  best buy v2Rm project  best buy v2
Rm project best buy v2Rajendra Inani
 
Social Media Brand Positioning Workflow- David Gerson
Social Media Brand Positioning Workflow- David GersonSocial Media Brand Positioning Workflow- David Gerson
Social Media Brand Positioning Workflow- David GersonPyData
 
Best buy-analysis
Best buy-analysisBest buy-analysis
Best buy-analysisTaposh Roy
 
Pricing Strategy: How To Win With Subscription Pricing Models
Pricing Strategy: How To Win With Subscription Pricing ModelsPricing Strategy: How To Win With Subscription Pricing Models
Pricing Strategy: How To Win With Subscription Pricing ModelsZuora, Inc.
 
Virgin mobile USA pricing first time case analysis
Virgin mobile USA pricing first time case analysisVirgin mobile USA pricing first time case analysis
Virgin mobile USA pricing first time case analysisSiddharth Dhamija
 

What's hot (20)

Win-Win – Give telecoms customers the freedom they want and they’ll reward yo...
Win-Win – Give telecoms customers the freedom they want and they’ll reward yo...Win-Win – Give telecoms customers the freedom they want and they’ll reward yo...
Win-Win – Give telecoms customers the freedom they want and they’ll reward yo...
 
Pricing research
Pricing researchPricing research
Pricing research
 
Marketing Research - Perceptual Map
Marketing Research - Perceptual MapMarketing Research - Perceptual Map
Marketing Research - Perceptual Map
 
[Project] Retail Management Report Brands Versus Private Labels- Fighting to Win
[Project] Retail Management Report Brands Versus Private Labels- Fighting to Win[Project] Retail Management Report Brands Versus Private Labels- Fighting to Win
[Project] Retail Management Report Brands Versus Private Labels- Fighting to Win
 
Tweeter Electronics: Marketing Case Analysis
Tweeter Electronics: Marketing Case AnalysisTweeter Electronics: Marketing Case Analysis
Tweeter Electronics: Marketing Case Analysis
 
From Online to Mobile - Impact of Consumers' Online Purchase Behaviors on Mob...
From Online to Mobile - Impact of Consumers' Online Purchase Behaviors on Mob...From Online to Mobile - Impact of Consumers' Online Purchase Behaviors on Mob...
From Online to Mobile - Impact of Consumers' Online Purchase Behaviors on Mob...
 
BCG matrix with example
BCG matrix with exampleBCG matrix with example
BCG matrix with example
 
Repositioning of Wendy's
Repositioning of Wendy'sRepositioning of Wendy's
Repositioning of Wendy's
 
[foresight research] Introduction to Brand Health Tracking
[foresight research] Introduction to Brand Health Tracking[foresight research] Introduction to Brand Health Tracking
[foresight research] Introduction to Brand Health Tracking
 
Best Buy International Strategic Update
Best Buy International Strategic UpdateBest Buy International Strategic Update
Best Buy International Strategic Update
 
Pricing research
Pricing researchPricing research
Pricing research
 
Praag 280809v 0.9
Praag 280809v 0.9Praag 280809v 0.9
Praag 280809v 0.9
 
Rm project best buy v2
Rm project  best buy v2Rm project  best buy v2
Rm project best buy v2
 
Azhar
Azhar Azhar
Azhar
 
Social Media Brand Positioning Workflow- David Gerson
Social Media Brand Positioning Workflow- David GersonSocial Media Brand Positioning Workflow- David Gerson
Social Media Brand Positioning Workflow- David Gerson
 
Best buy-analysis
Best buy-analysisBest buy-analysis
Best buy-analysis
 
Pricing Strategy: How To Win With Subscription Pricing Models
Pricing Strategy: How To Win With Subscription Pricing ModelsPricing Strategy: How To Win With Subscription Pricing Models
Pricing Strategy: How To Win With Subscription Pricing Models
 
Chap4pp
Chap4ppChap4pp
Chap4pp
 
Virgin mobile USA pricing first time case analysis
Virgin mobile USA pricing first time case analysisVirgin mobile USA pricing first time case analysis
Virgin mobile USA pricing first time case analysis
 
One Experience
One ExperienceOne Experience
One Experience
 

Similar to MRS Advanced Analytics Innovation Symposium Presentation

Importing, Exporting and SourcingWhat’s this chapter about
Importing, Exporting and SourcingWhat’s this chapter aboutImporting, Exporting and SourcingWhat’s this chapter about
Importing, Exporting and SourcingWhat’s this chapter aboutMalikPinckney86
 
FMCG for Management Consultants and Business Analysts
FMCG for Management Consultants and Business AnalystsFMCG for Management Consultants and Business Analysts
FMCG for Management Consultants and Business AnalystsAsen Gyczew
 
Group 17 - Immaculate - Disruptive Innovation in the Beauty Industry
Group 17 - Immaculate - Disruptive Innovation in the Beauty IndustryGroup 17 - Immaculate - Disruptive Innovation in the Beauty Industry
Group 17 - Immaculate - Disruptive Innovation in the Beauty Industrygroup17teen
 
2.5 advertising for all
2.5 advertising for all2.5 advertising for all
2.5 advertising for allVasili Andrews
 
Immaculate | Disruptive Innovation in the Beauty Industry
Immaculate | Disruptive Innovation in the Beauty IndustryImmaculate | Disruptive Innovation in the Beauty Industry
Immaculate | Disruptive Innovation in the Beauty IndustryJakki Magowan
 
Group 17 - IMMACULATE - Disruptive Innovation in the Beauty Industry
Group 17 - IMMACULATE - Disruptive Innovation in the Beauty IndustryGroup 17 - IMMACULATE - Disruptive Innovation in the Beauty Industry
Group 17 - IMMACULATE - Disruptive Innovation in the Beauty Industry7teen2
 
Group 17 - Immaculate - Disruptive Innovation in the Beauty Industry
Group 17 - Immaculate - Disruptive Innovation in the Beauty IndustryGroup 17 - Immaculate - Disruptive Innovation in the Beauty Industry
Group 17 - Immaculate - Disruptive Innovation in the Beauty Industrygroup17teen
 
Emerce eTravel - 5 startups lessons to build better products faster
Emerce eTravel - 5 startups lessons to build better products fasterEmerce eTravel - 5 startups lessons to build better products faster
Emerce eTravel - 5 startups lessons to build better products fasterMeasureWorks
 
Introduction to marketing management
Introduction to marketing managementIntroduction to marketing management
Introduction to marketing managementManish Parihar
 
The World of Marketing
The World of MarketingThe World of Marketing
The World of Marketingjohn3092
 
Turning Good Ideas into Great Products
Turning Good Ideas into Great ProductsTurning Good Ideas into Great Products
Turning Good Ideas into Great ProductsChristina Wodtke
 
Brand equity
Brand equityBrand equity
Brand equityamitgurus
 
Week 6, Using Marketing Channels and Price to Create Value for Cu.docx
Week 6, Using Marketing Channels and Price to Create Value for Cu.docxWeek 6, Using Marketing Channels and Price to Create Value for Cu.docx
Week 6, Using Marketing Channels and Price to Create Value for Cu.docxcockekeshia
 
Media, advertising & consumer
Media, advertising & consumerMedia, advertising & consumer
Media, advertising & consumerRbk Asr
 

Similar to MRS Advanced Analytics Innovation Symposium Presentation (20)

Importing, Exporting and SourcingWhat’s this chapter about
Importing, Exporting and SourcingWhat’s this chapter aboutImporting, Exporting and SourcingWhat’s this chapter about
Importing, Exporting and SourcingWhat’s this chapter about
 
FMCG for Management Consultants and Business Analysts
FMCG for Management Consultants and Business AnalystsFMCG for Management Consultants and Business Analysts
FMCG for Management Consultants and Business Analysts
 
