0% found this document useful (0 votes)
45 views123 pages

Marketing Analytics Notes

Uploaded by

Abc Def
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
45 views123 pages

Marketing Analytics Notes

Uploaded by

Abc Def
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 123

Marketing analytics notes

Data
Data is ubiquitous. It’s collected at every purchase made, flight taken, ad clicked, and social
media post liked—which means it’s never been more accessible to organizations.
Yet, access to data isn’t all it takes to set a business on the path to success; it also takes
employees who understand and know how to leverage data.

There’s now an increased demand for data-literate business professionals who can
handle, analyze, and interpret data to drive decision-making.

“In this world of big data, basic data literacy—the ability to analyze, interpret, and even
question data—is an increasingly valuable skill,”
With the right skills, data can allow you to gain and act on customer insights, predict future
financial and market trends, and enact systemic change for social good.
Marketing Analytics

• Your client has invited you to a meeting on 29 th of December, January 5th , Feb 2nd
• Is there a pattern here?
• Marketing Analytics is a discipline that uses Statistics and Machine Learning models and
visualizations to provide data-driven inference for marketing decisions.
• Marketing analytics can help increase return on investment (ROI) and create better customer
experiences.
• here are different types of marketing analytics tools and models that you can use to collect, analyze,
and interpret your marketing data. Some of the common tools are web analytics, email analytics,
social media analytics, mobile analytics, and lead generation analytics.
• Some of the common models are descriptive, predictive, and prescriptive analytics. Each of these
tools and models can help you answer different questions and achieve different goals.
• For example, if you want to know how many people visited your website, how long they stayed, and
what pages they viewed, you can use web analytics tools like Google Analytics.

3
How Data helps
Data science is the process of building, cleaning, and structuring
datasets to analyze and extract meaning.

Data analytics, on the other hand, refers to the process and


practice of analyzing data to answer questions, extract insights, and
identify trends.

If your dataset isn’t structured, cleaned, and wrangled, how will you be
able to draw accurate, insightful conclusions?
How data helps
• Gain customer insights: Data about your customers can reveal details about
their habits, demographics, preferences, and aspirations.

• Increase security: You can also use data science to increase your business’s
security and protect sensitive information. For example, machine-learning
algorithms can detect bank fraud faster and with greater accuracy than humans,
simply because of the sheer volume of data generated every day.

• Inform internal finances: Your organization’s financial team can utilize data
science to create reports, generate forecasts, and analyze financial trends. Data
on a company’s cash flows, assets, and debts is constantly gathered, which
financial analysts use to manually or algorithmically detect trends in financial
growth or decline.
How data helps
• Streamline manufacturing/Services : Manufacturing machines gather data
from production processes at high volumes. In cases where the volume of data
collected is too

• Predict future market trends: Collecting and analyzing data on a larger scale
can enable you to identify emerging trends in your market. By staying up to date
on the behaviors of your target market, you can make business decisions that
allow you to get ahead of the curve.
Segmentation
• Process of systematically grouping sets of consumers into groups based on similar
characteristics.
• Segments can be based on demography, Psychographic, geography, category or brand related
attitude and behaviors.
• The process of subdividing a market into distinct subsets of customers that behave in the same
way or have similar needs. Each subset may conceivably be chosen as a market target to be
reached with a different marketing strategy.
• A target market is a group of customers a business has decided to aim its marketing efforts
and ultimately its merchandise towards. A well-defined target market is the first element of a
marketing strategy.
• Uses in advertising :
• Help in defining Target groups.
• Consumers can be segmented by usage or income.
Creating target groups

• Demographics
• It is a way of describing audience based on measurable and quantifiable
factors such as
• -gender
• age
• income level
• marital status
• education
• race
• religion
Creating target groups

• Psychographic Segmentation
• Psychographic segmentation divides the target market based on socio-
economic class, personality, or lifestyle preferences.
- attitudes
- aspirations
- habits including media & product consumption
Creating target groups

• Socio-Economic Class ( SEC)


• Categorization of a household based on two criteria:
• Education of the Chief Wage Earner

Number of Consumer Durables (from a predefined list) owned by the
family. The list has 11 items, ranging from electricity connection to cars
and air conditioners; including agricultural land in rural areas.

