Department of MBA

IMS Engineering College, Ghaziabad

Marketing Analytics
Ashish Awasthi, Assistant Professor,
UNIT 1
Department of MBA

Introduction to Marketing Analytics

Marketing Analytics: Data Driven Techniques with Microsoft Excel Wayne L
Winston, Wiley India Pvt Ltd.
 Marketing Analytics – Meaning
• Marketing analytics is the practice of measuring and
analyzing data and metrics (parameters) to
understand the impact of marketing activities,
maximize ROI (Return on Investment) and identify
areas of improvement.
• Peter Drucker has said “what gets measured gets
managed”
• An effective marketing analytics practice tracks and
collects data across multiple marketing channels and
consolidates it into a single view.
 Marketing Analytics – Characteristics

• Marketing analytics is data driven

• Marketing analytics portrays customer insights
and trends
• Marketing analytics facilitates marketing
decision making
• Marketing analytics provides a picture of
marketing efforts of a company
• Marketing analytics requires experts to be done
properly
 Marketing Analytics –
Advantages
• Marketing analytics is important because:
 Understand customer and market trend is really
important I today’s times, marketing analytics gives
big picture trend by following every single detail
 Marketing analytics allows to easily depict the
programs / campaigns that worked / failed along
with the reasons.
 Marketing analytics allows monitoring of trends
over time
 Marketing Analytics –
Advantages
• It helps a marketer to understand its target
audience better
• It helps a marketer to understand where the
competitors are investing their efforts
• It helps a marketer to understand how well the
marketing campaigns are performing
• It helps the marketers to monitor current trends
and assess future trends
• It helps the marketer to use the data to decide
future course of action
 Marketing Analytics –
Disadvantages
• As such, marketing analytics offers very less
disadvantages, still some of the disadvantages that
it may offer are:
 Misinterpretation of market data: Collecting a lot of
data from the market is one things, interpreting the
data correctly is another
 Evaluating market growth without market share: An
analysis of the market size alone is not enough to
indicate your opportunities, market share must also
be evaluated correctly
 Marketing Analytics –
Disadvantages
 Market segmentation Vs Target markets: You must
identify the segments of the market that have
potential customers for your products or services.
Few businesses can afford to market to every single
potential customer. Identify a target market that you
choose from among the available segments, and go
after that target market in a focused manner.
 Misidentifying market needs: You may overestimate
how well your competition is meeting the customers'
needs and quit before you even try to market. You
also may misidentify the need that is being met.
 Sources of Market Data
Sources of Market Data
INTERNAL EXTERNAL

Internal Sources
Primary Sources Secondary Sources
(within the org.)

Accounting Periodicals and

Salesmen
Information Newspapers

Govt. Publications and

Sales Reports Dealers
Reports

Expenditure data of Published Market

Consumers
various types Surveys

Statistics of various Data released by

kinds International agencies
 The new realities of Marketing Decision
Making
• Marketing decision making relates to areas like –
sales and distribution, advertising and promotion,
supply chain management, customer service and
support etc
• Marketing decision making has undergone a drastic
change in last one decade, due to availability of new
and improved tools of data analysis
• Marketing decisions being made today are based on
data rather than intuition
• Experience is backed with factual insights in today’s
marketing decision making process.
 The new realities of Marketing Decision
Making
• Marketing decision making in modern times has not
remained simple anymore, it has become complex
where multiple factors play a role at once
• Modern marketing decision are made considering
the competitors and their actions
• Marketing decisions being made today more
consumer oriented, they aim to satisfy consumers
more and more
• Data has presumed utmost importance, analytics and
marketing research has become the foundation for
making sound decisions
 Market Sizing
• The "market size" is made up of the total number of potential
buyers of a product or service within a given market, and the
total revenue that these sales may generate. Market sizing is
the measurement of market potential in terms of customers
and revenue it can generate for a business
• It's important to calculate and understand market size for
several reasons – Market sizing can also help you to estimate
the number of people that you may need to hire before you
launch a new product or service, rather than "feeling your
way" as you test your new market. If you know this from the
start, you can optimize your approach to recruitment, so that
you have the right people in place when you need them.
 Market Sizing Methods
Top Down Approach
• There are two methods that are commonly used for
market sizing: top-down and bottom-up.
• Although the top-down method is simple, it's often
unreliable and overly optimistic. It looks at the
"relevant" market size for your product or service,
and then calculates how much your organization
might earn from it.
• For example, imagine that your organization markets
learning resources to schools. Your research shows
that there are 6,000 relevant schools in your country.
 Market Sizing Methods
Top Down Approach
• You know that the average sale per school is
around Rs 50,000, which means that your
market size is Rs 300 million.
• Of course, this is an incredibly optimistic and
unrealistic figure. Not every school needs your
products, and they're unlikely to purchase
• A top-down approach gives you inflated data,
and you often can't rely on it to make good
decisions.
 Market Sizing Methods
Bottom Up Approach
• The approach can be summarized in the
following steps:
Calculate Potential Sales