Group 17 - Immaculate - Disruptive Innovation in the Beauty Industry
Group 17 - Immaculate - Disruptive Innovation in the Beauty IndustryGroup 17 - Immaculate - Disruptive Innovation in the Beauty Industry
Group 17 - Immaculate - Disruptive Innovation in the Beauty Industry
 
2.5 advertising for all
2.5 advertising for all2.5 advertising for all
2.5 advertising for all
 
Immaculate | Disruptive Innovation in the Beauty Industry
Immaculate | Disruptive Innovation in the Beauty IndustryImmaculate | Disruptive Innovation in the Beauty Industry
Immaculate | Disruptive Innovation in the Beauty Industry
 
Group 17 - IMMACULATE - Disruptive Innovation in the Beauty Industry
Group 17 - IMMACULATE - Disruptive Innovation in the Beauty IndustryGroup 17 - IMMACULATE - Disruptive Innovation in the Beauty Industry
Group 17 - IMMACULATE - Disruptive Innovation in the Beauty Industry
 
Group 17 - Immaculate - Disruptive Innovation in the Beauty Industry
Group 17 - Immaculate - Disruptive Innovation in the Beauty IndustryGroup 17 - Immaculate - Disruptive Innovation in the Beauty Industry
Group 17 - Immaculate - Disruptive Innovation in the Beauty Industry
 
Emerce eTravel - 5 startups lessons to build better products faster
Emerce eTravel - 5 startups lessons to build better products fasterEmerce eTravel - 5 startups lessons to build better products faster
Emerce eTravel - 5 startups lessons to build better products faster
 
Introduction to marketing management
Introduction to marketing managementIntroduction to marketing management
Introduction to marketing management
 
The World of Marketing
The World of MarketingThe World of Marketing
The World of Marketing
 
Turning Good Ideas into Great Products
Turning Good Ideas into Great ProductsTurning Good Ideas into Great Products
Turning Good Ideas into Great Products
 
Brand equity
Brand equityBrand equity
Brand equity
 
Marketing
MarketingMarketing
Marketing
 
Brand equity
Brand equityBrand equity
Brand equity
 
2015 stima holaba
2015 stima holaba2015 stima holaba
2015 stima holaba
 
2015 stima holaba
2015 stima holaba2015 stima holaba
2015 stima holaba
 
2015 Stima holaba @ Palm
2015 Stima holaba @ Palm 2015 Stima holaba @ Palm
2015 Stima holaba @ Palm
 
L4
L4L4
L4
 
Week 6, Using Marketing Channels and Price to Create Value for Cu.docx
Week 6, Using Marketing Channels and Price to Create Value for Cu.docxWeek 6, Using Marketing Channels and Price to Create Value for Cu.docx
Week 6, Using Marketing Channels and Price to Create Value for Cu.docx
 
Media, advertising & consumer
Media, advertising & consumerMedia, advertising & consumer
Media, advertising & consumer
 

More from MRS

Covid 19 research - wave 2
Covid 19 research - wave 2 Covid 19 research - wave 2
Covid 19 research - wave 2 MRS
 
Respondi whitepaper the 'appiness project
Respondi whitepaper the 'appiness projectRespondi whitepaper the 'appiness project
Respondi whitepaper the 'appiness projectMRS
 
Supporting good Mental Health at Work
Supporting good Mental Health at Work Supporting good Mental Health at Work
Supporting good Mental Health at Work MRS
 
MRS Code of Conduct 2019 - Changes to Fair Data
MRS Code of Conduct 2019 - Changes to Fair DataMRS Code of Conduct 2019 - Changes to Fair Data
MRS Code of Conduct 2019 - Changes to Fair DataMRS
 
Big Semiotics - May 2019
Big Semiotics - May 2019Big Semiotics - May 2019
Big Semiotics - May 2019MRS
 
Digital Darwinism: How online communities can survive and thrive three waves ...
Digital Darwinism: How online communities can survive and thrive three waves ...Digital Darwinism: How online communities can survive and thrive three waves ...
Digital Darwinism: How online communities can survive and thrive three waves ...MRS
 
How to write an Oppies Award Entry
How to write an Oppies Award EntryHow to write an Oppies Award Entry
How to write an Oppies Award EntryMRS
 
MRS Roadshow 2019
MRS Roadshow 2019MRS Roadshow 2019
MRS Roadshow 2019MRS
 
BBC Media Action - 2019
BBC Media Action - 2019BBC Media Action - 2019
BBC Media Action - 2019MRS
 
Operations network meeting 22 January 2019
Operations network meeting 22 January 2019Operations network meeting 22 January 2019
Operations network meeting 22 January 2019MRS
 
Using VR for immersion and audience engagement
Using VR for immersion and audience engagementUsing VR for immersion and audience engagement
Using VR for immersion and audience engagementMRS
 
Humans v tech
Humans v tech Humans v tech
Humans v tech MRS
 
Planning for new communities
Planning for new communitiesPlanning for new communities
Planning for new communitiesMRS
 
Women in Ads
Women in AdsWomen in Ads
Women in AdsMRS
 
Grooming and well-being
Grooming and well-beingGrooming and well-being
Grooming and well-beingMRS
 
MRS Operations Network: GDPR - Organisational Measures
MRS Operations Network: GDPR - Organisational MeasuresMRS Operations Network: GDPR - Organisational Measures
MRS Operations Network: GDPR - Organisational MeasuresMRS
 
GDPR master class - transparent research projects
GDPR master class - transparent research projectsGDPR master class - transparent research projects
GDPR master class - transparent research projectsMRS
 
GDPR master class accountable research organisations (january 2018)
GDPR master class   accountable research organisations (january 2018)GDPR master class   accountable research organisations (january 2018)
GDPR master class accountable research organisations (january 2018)MRS
 
Operations network - consent under gdpr 24.01.2018
Operations network - consent under gdpr 24.01.2018Operations network - consent under gdpr 24.01.2018
Operations network - consent under gdpr 24.01.2018MRS
 
Leveragin research, behavioural and demeographic data
Leveragin research, behavioural and demeographic dataLeveragin research, behavioural and demeographic data
Leveragin research, behavioural and demeographic dataMRS
 

More from MRS (20)

Covid 19 research - wave 2
Covid 19 research - wave 2 Covid 19 research - wave 2
Covid 19 research - wave 2
 
Respondi whitepaper the 'appiness project
Respondi whitepaper the 'appiness projectRespondi whitepaper the 'appiness project
Respondi whitepaper the 'appiness project
 
Supporting good Mental Health at Work
Supporting good Mental Health at Work Supporting good Mental Health at Work
Supporting good Mental Health at Work
 
MRS Code of Conduct 2019 - Changes to Fair Data
MRS Code of Conduct 2019 - Changes to Fair DataMRS Code of Conduct 2019 - Changes to Fair Data
MRS Code of Conduct 2019 - Changes to Fair Data
 
Big Semiotics - May 2019
Big Semiotics - May 2019Big Semiotics - May 2019
Big Semiotics - May 2019
 
Digital Darwinism: How online communities can survive and thrive three waves ...
Digital Darwinism: How online communities can survive and thrive three waves ...Digital Darwinism: How online communities can survive and thrive three waves ...
Digital Darwinism: How online communities can survive and thrive three waves ...
 