(-MEDIA RESEARCH USERS COUNCIL AND BARC )


Durables
• Electricity Connection
• Ceiling Fan
• LPG Stove
• Two Wheeler
• Colour TV
• Refrigerator
• Washing Machine
• Personal Computer/ Laptop Car/
• Jeep/Van Air Conditioner
• Agricultural Land
The SEC grid
Audience measurement
• Audience measurement measures how many people are in an audience,
usually in relation to radio listenership and television viewership, but also in
relation to newspaper and magazine readership and web traffic on websites.
• Measurements are broken down by media market, which for the most part
corresponds to metropolitan areas, both large and small.
• IRS/TAM/RAM
• Diary method : researcher goes to the audience with a booklet and records
TV audience.
TG /ta

• FEM 15+ SEC ABC HSM.


• SEC A1B1MF 15 -25 METROS AND MINI METROS
• Formula in the Indian Context
• SEC (New Consumer Classification System – nccs)
• Gender
• Age group
• Markets/Geography
Audience Measurement on Digital
• 1. Website Traffic
• 2. Demographics
• 3. Audience Interests/Behaviours
• 4. Search Keywords
• 5. Digital Media Consumption
• 6. Clicks
• 7. Conversions
Digital/ web analytics

• Web analytics
• Techniques used to assess and improve the contribution of digital marketing to
a business reviewing traffic volume, clickstreams, online reach data, customer
satisfaction surveys, leads and sales

• https://www.smartinsights.com/

• https://www.digitalanalyticsassociation.org/
Key tools for digital advertising measurement
• https://marketingplatform.google.com/about/analytics/
• Google Analytics is the most-used tool for web analytics on the internet
• https://www.youtube.com/user/googleanalytics
• https://www.hubspot.com/
• -It's a one in all solution that provides plenty of feedback and support for analysis and evaluation.
• https://hootsuite.com/en-gb/
• - Manages multiple social platforms
• https://mixpanel.com/
• Mixpanel is a tool for mobile, allowing you to measure the ways people are using app or site, whether
they're coming back and using them again after their first visit
• https://www.megalytic.com/
• Megalytic lets you pull in data directly from a number of different sources such as Google Analytics or
Adwords , can be used to create integrated reports
• https://www.salesforce.com
• https://www.salesforce.com/products/einstein-analytics/overview/
• The platform features analytics tools called Einstien Analytics that can help you build up a clear picture of
how your marketing activity is impacting sales
Overview of Marketing Analytics
• 1.Marketing analytics can be termed as a discipline that uses appropriate statistical, machine learning models, and
visualizations to inform about key marketing decisions.

• 2. Marketing analytics possibilities start with availability, depth of data. Data can be from firm’s internal sources or
can be bought from market research.

• 3. Marketing analytics use case begins with exploratory data analysis (EDA) – this typically involves checking
data sanctity, useful variations and the applicable variance in data.

• Exploratory Data Analysis (EDA) is a process used by data scientists to analyze and investigate data sets and
summarize their main characteristics, often employing data visualization methods. It involves a variety of
techniques to:

• Discover patterns
• Spot anomalies
• Test hypotheses
• Check assumptions
• 4. Descriptive analysis involves checking variation in each variable using
mean, variance, and covariance (or correlations) – usually descriptive analysis is
done as part of EDA.

• 5. Correlation is a one of the primary techniques to examine how two variables


move together.

• 6. Predictive analytics is a discipline focused on estimating the values of a


metric for future. Machine learning technologies play a critical role in predictive
analytics.