Use Market Research to assess interest in your product

Define You Target Market

 Market Sizing Methods
Bottom Up Approach – Example
• You've determined that 1,800 grocery stores might
invest in your software, which costs Rs 30,000. If 100
percent of these stores purchase the software, this is a
return of Rs 54 million.
• Your organization has already estimated that it will have
to invest at least Rs 7 million to develop, test, and
market the new software.
• This investment is only 13 percent of potential annual
revenues, so the risk is low, even if the response isn't as
positive as predicted. Your organization therefore
decides to move forward with the development of new
software.
 Department of MBA
IMS Engineering College, Ghaziabad
Marketing Analytics
Ashish Awasthi, Assistant Professor,
UNIT 2
Department of MBA

Pricing Analytics
Marketing Analytics: Data Driven Techniques with Microsoft Excel Wayne L
Winston, Wiley India Pvt Ltd.
 Pricing Analytics
Introduction
• Pricing impacts revenues and revenues impact
profitability
• Understanding how pricing impacts revenues &
profitability is one of the most important issues
faced by managers
• To do so, managers need to understand how
consumers’ willingness to purchase changes at
different price levels and how these changes
impacts profitability
• In short, managers need to understand the demand
curve
 Introduction to
Linear and Power Demand Curves
• As we know there are two types of demand curves
– Linear and Power (non linear) demand curves
• Linear demand curve is a straight line, power
demand curve is a downward sloping curve
• Linear demand curve has a equation q=a – b*p,
where a and b are constant, q is the quantity
demanded and p is the price
• Power demand curve has the equation q = a*pb
here – b is the price elasticity
• These are two most frequently used demand curves
 Introduction to
Linear and Power Demand Curves
• In case of linear demand curve, the Price elasticity of
demand, Ep is variable, while in case of power demand
curve it is constant
• To estimate a profit maximizing price, two things are
required to be known:
 The variable cost to produce a unit product = UC
 The demand curve of the product. Demand curve tells the
number of units of a product a consumer will demand at
each price.
 E.g. At price of Rs. p, the no. of units demand will be D(p)
 A firm’s demand curve is constantly changing and depends
of factors which are beyond firm’s control.
 Calculation of Profit
• When UC, p and D(p) are known;
Profit = (p – UC)*D(p)
• When you have an equation that gives the
quantity of the product demanded for each
price, you can find out the profit maximizing
price with the help of Microsoft Excel solver.
 Price Elasticity of Demand
• Given a demand curve, the price elasticity of
demand is the decrease in demand resulting from
1% increase in price.
• When elasticity is larger than 1, the demand is price
elastic. When elasticity is less than 1, the demand is
price inelastic.
• When demand is price elastic, a price cut will result
into increase in revenue
• When demand is price inelastic, a price cut will
decrease revenue
 Estimates of elasticity
• Studies by economists have found the following
estimates of elasticity:
 Salt: 0.1 (Highly inelastic)
 Coffee: 0.25 (inelastic)
 Legal fees: 0.4 (inelastic)
 TV sets: 1.2 (slightly elastic)
 Restaurant meals: 2.3 (elastic)
 Foreign travels: 4.0 (highly elastic)
• A 1% decrease in price of foreign travel can
increase the demand of foreign travel by 4%
 Estimating a
Linear Demand Curve
• Suppose that a product follows a linear demand curve
• Given the current price, the demand and price elasticity,
it is a simple matter to estimate the demand curve of the
product
• E.g. the current price is Rs. 100, Demand is 500 units
and EP = 2
• A 1% decrease in price will decrease demand by 2%
• Two points are needed on Demand curve one is (100,
500) other is (101, 490)
• Using this information, we will estimate the linear
demand curve
 Estimating a
Linear Demand Curve
• Demo link on estimation of Linear demand
curve:
https://www.youtube.com/watch?v=XI3MJ7bP
-CQ
• Demo link on estimation of power demand
curve:
https://www.youtube.com/watch?v=XI3MJ7bP
-CQ
 Complementary Products
• A complement refers to a complementary product or
service used in conjunction with another product or
service. Usually, the complementary product has little
to no value when consumed alone, but when
combined with another product or service, it adds to
the overall value of the offering
• The joint demand nature of complementary products
causes an interplay between the consumer need for
the second product as the price of the first product
fluctuates. In economics, this connection is called
negative cross-elasticity of demand
 Complementary Products
• For example, should the price of hot dogs
increase, it can cause a decrease in the demand
for hot dog buns. Since the cost of hot dogs
has an inverse relationship with the demand
for hot dog buns, they are considered
complementary products
 Pricing Bundling
• In a bundle pricing, companies sell a package or
set of goods or services for a lower price than
they would charge if the customer bought all of
them separately. Common examples include
option packages on new cars, value meals at
restaurants and cable TV channel plans.
• Pursuing a bundle pricing strategy allows you to
increase your profit by giving customers a
discount.
 Pricing Bundling
• Bundle pricing is built on the idea of consumer
surplus. Every customer has a price that he is
willing to pay for a particular good or service
• The difference between what the customer
pays and what the customer was willing to pay
is known in economics as the consumer
surplus. Bundle pricing is an attempt to
capture more of your customers’ consumer
surplus.
 Pricing Skimming and Sales
• Price skimming is a product pricing strategy
by which a firm charges the highest initial
price that customers will pay and then lowers
it over time. As the demand of the first
customers is satisfied and competition enters
the market, the firm lowers the price to attract
another, more price-sensitive segment of the
population
 Pricing Skimming and Sales
• The skimming strategy gets its name from
“skimming” successive layers of cream, or
customer segments, as prices are lowered over
time.
• Price skimming is often used when a new type
of product enters the market. The goal is to
gather as much revenue as possible while
consumer demand is high and competition has
not entered the market.
 Pricing Skimming and Sales
• Firms often use skimming to recover the cost of
development. Skimming is a useful strategy in the
following contexts:
• There are enough prospective customers willing to
buy the product at a high price.
• The high price does not attract competitors.
• Lowering the price would have only a minor effect
on increasing sales volume and reducing unit costs.
• The high price is interpreted as a sign of high
quality.
 Pricing Skimming and Sales
• Demo Link for price skimming:
https://www.youtube.com/watch?v=DK0IWT
mFXL0
 Department of MBA
IMS Engineering College, Ghaziabad
Marketing Analytics
Ashish Awasthi, Assistant Professor,
UNIT 4
Department of MBA