How to write an Oppies Award Entry
How to write an Oppies Award EntryHow to write an Oppies Award Entry
How to write an Oppies Award Entry
 
MRS Roadshow 2019
MRS Roadshow 2019MRS Roadshow 2019
MRS Roadshow 2019
 
BBC Media Action - 2019
BBC Media Action - 2019BBC Media Action - 2019
BBC Media Action - 2019
 
Operations network meeting 22 January 2019
Operations network meeting 22 January 2019Operations network meeting 22 January 2019
Operations network meeting 22 January 2019
 
Using VR for immersion and audience engagement
Using VR for immersion and audience engagementUsing VR for immersion and audience engagement
Using VR for immersion and audience engagement
 
Humans v tech
Humans v tech Humans v tech
Humans v tech
 
Planning for new communities
Planning for new communitiesPlanning for new communities
Planning for new communities
 
Women in Ads
Women in AdsWomen in Ads
Women in Ads
 
Grooming and well-being
Grooming and well-beingGrooming and well-being
Grooming and well-being
 
MRS Operations Network: GDPR - Organisational Measures
MRS Operations Network: GDPR - Organisational MeasuresMRS Operations Network: GDPR - Organisational Measures
MRS Operations Network: GDPR - Organisational Measures
 
GDPR master class - transparent research projects
GDPR master class - transparent research projectsGDPR master class - transparent research projects
GDPR master class - transparent research projects
 
GDPR master class accountable research organisations (january 2018)
GDPR master class   accountable research organisations (january 2018)GDPR master class   accountable research organisations (january 2018)
GDPR master class accountable research organisations (january 2018)
 
Operations network - consent under gdpr 24.01.2018
Operations network - consent under gdpr 24.01.2018Operations network - consent under gdpr 24.01.2018
Operations network - consent under gdpr 24.01.2018
 
Leveragin research, behavioural and demeographic data
Leveragin research, behavioural and demeographic dataLeveragin research, behavioural and demeographic data
Leveragin research, behavioural and demeographic data
 

Recently uploaded

Call Girls In Holiday Inn Express Gurugram➥99902@11544 ( Best price)100% Genu...
Call Girls In Holiday Inn Express Gurugram➥99902@11544 ( Best price)100% Genu...Call Girls In Holiday Inn Express Gurugram➥99902@11544 ( Best price)100% Genu...
Call Girls In Holiday Inn Express Gurugram➥99902@11544 ( Best price)100% Genu...lizamodels9
 
Enhancing and Restoring Safety & Quality Cultures - Dave Litwiller - May 2024...
Enhancing and Restoring Safety & Quality Cultures - Dave Litwiller - May 2024...Enhancing and Restoring Safety & Quality Cultures - Dave Litwiller - May 2024...
Enhancing and Restoring Safety & Quality Cultures - Dave Litwiller - May 2024...Dave Litwiller
 
A DAY IN THE LIFE OF A SALESMAN / WOMAN
A DAY IN THE LIFE OF A  SALESMAN / WOMANA DAY IN THE LIFE OF A  SALESMAN / WOMAN
A DAY IN THE LIFE OF A SALESMAN / WOMANIlamathiKannappan
 
VIP Call Girls Gandi Maisamma ( Hyderabad ) Phone 8250192130 | ₹5k To 25k Wit...
VIP Call Girls Gandi Maisamma ( Hyderabad ) Phone 8250192130 | ₹5k To 25k Wit...VIP Call Girls Gandi Maisamma ( Hyderabad ) Phone 8250192130 | ₹5k To 25k Wit...
VIP Call Girls Gandi Maisamma ( Hyderabad ) Phone 8250192130 | ₹5k To 25k Wit...Suhani Kapoor
 
It will be International Nurses' Day on 12 May
It will be International Nurses' Day on 12 MayIt will be International Nurses' Day on 12 May
It will be International Nurses' Day on 12 MayNZSG
 
VIP Call Girls In Saharaganj ( Lucknow ) 🔝 8923113531 🔝 Cash Payment (COD) 👒
VIP Call Girls In Saharaganj ( Lucknow  ) 🔝 8923113531 🔝  Cash Payment (COD) 👒VIP Call Girls In Saharaganj ( Lucknow  ) 🔝 8923113531 🔝  Cash Payment (COD) 👒
VIP Call Girls In Saharaganj ( Lucknow ) 🔝 8923113531 🔝 Cash Payment (COD) 👒anilsa9823
 
MONA 98765-12871 CALL GIRLS IN LUDHIANA LUDHIANA CALL GIRL
MONA 98765-12871 CALL GIRLS IN LUDHIANA LUDHIANA CALL GIRLMONA 98765-12871 CALL GIRLS IN LUDHIANA LUDHIANA CALL GIRL
MONA 98765-12871 CALL GIRLS IN LUDHIANA LUDHIANA CALL GIRLSeo
 
Best VIP Call Girls Noida Sector 40 Call Me: 8448380779
Best VIP Call Girls Noida Sector 40 Call Me: 8448380779Best VIP Call Girls Noida Sector 40 Call Me: 8448380779
Best VIP Call Girls Noida Sector 40 Call Me: 8448380779Delhi Call girls
 
Monte Carlo simulation : Simulation using MCSM
Monte Carlo simulation : Simulation using MCSMMonte Carlo simulation : Simulation using MCSM
Monte Carlo simulation : Simulation using MCSMRavindra Nath Shukla
 
The Path to Product Excellence: Avoiding Common Pitfalls and Enhancing Commun...
The Path to Product Excellence: Avoiding Common Pitfalls and Enhancing Commun...The Path to Product Excellence: Avoiding Common Pitfalls and Enhancing Commun...
The Path to Product Excellence: Avoiding Common Pitfalls and Enhancing Commun...Aggregage
 
A305_A2_file_Batkhuu progress report.pdf
A305_A2_file_Batkhuu progress report.pdfA305_A2_file_Batkhuu progress report.pdf
A305_A2_file_Batkhuu progress report.pdftbatkhuu1
 
Boost the utilization of your HCL environment by reevaluating use cases and f...
Boost the utilization of your HCL environment by reevaluating use cases and f...Boost the utilization of your HCL environment by reevaluating use cases and f...
Boost the utilization of your HCL environment by reevaluating use cases and f...Roland Driesen
 
Call Girls Jp Nagar Just Call 👗 7737669865 👗 Top Class Call Girl Service Bang...
Call Girls Jp Nagar Just Call 👗 7737669865 👗 Top Class Call Girl Service Bang...Call Girls Jp Nagar Just Call 👗 7737669865 👗 Top Class Call Girl Service Bang...
Call Girls Jp Nagar Just Call 👗 7737669865 👗 Top Class Call Girl Service Bang...amitlee9823
 
M.C Lodges -- Guest House in Jhang.
M.C Lodges --  Guest House in Jhang.M.C Lodges --  Guest House in Jhang.
M.C Lodges -- Guest House in Jhang.Aaiza Hassan
 
KYC-Verified Accounts: Helping Companies Handle Challenging Regulatory Enviro...
KYC-Verified Accounts: Helping Companies Handle Challenging Regulatory Enviro...KYC-Verified Accounts: Helping Companies Handle Challenging Regulatory Enviro...
KYC-Verified Accounts: Helping Companies Handle Challenging Regulatory Enviro...Any kyc Account
 
B.COM Unit – 4 ( CORPORATE SOCIAL RESPONSIBILITY ( CSR ).pptx
B.COM Unit – 4 ( CORPORATE SOCIAL RESPONSIBILITY ( CSR ).pptxB.COM Unit – 4 ( CORPORATE SOCIAL RESPONSIBILITY ( CSR ).pptx
B.COM Unit – 4 ( CORPORATE SOCIAL RESPONSIBILITY ( CSR ).pptxpriyanshujha201
 
Best Basmati Rice Manufacturers in India
Best Basmati Rice Manufacturers in IndiaBest Basmati Rice Manufacturers in India
Best Basmati Rice Manufacturers in IndiaShree Krishna Exports
 
7.pdf This presentation captures many uses and the significance of the number...
7.pdf This presentation captures many uses and the significance of the number...7.pdf This presentation captures many uses and the significance of the number...
7.pdf This presentation captures many uses and the significance of the number...Paul Menig
 
Unlocking the Secrets of Affiliate Marketing.pdf
Unlocking the Secrets of Affiliate Marketing.pdfUnlocking the Secrets of Affiliate Marketing.pdf
Unlocking the Secrets of Affiliate Marketing.pdfOnline Income Engine
 

Recently uploaded (20)

Call Girls In Holiday Inn Express Gurugram➥99902@11544 ( Best price)100% Genu...
Call Girls In Holiday Inn Express Gurugram➥99902@11544 ( Best price)100% Genu...Call Girls In Holiday Inn Express Gurugram➥99902@11544 ( Best price)100% Genu...
Call Girls In Holiday Inn Express Gurugram➥99902@11544 ( Best price)100% Genu...
 