• 7. Prescriptive analytics aims at suggesting a course of action and outcome


associated with each action (say pricing change or increase in promotion spends
in a channel)
Mean
• 1.Mean:
• The mean is the most used measure of central tendency.
• It represents the average of a given collection of data.
• The mean can be calculated for both continuous and discrete data.
• Mathematically, the mean is equal to the sum of all the values in the data set
divided by the total number of values.
• Data – 4,8,9,16,23
• - 12
Median
• Median:
• The median is the middle value of a data set when the values are arranged
in ascending order.
• It is the value that separates the lower half from the upper half of the data.
• 4,8,9,16,23
• Calculate median –
• 5+1 - 6
• 6/2 -3
• So median is 9
Mode
• Mode:
• The mode is the data value that appears most frequently in the data set.
• It represents the value with the highest frequency.
• A data set can have one mode (unimodal), multiple modes (multimodal), or
no mode (if all values occur with equal frequency).
• 4,8,9,16,23
Developing Marketing Information System (MKIS)

Marketing information is needed to take business decisions


Sample Business Decisions

 How to keep price constant when costs increase?


 How many quantities should be made available in soft drink?
 How many car variants should be made?
 How many phone plans a company should offer?
 What product ideas should be converted to prototype?
 What should be the next market to be explored?
 …and many more like these…
Sources of Marketing Information

Users of Marketing Information External Database

 Marketing Department People  Hiring MR Agency

 Management People  Government Reports

 Heads of other Departments  Commercial Databases


Internal Database  Info from channel members
 Competitors’ data
 Order history
 Sales and stock records
 Internal Research findings
Touchpoint and Profiling of the Customer

Idea of Touch Point


Every source (Touch Point) you tap for information becomes
source of your information to the seller Browser data
Customer Information for profiling the customer Social media accounts
Bank account Purchasing power
Credit Card Education
Email account Vehicle information
Apps used Professional
Institute data memberships
Company data
Customer Engagement

eMail
Mobile Messaging including whatsapp
Social Media – Facebook
Mobile Apps
Tele calling
Make customer ‘Subscribe’
Ask for referrals and reward for references
WHY IS MARKETING ANALYTICS IMPORTANT?
• 1. Improve the User Experience
• Collecting and analyzing your users’ first-party data can reveal how they feel about
their interactions with your product and website.
• Whether their feelings are explicitly stated (for instance, in a survey) or implicit in
their behaviors(for instance, leaving the website shortly after loading the page), having
this qualitative and quantitative information can allow any organization to make
changes that address their needs and increase the potential for leads to become
customers.

• 2. Calculate the Return on Investment of Marketing Efforts


• Another important function of marketing analytics is calculating monetary gain that
can be attributed to specific marketing channels or campaigns.
• calculate the return on investment
Calculating ROI

• for a specific marketing effort, use the following formula:


• ROI = (Net Profit / Cost of Investment) x 100
• Net profit is 10
• Investment is 100
• ROI will be ?
• 10/100x100
• = 10%
Importance of Marketing analytics

• 3. Plan Future Marketing Strategies


• With knowledge of your customers and the ability to track marketing efforts marketing
analytics provides an opportunity to create data-driven strategies for your organization.

• By analyzing marketing data, a company can discover what’s working, what hasn’t
worked, and how your customers feel about their experiences with your product and
website. You can also get a full picture of the impact that marketing efforts are having on
your company.

• With that information, you can plan. What should you do more of to reach your
quantitative goals? Which effort failed to generate new leads and should be dropped from
future plans? Data analytics helps you strategize and answer these kinds of questions.
Importance of Marketing analytics

• 4. Gain a better understanding of your customers.