Retailing and Advertising Analytics

Marketing Analytics: Data Driven Techniques with Microsoft Excel Wayne L
Winston, Wiley India Pvt Ltd.
 Market Basket Analysis
• Retailers use scanners to create data that lists
the items purchased by each customer on a
given transaction. This data can be used to
obtain information that can be used for
increasing profits.
• E.g. Bloomingdale found that women who
purchased cosmetics also purchased hand
bags, so they placed cosmetics and hand bags
together in the retail store.
 Market Basket Analysis
• In market basket analysis we study how
identify pairs or sets of products that customer
tend to purchase together and how this
knowledge can help the retailer increase
profits
• Demo Link:
https://www.youtube.com/watch?v=BKER4dA
yqBk
 Market Basket Analysis
Computing ‘lift’ for two products
• In market basket analysis we study how identify pairs or
sets of products that customer tend to purchase together
and how this knowledge can help the retailer increase
profits
• Market basket – list of products purchased by the
consumer
• Market basket analysis – looks at association between the
products brought together.
• Lift – commonly used tool in the market basket analysis
to identify the combination of items that tend to be
purchased together
 Market Basket Analysis
Computing ‘lift’ for two products
• Formula for lift is given as under:

• Two way product lift involves combination of

2 products and can easily be computed in
excel
 Market Basket Analysis
Computing ‘lift’ for two products
The following data set can be used to compute 2 way lift, here
1,2,3,4,5,6,7 represent days of the week staring from Monday
 Market Basket Analysis
Computing ‘lift’ for two products
• For a superstore, the lift for meat and vegetable
would be equal