Enhancing and Restoring Safety & Quality Cultures - Dave Litwiller - May 2024...
Enhancing and Restoring Safety & Quality Cultures - Dave Litwiller - May 2024...Enhancing and Restoring Safety & Quality Cultures - Dave Litwiller - May 2024...
Enhancing and Restoring Safety & Quality Cultures - Dave Litwiller - May 2024...
 
A DAY IN THE LIFE OF A SALESMAN / WOMAN
A DAY IN THE LIFE OF A  SALESMAN / WOMANA DAY IN THE LIFE OF A  SALESMAN / WOMAN
A DAY IN THE LIFE OF A SALESMAN / WOMAN
 
VIP Call Girls Gandi Maisamma ( Hyderabad ) Phone 8250192130 | ₹5k To 25k Wit...
VIP Call Girls Gandi Maisamma ( Hyderabad ) Phone 8250192130 | ₹5k To 25k Wit...VIP Call Girls Gandi Maisamma ( Hyderabad ) Phone 8250192130 | ₹5k To 25k Wit...
VIP Call Girls Gandi Maisamma ( Hyderabad ) Phone 8250192130 | ₹5k To 25k Wit...
 
It will be International Nurses' Day on 12 May
It will be International Nurses' Day on 12 MayIt will be International Nurses' Day on 12 May
It will be International Nurses' Day on 12 May
 
VIP Call Girls In Saharaganj ( Lucknow ) 🔝 8923113531 🔝 Cash Payment (COD) 👒
VIP Call Girls In Saharaganj ( Lucknow  ) 🔝 8923113531 🔝  Cash Payment (COD) 👒VIP Call Girls In Saharaganj ( Lucknow  ) 🔝 8923113531 🔝  Cash Payment (COD) 👒
VIP Call Girls In Saharaganj ( Lucknow ) 🔝 8923113531 🔝 Cash Payment (COD) 👒
 
MONA 98765-12871 CALL GIRLS IN LUDHIANA LUDHIANA CALL GIRL
MONA 98765-12871 CALL GIRLS IN LUDHIANA LUDHIANA CALL GIRLMONA 98765-12871 CALL GIRLS IN LUDHIANA LUDHIANA CALL GIRL
MONA 98765-12871 CALL GIRLS IN LUDHIANA LUDHIANA CALL GIRL
 
Best VIP Call Girls Noida Sector 40 Call Me: 8448380779
Best VIP Call Girls Noida Sector 40 Call Me: 8448380779Best VIP Call Girls Noida Sector 40 Call Me: 8448380779
Best VIP Call Girls Noida Sector 40 Call Me: 8448380779
 
Monte Carlo simulation : Simulation using MCSM
Monte Carlo simulation : Simulation using MCSMMonte Carlo simulation : Simulation using MCSM
Monte Carlo simulation : Simulation using MCSM
 
The Path to Product Excellence: Avoiding Common Pitfalls and Enhancing Commun...
The Path to Product Excellence: Avoiding Common Pitfalls and Enhancing Commun...The Path to Product Excellence: Avoiding Common Pitfalls and Enhancing Commun...
The Path to Product Excellence: Avoiding Common Pitfalls and Enhancing Commun...
 
A305_A2_file_Batkhuu progress report.pdf
A305_A2_file_Batkhuu progress report.pdfA305_A2_file_Batkhuu progress report.pdf
A305_A2_file_Batkhuu progress report.pdf
 
Boost the utilization of your HCL environment by reevaluating use cases and f...
Boost the utilization of your HCL environment by reevaluating use cases and f...Boost the utilization of your HCL environment by reevaluating use cases and f...
Boost the utilization of your HCL environment by reevaluating use cases and f...
 
Call Girls Jp Nagar Just Call 👗 7737669865 👗 Top Class Call Girl Service Bang...
Call Girls Jp Nagar Just Call 👗 7737669865 👗 Top Class Call Girl Service Bang...Call Girls Jp Nagar Just Call 👗 7737669865 👗 Top Class Call Girl Service Bang...
Call Girls Jp Nagar Just Call 👗 7737669865 👗 Top Class Call Girl Service Bang...
 
M.C Lodges -- Guest House in Jhang.
M.C Lodges --  Guest House in Jhang.M.C Lodges --  Guest House in Jhang.
M.C Lodges -- Guest House in Jhang.
 
KYC-Verified Accounts: Helping Companies Handle Challenging Regulatory Enviro...
KYC-Verified Accounts: Helping Companies Handle Challenging Regulatory Enviro...KYC-Verified Accounts: Helping Companies Handle Challenging Regulatory Enviro...
KYC-Verified Accounts: Helping Companies Handle Challenging Regulatory Enviro...
 
B.COM Unit – 4 ( CORPORATE SOCIAL RESPONSIBILITY ( CSR ).pptx
B.COM Unit – 4 ( CORPORATE SOCIAL RESPONSIBILITY ( CSR ).pptxB.COM Unit – 4 ( CORPORATE SOCIAL RESPONSIBILITY ( CSR ).pptx
B.COM Unit – 4 ( CORPORATE SOCIAL RESPONSIBILITY ( CSR ).pptx
 
Best Basmati Rice Manufacturers in India
Best Basmati Rice Manufacturers in IndiaBest Basmati Rice Manufacturers in India
Best Basmati Rice Manufacturers in India
 
7.pdf This presentation captures many uses and the significance of the number...
7.pdf This presentation captures many uses and the significance of the number...7.pdf This presentation captures many uses and the significance of the number...
7.pdf This presentation captures many uses and the significance of the number...
 
Unlocking the Secrets of Affiliate Marketing.pdf
Unlocking the Secrets of Affiliate Marketing.pdfUnlocking the Secrets of Affiliate Marketing.pdf
Unlocking the Secrets of Affiliate Marketing.pdf
 
VVVIP Call Girls In Greater Kailash ➡️ Delhi ➡️ 9999965857 🚀 No Advance 24HRS...
VVVIP Call Girls In Greater Kailash ➡️ Delhi ➡️ 9999965857 🚀 No Advance 24HRS...VVVIP Call Girls In Greater Kailash ➡️ Delhi ➡️ 9999965857 🚀 No Advance 24HRS...
VVVIP Call Girls In Greater Kailash ➡️ Delhi ➡️ 9999965857 🚀 No Advance 24HRS...
 

MRS Advanced Analytics Innovation Symposium Presentation

  • 1. MRS Advanced Analytics Innovation Symposium 30th April 2015 #MRSlive
  • 2. Brand Share & Industry Size: Will the twain ever meet? Using a portfolio of techniques to improve accuracy of market volume evolution in price change scenarios April 2015 Sreeram Srinivasan, IMRB International Ranjana Gupta, IMRB International
  • 3. 30 40 Marketer’s pricing dilemma 30 Market volum e 100 bn Marke t volum e 70 bn With new prices, my share will grow… … but will my volume also grow? Market share grows….. ….but, market size shrinks
  • 4. Both the consumer and the macro variables needed to answer the questions.. C o n s u m e r M a c r o Brand choice Same brand, switching etc. In context of price Category choice Frequency, consumption, substitutes etc. Industry Past volumes, substitutes, prices etc. Economy Income, affordability, Inflation etc. Relies on past trends Future may be different Future oriented. Past learnings not fully leveraged
  • 5. Hence for accurate volume forecasts, no single methodology can provide complete answer Consumer research and macro-economic models should be integrated Choice Model Consumer Research • Brand shares • Switching Econometric Macro Modelling • Market size
  • 6. An approach to integration of choice and econometric model However, before integration, the individual tools need to be refined, adapted…
  • 7. …to account for the nature of the category How to account for occasion? Is it an impulse or considered purchase? Is it a repertoire or non- repertoire category? Do number of units matter? What about frequency of purchase? How can we ensure that the respondent reacts to only relevant offers?
  • 9. The questioning technique uses FORCE principle to make the consumer response more realistic… Familiar Only brands that the respondent interacts with / likely to interact with shown – customized for each respondent real time Evoked set created using respondent’s current repertoire, past usage and future disposition Occasio n Respondent’s answers using an occasion as a context in occasion led categories like CSDs, Snacks etc. There can be other household purchase categories where occasion is irrelevant Repertoir e Respondents allowed to select multiple brands, SKUs and units, as they might in real life Channel Primary channel of purchase identified and specified in the questioning Event Recurren ce Frequency of purchase
  • 10. Here’s an example… around 20 such choice tasks shown Imagine you are doing your monthly grocery shopping from the supermarket and you have to buy bathing soaps. On the shelf you see the following brands with the given prices? Which brands are you likely to buy? You can choose as many brands as you like. Or you can walk out of the shop without buying any. Do state the number of packs that you would be buying. I will not buy any anything 2 Johnson’s Baby Soap 75 gms Rs. 30 Dove 75 gms Rs. 40 Santoor – pack of 4 100gms X 4 Rs. 50 Now at Rs. 40 Pears 100 gms + 25 gms extra free Rs. 25 Rexona 200 gms Rs. 25 I would buy a shower gel Lux 100 gms Rs. 10 1
  • 11. Consumer choices are converted into utilities – two levels of calculations Main Effect For every level within each attribute at a respondent level Example: Utility or preference for Brands like Dove, Pears etc., for SKUs like 100 gms,, 75 gms and for various price levels Cross Effect Interaction between attributes Example: Utility for Dove by itself and for Dove at a particular price may be different The utilities are transformed into share of preference and weighted to give the shares in various scenarios For all existing brands & SKUs, the current levels of distribution are built into the model – to ensure current scenario shares are in line with actual market shares Respondent level estimation of preference helps in calculation of gains and losses from one brand to another
  • 12. The output Current Scenario New price scenario (Client’s brands cut prices by 10%) Company Brand X 20.0% 21.8% Brand Y 17.0% 17.2% Brand A 10.0% 9.9% Brand B 5.0% 4.4% Brand P 8.0% 7.6% Brand Q 15.0% 14.8% Brand R 10.0% 9.7% Brand S 15.0% 14.6% 39%37% Brand X Gaining From Brand B, Brand S Net Gain/Loss 1.8%
  • 14. The input Past volumes Population Substitute categories (Real) Price Purchase frequency No. of packs(Real) Income Basket size
  • 15. The statistics Options Simple Regression Volume = fn (Price) Easy but can lead to situations like increase price to increase volumes! Simple Time Series Volume = fn (Past volumes) Builds in past volume trends but assumes that history will definitely repeat itself Price Volume
  • 16. The statistics Adopted method ARIMA (Auto Regressive Integrated Moving Average) Volume = fn (Real Price, Past Volume, Real Income, etc….) Moving average included – accounts for any possible prediction error in previous time periods Accounts for autocorrelation Better prediction accuracy
  • 17. The statistics Adopted method ARIMA (Auto Regressive Integrated Moving Average) Deseasonalized data - predict organic change in volumes Model by major sub-groups to account for different trends – break the market Price gap between sub-groups used – inter-movement built in
  • 18. The output Market volumes 125000 tonnes Current scenario 126250 tonnes New price scenario 1%
  • 19. Bringing the twain together
  • 20. It’s quite simple actually… Shares X Market Volumes = Company Volumes 37 Current Scenario 125000 tonnes New Scenario Share Market Volumes 46250 tonnesCompany volumes 39 126250 tonnes Share Market Volumes 49238 tonneCompany volumes Share change: 5% Volume change: 6%
  • 21. Proof of the pudding
  • 22. Results validated across markets 0 20 40 60 80 100 Predicted volume accuracy by brands in various markets (74 data points in this graph) Average : 88% The trick: improve accuracy of the individual models
  • 23. Identifying the relative impact of touchpoints: A tailored statistical technique for real-time data Shane Baxendale, Cranfield School of Management Heval Ceylan-Gilchrist, MESH MRS ADVANCED ANALYTICS NETWORK 30th April 2015
  • 24. Why are we here today?  Real-time Experience Tracking Methodology  Analysis - using linear mixed-effects regression 24
  • 25. Our thinking  Consumers experience brands through multiple channels (not just advertising!)  Brand experiences influence a consumer’s attitude toward brands  The majority of existing literature focusses on the impact of one or two types of experience What impact are different encounters having on consumer attitudes toward the brand? 25 *Baxendale S., Macdonald E.K., Wilson H.N., (2015), The impact of different touchpoints on brand consideration, Journal of Retailing, 37(2), 203.
  • 26. Real-time Experience Tracking (RET) ONLINE REAL-TIME ONLINE Day 9Day 2 - 8Day 1
  • 27. Text us whenever you see, hear or experience anything to do with the following brands… Text framework 27 BRAND: A)Brand A B)Brand B C)Brand C D)Brand D E)Brand E F) Other OCCASION: A)TV B)Poster/Billboard C)Radio D)In store E)Cinema F)Newspaper G)Magazine H)Conversation I)Online/Mobile J)Mailing/leaflet K)Me Purchasing L)Me using M)Someone else using N)Sponsorship O)Other FEELING: 5)Very positive 4)Fairly positive 3)Neutral 2)Fairly negative 1)Very Negative CHOICE: 5)Much more likely to choose 4)Slightly more likely to choose 3)No difference 2)Slightly less likely to choose 1) Much less likely to choose
  • 28. Imagine you experienced Brand A Online… …you would text: 28 a 5i 5 CHOICE: 5) Much more likely to choose 4) Slightly more likely to choose 3) No change 2) Slightly less likely to choose 1) Much less likely to choose ENGAGEMENT: 5) Very positive 4) Fairly positive 3) Neutral 2) Fairly negative 1) Very negative BRAND: a) Brand A b) Brand B c) Brand C d) Brand D e) Brand E f) Other OCCASION: a) TV b) Poster/Billboard c) Radio d) In store e) Cinema f) Newspaper g) Magazine h) Conversation i) Online/Mobile j) Mailing/Leaflet k) Me purchasing l) Me using m) Someone else using n) Sponsorship o) Other Which brand was it? Where did you experience it? How likely did it make you to choose the brand next time? How did it make you feel?
  • 29. Now tell us more in an online diary… 29 This is an individual’s experience log By clicking on each entry, the experience can be expanded upon in detail Wednesday 13th February 2012,11:54 Wednesday13th February2012,11:54 Wednesday13th February2012,10:22 Tuesday512h February2012,18:46 Tuesday12th February2012,13:05 Tuesday12th February2012,08:38 Brand A,Online, Very Positive, MuchMoreLikely toChoose Brand C,Conversation, Very Negative, MuchLess Likely toChoose Brand E, TV, Fairly Positive, Slightly More Likely to Choose Other, Instore, Fairly Positive, Slightly morelikely toChoose Brand B, Mailing/Leaflet, Slightly Negative, Nochange Brand A Brand’s website Very positive Much more likely to choose I was looking on the brand website to find out more information about the product range. Looks like there are some good options. Look for product info 13/02/2012, 11:54 Please tell us exactly what yousaw? : What was the purpose ofyour online activity? : Brand’s website Ad from brand In the news Social networking site Price comparison site Other For each level of data captured in real-time we can tailor extra questions to get more granular information in near-time
  • 30. Data For one individual Brand A Brand B … Brand N Consideration Wk0 Consideration Wk1 Consideration Wk0 Consideration Wk1 Consideration Wk0 Consideration Wk1 Freq. & Pos. Brand Ad Retailer Ad In Store WOM … Freq. & Pos. Brand Ad Retailer Ad In Store WOM … Freq. & Pos. Brand Ad Retailer Ad In Store WOM … 30 +ve -ve -ve