• Product combinations with lift greater than 1

have a tendency of being purchased together
• Thus, the retailer can place product
combinations with higher lifts near each other
Excel Numerical:
Page 447 – 449
 Using Lift to
Optimize Store Layout
• To maximize revenues, the store should adopt
a layout in which high lift items are placed
near each other.
• Given a lift matrix for different product
categories, a function called ‘evolutionary
solver’ can be used to maximize the total lift of
product categories
• On the basis of this information, optimum
store layout can be calculated Excel Numerical:
Page 454 – 456
 Allocating Resources
• Marketing managers must determine the profit
making allocation of scarce resources such as
advertizing money, shelf space in a store or
sales force to a territory.
• It is important to understand how a change in a
resource allocated to a product impacts its sales.
• This can be achieved through establishing the
relationship between the level scarce resource
and response (sales)
 Identifying sales to
marketing effort relationship
• E.g. a company must determine how to allocate an
advertising budget between its many magazines in
which ads are published in order to generate
maximum sales.
• How much shelf space be allocated to each product
category to increase profits?
• The relationship between the sales effort and
response can be modeled visually using three types
of curves: power curve, ADBUDG curve and
Gompertz curve
 The power curve
• The power curve has the equation y = a.xb
• The values a, b are constant which can be
found using excel trend curve
• The value of b is assumed to be between 0 and
1
 The ADBUDG curve
• The power curve has the equation
• This curve was developed by Prof. John Little on
MIT USA.
• It has a ‘S’ shape
• It shows that for small amount of marketing efforts,
little response is generated, for intermediate effort,
increasing returns are observed while for efforts
beyond a point, decreasing returns are observed
• https://www.youtube.com/watch?v=ugtMLICdDT4
Excel Numerical:
Page 486 – 489
 The Gompertz curve
• The power curve has the equation

• This curve is used to model change is profit

resulting from additional shelf space allocation
• It also has a ‘S’ shape
• It shows that for small amount of marketing efforts,
little response is generated, for intermediate effort,
increasing returns are observed while for efforts
beyond a point, decreasing returns are observed
• https://www.youtube.com/watch?v=Ewp5CF5ba_w
 Optimizing Allocation of
Sales efforts
• Suppose you know that a power curve can model
the response to sales efforts for four drugs, how
would you allocate sales effort between four drugs?
• Suppose sales effort = calls in thousands (‘000)
made, equation of the 4 power curves are:
• Drug 1, sales = 50(calls).5
• Drug 2, sales = 10(calls).75
• Drug 3, sales = 15(calls).6
• Drug 4, sales = 20(calls).3
Excel Numerical:
Page 490 – 491
Measuring effectiveness of Advertisements

• Companies have trouble measuring

effectiveness of ads, this is because of the
time lag between consumer exposure and
consumer response to an ad
• Effective allocation of ad budget is important
so that the money spent on ads does not go
waste
Measuring effectiveness of Advertisements

Ad budget can be spent in two ways:

• Pulsing: short spells of intensive ads followed
by no ads
• Continuous spending: advertising at all times
at a fairly constant rate
• https://www.youtube.com/watch?v=Y1w6ng5
73ks
 Models for Measuring
effectiveness of Advertisements
• 1) Adstock model:
• Adstock is an important component of marketing-mix
models. The term "adstock" was coined by Simon
Broadbent.
• Adstock is a model of how response to advertising builds
and decays in consumer markets. Advertising tries to expand
consumption in two ways; it both reminds and teaches.
• It assumes that the effect of an ad decays or wears away
with passage of time
• https://analyticsartist.wordpress.com/2013/11/02/calculating
-adstock-effect/
 Models for Measuring
effectiveness of Advertisements
• Advertising adstock is a term used to measure
the memory effect of advertising carried over
from start of advertising. For example, if a
company advertises at a certain level in week
1, week 2 will have a portion of week 1 level.
Week 3, in turn, will have a portion of week 2
level. In other words, adstock is a percentage
term that measures the decaying effect of
advertising throughout the weeks.
 Pay Per Click
Online Advertising
• PPC is an online advertising model in which
advertisers pay each time a user clicks on one of
their online ads.
• There are different types of PPC ads, but one of
the most common types is the paid search ad.
These ads appear when people search for things
online using a search engine like Google –
especially when they are performing commercial
searches, meaning that they're looking for
something to buy.
 Pay Per Click
Online Advertising
• In pay-per-click advertising, businesses
running ads are only charged when a user
actually clicks on their ad, hence the name
“pay-per-click.”
• Other forms of PPC advertising include
display advertising (typically, serving
banner ads) and remarketing.
• https://www.wordstream.com/pay-per-click-ad
vertising
 Department of MBA
IMS Engineering College, Ghaziabad
Marketing Analytics
Ashish Awasthi, Assistant Professor,
UNIT 5
Department of MBA