  • 31. Model © Cranfield University 31 Change in Consideration = ??? Data Rationale Considerations Implications Demographics / Participant information Certain consumer groups may be more / less likely to change their opinion towards brands over time Multiple responses per participant means that we can learn more about individual tendencies Need to account for the homogeneity in repeated responses, therefore random effects modelling Frequency of experience More experiences can be a positive impact (via reinforcement of messaging) or negative (over-exposure) There could be many potential ways of including this in the model; Constant effect Diminishing returns Check the validity of the model by testing multiple approaches Positivity of experience A consumers’ perception of an experience can determine the impact it has on them How do we account for positivity over multiple encounters? Average? Check the validity of the model by testing multiple approaches
  • 32. Parameter operation Frequency © Cranfield University 32 0 1 2 3 4 5 0 1 2 3 4 5 0 1 2 3 4 5 Exposure Impact = x if Freq.>0 Increasing Impact Impact = x*Freq. Diminishing Returns Impact = x*ln(Freq.+1) Impact = x*Freq. - y*Freq.^2
  • 33. Parameter operation Positivity 1. Average positivity across experiences 2. Average positivity and variance of positivity 3. Positivity of last experience 4. Freq. of positive and Freq. of negative experiences © Cranfield University 33
  • 34. Results Focal Frequency Positivity In-store communications =1 1 Brand advertising =1 =2 Retailer advertising =1 =2 Peer observation =1 =4 Traditional earned =5 =4 WOM =5 6 © Cranfield University 34 Competitor Frequency Positivity In-store communications 1 =1 WOM =2 =1 Peer observation =2 =1 Retailer advertising =4 =1 Brand advertising =4 =1 Traditional earned 6 =1
  • 36. The contents of this document are the sole and confidential property of Lieberman Research Worldwide, and may not be reproduced or distributed without the express written permission of Lieberman Research Worldwide. Prepared for CLIENT TITLE LRW Europe BAYESIAN ANALYSIS FOR MARKETING IMPACT April 2015 LRW Europe 1, Heathcock Court, 415, Strand London WC2R 0NT Prepared by: Adele Gritten & Graham Williams for MRS Advanced Analytics Conference
  • 37. ‹#› © 2015 Lieberman Research Worldwide. All rights reserved. CONFIDENTIAL.  Why should our industry care about BayesNets?  What is BayesNets?  What are the Advantages of BayesNets?  Illustrative outputs  Live UK Case Study  Summary TODAY’S AGENDA
  • 38. ‹#› © 2015 Lieberman Research Worldwide. All rights reserved. CONFIDENTIAL. Why should our Industry Care about BayesNets?
  • 39. ‹#› © 2015 Lieberman Research Worldwide. All rights reserved. CONFIDENTIAL. Why Should Our Industry Care About BayesNets?  BayesNets is a unique and more comprehensive driver analysis to assist Marketers  Overcomes the shortcomings of traditional [drivers analysis] methods  Allows the integration of profiling, behavioural and other metrics with attitudinal/preference ratings to create a more holistic view of what drives the dependent variable “Bayesnets has played a major role in several recent wins. It’s especially helpful with brand positioning research where the complex relationships between brand attributes demands a more nuanced and flexible approach to analysis”. LRW Account Director
  • 40. ‹#› © 2015 Lieberman Research Worldwide. All rights reserved. CONFIDENTIAL. What is BayesNets?
  • 41. ‹#› © 2015 Lieberman Research Worldwide. All rights reserved. CONFIDENTIAL. What is BayesNets? Think of BayesNets as “drivers analysis on steroids” Let’s review the logic and goals of “key drivers analysis”:  There is a market attitude or behaviour – the “target outcome” – which:  A client needs to favourably influence in the marketplace, but which…  They cannot influence directly  So we need to find the best “levers to pull” through which to indirectly influence that attitude or behaviour; e.g. customer attitudes or perceptions:  Which we can influence through product/service design or marketing, and…  Which have strong “derived importance” in driving the “target outcome”  Ultimately, the purpose of “key drivers analysis” is to empirically identify the best “levers to pull” for maximum in market impact. BayesNets:  Can be thought of generally as a more powerful key drivers analysis methodology  Offers a number of significant advantages compared to hitherto commonly used key drivers analysis approaches
  • 42. Where Did BayesNets Come From? | Thomas Bayes “Bayesian” refers to Reverend Thomas Bayes’ Theorem from the 18th century that paved the way for data to be used in prediction. Bayes’ Theorem basically allows us to look at multi-directional probabilities.  18th Century English statistician, philosopher, and minister  Formulated Bayes’ Theorem: a mathematical expression of probabilities from observed data  Hotly debated and contested by Frequentists until recent years 42 © 2015 Lieberman Research Worldwide. All rights reserved. CONFIDENTIAL.
  • 43. ‹#› © 2015 Lieberman Research Worldwide. All rights reserved. CONFIDENTIAL. What are BayesNets? | A Model of Relationships Bayesian networks (BayesNets) are a type of “path analysis” model that simultaneously describes the relationships between variables in a network system based on joint probabilities between the variables. BayesNets improves upon models that use advanced correlation and regression techniques (such as most regression analyses and Structural Equation Modeling). Basically, it’s a better way of understanding interactions between independent variables as they drive dependent variables. WET PAVEMENT SPRINKLER RAIN SLIPPERY Probabilistic Relationship Probabilistic Relationship Variables or Factors Variables or Factors High-Dimensional Probability Hypercube
  • 44. ‹#› © 2015 Lieberman Research Worldwide. All rights reserved. CONFIDENTIAL. What are the Advantages of BayesNets?
  • 45. ‹#› © 2015 Lieberman Research Worldwide. All rights reserved. CONFIDENTIAL. What are Advantages of BayesNets? BayesNets’ advantages over more commonly used techniques are “technical” but nonetheless significant. BayesNets’ advantages include:  Does fully interactive, “multivariate” modeling  Not “confused” by multicollinearity  No implicit assumption of “linear” relationships  Siloed Regression Models don’t capture indirect or interaction effects BayesNets helps us find the best model:  We don’t have to “hypothesise” the structure of the multivariate network  Rather, BayesNets’ “machine learning” algorithms seek out the best network structures quickly & cost effectively  From there we bring in the “art” that mixes with the “science” to yield a highly actionable understanding of what drives the target outcome – the dependent variable – in the marketplace. Traditional Approach Approach with BayesNets Variables or Factors Independent Variables or Factors
  • 46. ‹#› © 2015 Lieberman Research Worldwide. All rights reserved. CONFIDENTIAL. Siloed Regression Models Don’t Capture Indirect or Interaction Effects Key Satisfaction/Loyalty Metric ProductsAtmosphere Store Experience Minutes in line Seconds Ordering Speed of Ordering Ordering Process Volume Customer Service Greeted by Employee Friendly Employees EXPERIENCE DOMAINS PERCEPTIONS OF BEHAVIORS QUANTIFIABLE MICRO- BEHAVIORS SUPPORTING IMPRESSIONS Metrics can impact other domains, not just those up the ladder in our hierarchy
  • 47. ‹#› © 2015 Lieberman Research Worldwide. All rights reserved. CONFIDENTIAL. The Former Best Approach: SEM Structural Equation Modeling  Tests Complex Structures  Interactive & Indirect Effects  Explanatory & Prescriptive  Multicollinearity may still be a problem  Can only test the “fit” of specifically hypothesized networks Siloed Regression Tree  Forces Simple Structure  Direct Effects Only  Diagnostic & Descriptive
  • 48. ‹#› © 2015 Lieberman Research Worldwide. All rights reserved. CONFIDENTIAL. Illustrative Outputs
  • 49. ‹#› © 2015 Lieberman Research Worldwide. All rights reserved. CONFIDENTIAL. Variables or Factors Colours identify “nodes” that belong to different factors Probabilistic Relationship “Arcs” connect the various “nodes” in the network What does it look like? | It Starts With Networks
  • 50. ‹#› © 2015 Lieberman Research Worldwide. All rights reserved. CONFIDENTIAL. BayesNets Satisfaction Key Drivers Analysis: A Case Study  Customer satisfaction surveys with >650,000 retail customers  Surveys conducted throughout 2013  Dependent variable: Overall quality of in-store experience rating  Independent variables:  Is a place for someone like me  Clothing was neatly displayed and well organized  The wait time in the checkout line was acceptable  Service you received in the fitting room met your needs  The cashier worked quickly and efficiently to check out all customers in line  Your experience in the store was more fun and engaging than other stores you typically shop  The signs clearly indicated what was on sale  Employees were easily accessible  Employees were willing to find style, color, size  Employees acknowledged and made you feel welcome  Employees seemed genuinely glad you were there  Overall clothing quality
  • 51. ‹#› © 2015 Lieberman Research Worldwide. All rights reserved. CONFIDENTIAL. BayesNets Customer Satisfaction Drivers Analysis - Retail Example Wait time at checkout Feel welcomed For someone like me Overall clothing quality Neat displays /organised Sale signs clear More fun and engaging than other stores Accessible employees Employees glad you were there Cashier worked quickly and efficiently Employees willing to find style, colour, size Service received in fitting room met needs Overall experience
  • 52. ‹#› © 2015 Lieberman Research Worldwide. All rights reserved. CONFIDENTIAL. BayesNets “total effects” analysis looks & feels like standard drivers analysis All Brands Retailer 1 Retailer 2 Retailer 3 Retailer 4 Retailer 5 Total Effects on Target Promoter_Rev Indexed Indexed Indexed Indexed Indexed Indexed Standardized Standardized Standardized Standardized Standardized Standardized Total Effects Total Effects Total Effects Total Effects Total Effects Total Effects F1: MERCHANDISE FOR ME 144.3 134.3 159.5 137.4 150.9 133.1 F0: EMOTIONALLY ENGAGED 131.9 114.6 127.8 110.5 115.4 111.7 F17: GREAT VALUE 125.8 121.6 126.8 112.7 129.7 118.6 F4: EXCITING AND STYLISH MERCHANDISE 125.0 111.6 147.3 115.9 102.7 116.7 F5: QUALITY BRANDS 120.1 111.1 132.7 152.7 111.7 115.6 F3: GREAT FIT & SIZES 112.8 114.9 112.0 132.3 119.1 115.3 F7: GREAT PRICES AND SAVINGS 110.0 120.6 141.9 117.6 95.1 114.7 F14: MERCHANDISE FOR MY HOME AND FAMILY 109.8 107.1 146.2 75.6 128.3 105.7 F11: GREAT SALES 105.1 111.6 102.2 95.8 117.5 106.3 F2: ENJOYABLE SHOPPING 103.9 94.0 56.8 101.3 105.5 100.5 I can always count on STORE to have what I want on sale 102.9 103.2 104.2 69.5 109.1 104.4 F12: BETTER DEALS 101.0 108.3 126.9 99.9 97.2 102.0 F13: EASY RETURN POLICY 93.4 93.4 104.9 107.1 91.4 101.5 F9: PRICES I TRUST 80.6 78.5 98.0 61.9 90.8 74.7 F6: LOYALTY PROGRAM 79.0 78.7 56.3 59.3 52.2 75.2 F8: COUPONS 72.2 93.5 39.0 85.7 75.1 89.1 F16: INSPIRING DISPLAYS 70.0 78.7 26.6 120.5 74.3 95.7 F18: SUPPORTS MY COMMUNITY 61.9 61.5 68.8 71.7 72.7 73.0 F10: EASY PROMOTION 50.5 62.7 22.1 72.5 61.4 46.3 Illustrative
  • 53. ‹#› © 2015 Lieberman Research Worldwide. All rights reserved. CONFIDENTIAL. The output here gives specific advice on which factors to affect first, and when it is optimal to focus on the next factor. Bayesian Analysis can also provide clear recommendations on where businesses should focus Initial Mean Rating Mean Rating After Improving Preceding Factors Target Mean Overall Opinion Mean Initial Value 4.45 HEALTHY 9.07 9.27 4.79 SELECTION/VARIETY 8.64 8.84 9.05 4.88 QUALITY 8.25 8.40 8.59 4.91 EASY/SIMPLE 8.05 8.56 8.86 4.92 First, the goal is to move the mean on the Healthy factor from 9.07 to 9.27 This would increase the Overall Opinion 0.34 points, from 4.45 to 4.79 Moving the Healthy mean from 9.07 to 9.27 also affects Selection/Variety, moving it from 8.64 to 8.84. Illustrative Moving the Selection mean from 8.84 to 9.05 similarly impacts both the Overall Opinion mean (up to 4.88) and the Quality mean (moving it to 8.40) and so on for each successive factor
  • 54. ‹#› © 2015 Lieberman Research Worldwide. All rights reserved. CONFIDENTIAL. UK Media Owner Client: Live Case Study Current data sources include: • Brand Tracking with image metrics • Industry audience measurement • Content audit v competitors • Content appreciation • Social Media tracking They tend to do a rough & ready comparison and they are doing some KDA in the tracking data, but so far no joined up stuff With so much data they worry about how much stakeholders trust or care about the data ‘I have an overload of data and metrics and I want to see what combination of factors drive audience growth (or decline).... At the moment I can’t see the wood for the trees – I’m hoping Bayesnet will help’ LRW are working with the client to initially conduct a Bayesnet analysis on the monthly tracker (which goes back over 2 years) to identify relationship between behaviour and the metrics and which ones are the ones that they really need to look at Ideal solution would be to stream line the cumbersome tracker – strip out metrics that don’t add value and then look to widen the Bayesnet analysis to other data sources and conduct a wider analysis
  • 55. ‹#› © 2015 Lieberman Research Worldwide. All rights reserved. CONFIDENTIAL. Summary
  • 56. ‹#› © 2015 Lieberman Research Worldwide. All rights reserved. CONFIDENTIAL. In Summary | Why BayesNet Modeling?  BayesNets modeling is often more effective than more traditional advanced modeling of derived importance analysis  BayesNet measures both direct impact on the dependent variables and indirect impacts through other independent variables in the model  BayesNets overcomes multicollinearity and makes no assumptions of either normal distributions of data or linear correlations between variables.  BayesNets mathematics and software allow for quicker creation of the model, optimizations and “what if” scenarios.  BayesNets offers more effective optimization modeling with target means to guide activation and appropriate levels of effort and investment.
  • 57. ‹#› © 2015 Lieberman Research Worldwide. All rights reserved. CONFIDENTIAL. If you’d like more info: We can set up a time for you to talk to one of our genuine experts! Mick McWilliams PhD, Sr. VP, Marketing Science Marketing scientist specializing in segmentation, brand engagement, database scoring, SEM, KDA and BayesNets 25+ years of MR experience with specialties in neuroscience studies & evolutionary psychology PhD, Sociology, Virginia Polytechnic Institute & State University Thank you!
  • 58. ‹#› © 2015 Lieberman Research Worldwide. All rights reserved. CONFIDENTIAL. Graham Williams Research Director, Europe gwilliams@LRWonline.com Lieberman Research Worldwide 1, Heathcock Court, 415 Strand London WC2R ONT www.lrwonline.com Direct Line: 0203 551 7075 Contact Information
  • 59. Copyright © Nepa All Rights Reserved Space Optimisation MRS Advanced Analytics 30 April 2015
  • 60. Copyright © Nepa All Rights Reserved 60 Space Optimisation is the process of maximising profit by allocating the appropriate amount of store shelf space to each product category
  • 61. Copyright © Nepa All Rights Reserved Typical client – a retail chain with a wealth of sales and loyalty club data 61 ...of different locations and demographic profiles Many hundred stores of different sizes Nearly 100 product categories
  • 62. Copyright © Nepa All Rights Reserved Many aspects determine the profit that a product category will yield. First, the most important ones are identified Significant factors, in selection : • Affluence in neighbourhood • Gender profile • Age profile • Proximity to low-cost competition • ... • ... • ... 62 Demography Location / competition Sales details Customer Satisfaction
  • 63. Copyright © Nepa All Rights Reserved Linear regression is used to isolate the relationship between space and profit, per product category 63 Space allocated ... ... Affluence in neighbourhood Gender profile Age profile Proximity to low-cost competition ... ... β1 + β2 + β3 + β4 + β5 + βs
  • 64. Copyright © Nepa All Rights Reserved ”Space elasticity” – not the same for all product types, illustrated by an intuitive example from a pharmacy 64 1 shelf Margin: £500 per day  £500 /shelf 4 shelves Margin: £1000 per day  £250 /shelf +? Margin Space Margin Space Due to its higher space elasticity, it is likely more profitable to add another beauty shelf than one for pain killers. This despite the fact that painkillers presntly give more profit per shelf unit.
  • 65. Copyright © Nepa All Rights Reserved Store-specific linear regression gives accurate space elasticity curves in steps, for each category 65 Shelf space allocated ... ... Affluence in neighbourhood Gender profile Age profile Proximity to low-cost competition ... ... βs
  • 66. Copyright © Nepa All Rights Reserved Store-specific linear regression gives accurate space elasticity curves in steps, for each category 66 0 20 40 60 80 100 120 0 20 40 60 Margin(£) Space (Shelf sections) Shelf space allocated ... ... Affluence in neighbourhood Gender profile Age profile Proximity to low-cost competition ... ... βs
  • 67. Copyright © Nepa All Rights Reserved Store-specific linear regression gives accurate space elasticity curves in steps, for each category 67 0 20 40 60 80 100 120 0 50 100 150 200 Margin(£) Space (Shelf sections) Shelf space allocated ... ... Affluence in neighbourhood Gender profile Age profile Proximity to low-cost competition ... ... βs
  • 68. Copyright © Nepa All Rights Reserved Store-specific linear regression gives accurate space elasticity curves in steps, for each category 68 0 10 20 30 40 50 60 70 80 90 0 50 100 150 200 Margin(£) Space (shelf sections) Shelf space allocated ... ... Affluence in neighbourhood Gender profile Age profile Proximity to low-cost competition ... ... βs
  • 69. Copyright © Nepa All Rights Reserved Store-specific linear regression gives accurate space elasticity curves in steps, for each category 69 0 10 20 30 40 50 60 70 80 90 0 50 100 150 200 Margin(£) Space (shelf sections) Shelf space allocated ... ... Affluence in neighbourhood Gender profile Age profile Proximity to low-cost competition ... ... βs
  • 70. Copyright © Nepa All Rights Reserved Store-specific linear regression gives accurate space elasticity curves in steps, for each category 70 0 10 20 30 40 50 60 70 80 0 10 20 30 40 50 Margin(£) Space (shelf sections) Shelf space allocated ... ... Affluence in neighbourhood Gender profile Age profile Proximity to low-cost competition ... ...
  • 71. Copyright © Nepa All Rights Reserved We will never start adding delicassy cheeses, since the start of the curve is so flat. Combining curves for the optimal space allocation – stepwise incremental assignment doesn’t always find the best solution available 71 0 10 20 30 40 50 60 70 80 90 100 0 1 2 3 4 5 6 7 8 9 10 Margin(£) ABBBC C D D E F GG/HIJK Store plan, 15 shelves 8 7 Stepwise adding products to shelves using the highest incremental value at each step will result in assigning 8 to vegetables and 7 to sauces
  • 72. Copyright © Nepa All Rights Reserved The optimal distribution includes many shelves of delicassy cheeses, giving a large profit at substantial space assignment 72 0 10 20 30 40 50 60 70 80 90 100 0 1 2 3 4 5 6 7 8 9 10 Margin(£) ABBBC C D D E F GG/HIJK Store plan, 15 shelves 96
  • 73. Copyright © Nepa All Rights Reserved We search through all possible combinations to find the best one – an enormous optimisation problem which we use logic to reduce 100 shelf units to allocate 73 30 categories ... ... ... ... ... ... ... ... 6 x 1028 combinations! ... ... Even this rather moderate number of shelf units and categories presents an enormous number of potential combinations. We need to use logic to reduce the computational complexity, and find the best solution available.
  • 74. Copyright © Nepa All Rights Reserved An online tool is used for space allocation, bespoke to each individual store 74
  • 75. Copyright © Nepa All Rights Reserved Thank you! kalle.backlund@nepa.com 0785-19 49 379 75