Forecasting and Co-joint Analysis

Marketing Analytics: Data Driven Techniques with Microsoft Excel Wayne L
Winston, Wiley India Pvt Ltd.
 Regression model to
forecast sales
• Often the marketing analysts need to determine how
variables are related, regression is a simple tool to
check that.
• The ‘trend line’ feature of MS excel helps in
showing relationship between a dependent and an
independent variable.
• The first step to determine how variables are related
is to graph the data points using scatter charts, with
independent variable on the x-axis and dependent
variable on the y-axis
Examples of relationships
Modeling a linear relationship
 Making a Regression Line
• Plot the scatter chart for the above data
• Add trend line that best fits the data points
• Note the equation of ‘line of best fit’ – also called
the least squares line or regression line
• Note the value of R2
• Regression line :
Daily bowl sales = -39.96*price + 787.7
• The slope -39.96 indicates that $1 increase in the
price of the bowl reduces the demand by 39.96
bowls
 Value of R 2

• R2 value is 0.721, which means that 72.1 %

variation in bowl sales is explained by the
regression line
• Higher value of R2 increases prediction
accuracy
• https://www.youtube.com/watch?v=Cltt47Ah3
Q4

Numerical Example:
WILEY page 163 – 166
 Ratio to moving average
Forecasting method
• Steps in ratio to moving average method:
 Estimate the de-seasonalized level of the series during
each period (using centered moving average)
 Fit a trend line to de-seasonalized estimates
 Determine the seasonal index for each quarter and
estimate the future level of the series by
‘extrapolation’
 Predict future sales by re-seasonalizing the trend line
estimates
Numerical Example:
 https://www.youtube.com/watch?v=mC1ARrtkObc
WILEY page 268 – 271
 Using ‘S’ curves for
Forecasting
• S-curves are great graphical project
management tools for planning, monitoring,
controlling, analyzing, and forecasting
project’s status, progress, & performance.
They show the progress of work over time and
form a historical record of project trends and
variations.
Using ‘S’ curves for
Forecasting
 Using ‘S’ curves for
Forecasting
• S-curves and s-curve patterns are a useful tool for
quickly analyzing systems, particularly when
looking at diffusion of trends and evolution of
innovations.
• They can heuristically identify solutions and
probabilities that would otherwise be quite time
consuming to figure out using something like a full
system or functional analysis.
• https://www.youtube.com/watch?v=qMXk6cJlUL
M
 Co Joint Analysis
• Conjoint analysis is the optimal market research
approach for measuring the value that consumers
place on features of a product or service.
• Subconsciously, one person might be more price-
sensitive while another is more feature-focused.
Understanding which elements consumers
consider being important, and which are trivial, is
the core purpose of conjoint analysis.
 Co Joint Analysis –
Importance
• Conjoint analysis is one of the most effective
models in extracting consumer preferences
during the purchasing process. This data is
then turned into a quantitative measurement
using statistical analysis.
• Analyzing the data gives you the ability to
peek into the minds of your target audience
and see what they value most in goods or
services and acts as a market simulator.
 Co Joint Analysis –
Types
• Two-Attribute Tradeoff Analysis
• Full-Profile Conjoint Analysis
• Adaptive Conjoint Analysis
• Choice-Based Conjoint Analysis
• Currently, Choice-Based Conjoint Analysis is the most
popular form of conjoint.
• Participants are shown a series of options and asked to
select the one they would be most likely to buy. Other
forms of conjoint include asking participants to rate or
rank products. Choosing a product to buy usually yields
more accurate results than ranking systems.
 Co Joint Analysis –
Advantages
• Useful in Market Segmentation
• Close Resemblance of Customer Decisions
• Measuring Price Sensitivity
• Ease of Calculating Attribute Interactions
• Enables Purchase Decisions
 Co Joint Analysis –
Disadvantages
• Complexity
• Difficult to Use
• Inability to quantitatively estimate Attitudes
• Over- or Undervaluation of Variables

• https://www.youtube.com/watch?v=aqsH4g0g
Yog