Editor's Notes

  1. When I talk to companies about this most people tell me they expected me to be older
  2. Volume = α + β. price + ¥. Last year Volume (Autoregressive component)+ Ɵet (Moving Average) Where ¥. Last year Volume is the Autoregressive component – accounts for the fact that Yt is dependent on Yt-1 I = Yt – Yt-1 (The difference in the volumes between two consecutive periods) Ɵet is the MA term which assumes et is dependent on et-1 Simple regression will not work due to autocorrelation AR is used to account for autocorrelation whereas MA is used to account for any errors in the previous time period We do not use the integration factor into our model and only use ARMA due to the fact that we already deseasonalize the input variables. Then there would be an issue of double correction.
  3. Rationale for the analysis
  4. But what data do we have? And what links do we hypothesise?
  5. So what do we want to include in the model, why, and what do we have to consider as a result of this?
  6. How should we include frequency? Which is best from the model? Does this make sense / are we comfortable with choosing this over other options?
  7. How should we include positivity? Which is best from the model? Does this make sense / are we comfortable with choosing this over other options?
  8. What are the results?
  9. Done to a large extent on gut feel. We will see that there are large money to make by relatively simple steps.
  10. EXAMPLES of categories Grocery store, like Co-op. Second in Sweden. Clarify that we are looking at allocating space for each of those product categories. Done to a large extent on gut feel. We will see that there are large money to make
  11. Go through the matrix quickly! Two criteria: Drive margin in categories (+ or -), and be relatively un-correlated to each other
  12. R^2, around 0.6 (for the best categories)