• Data can provide actionable answers about your customer base, including who they
are, what actions they commonly take, what their pain points tend to be, and more.
Data can help you understand what improvements your team can make to improve
their experience.
• 5. Refine your marketing strategy.
• Data tells you what works and why, so you can refine your marketing strategy in real
time, replicating certain efforts because they're performing well and eliminating those
that are under-delivering.
• 6. Predict the success of future marketing campaigns.
• With predictive scoring based on past marketing campaigns, data can often predict
how customers will respond to future campaigns and overall advertising and
marketing efforts.
Popular marketing analytics software
• Hubspot: measures the performance of all marketing campaigns
• https://www.hubspot.com/products/marketing/analytics
• Sprout Social: manages, listens, and tracks social media engagement
• https://sproutsocial.com/
• SEMRush: tracks and measures content marketing efforts
• https://www.semrush.com/
• Brandwatch: finds trends, gathers consumer insights, and tracks marketing campaign
performance
• https://www.brandwatch.com/

• Salesforce: marketing campaign performance across all channels

• https://www.salesforce.com/products/marketing-intelligence/
Key tools for digital advertising measurement
• https://marketingplatform.google.com/about/analytics/
• Google Analytics is the most-used tool for web analytics on the internet
• https://www.youtube.com/user/googleanalytics
• https://www.hubspot.com/
• -It's a one in all solution that provides plenty of feedback and support for analysis and evaluation.
• https://hootsuite.com/en-gb/
• - Manages multiple social platforms
• https://mixpanel.com/
• Mixpanel is a tool for mobile, allowing you to measure the ways people are using app or site, whether
they're coming back and using them again after their first visit
• https://www.megalytic.com/
• Megalytic lets you pull in data directly from a number of different sources such as Google Analytics or
Adwords , can be used to create integrated reports
• https://www.salesforce.com
• https://www.salesforce.com/products/einstein-analytics/overview/
• The platform features analytics tools called Einstien Analytics that can help you build up a clear picture of
how your marketing activity is impacting sales
Marketing Analytics Process
Marketing research
MR is the function that links the consumer, customer, and public to the
marketer through information–information used to identify and define
marketing opportunities and problems; generate, refine, and evaluate marketing
actions; monitor marketing performance; and improve understanding of
marketing as a process.
Marketing research specifies the information required to address these issues,
designs the method for collecting information, manages and implements the
data collection process, analyzes the results, and communicates the findings and
their implications.”

• AMA.
Content analysis
• Research tool used to determine the presence of certain words or concepts within texts or
sets of texts and make inferences from it.
• Text can be defined from books, magazines, articles, speeches, conversations.
• It is a systematic, objective and quantitative analysis of advertising.
• Content is analyzed to understand the meaning it conveys to the audience.
• Uses of Content Analysis ;
• Reveal international differences in communication content
• Detect the existence of propaganda
• Identify intention focus or trends in an individual, group or institution
• Describe attitudinal and behavioral responses to communications
• Determine Psychological or emotional state of persons or groups
Methodologies of marketing research
• Content Analysis
• Perceptual Mapping
• Segmentation
• Experimental Design
• Concept and Benefit testing
• Copy Testing
• Audience Measurement
Perceptual mapping
• Perceptual mapping is used to compare products and brands based on the
consumer perceptions.
• The purpose of a perceptual map is to identify the images that consumers have
and their reactions to brands, products, services and other market offerings.
• USE : It graphically represents what consumers think about their brands and their
position relative to competition
• Main aspects/USES
• Pictorial representation of tabulated data
• Via correspondence analysis, biplot or multidimensional scaling
• Plots brands, products or companies against multiple attributes, even
demographics
• Useful for identifying gaps in a market and competitor strengths/weaknesses
• Can see the shift in perceptions brought about by a new entry into the market
• Has visual impact
Types of perceptual maps

• Simple perceptual - map using some key attributes in comparing specific


products and services . It is a standard perceptual map has two-axis and
visualizes collected data to communicate key findings.

• Multi dimensional scaling (MDS) – A variety of techniques for producing


perceptual maps of competitive products or brands e.g. A computer
manufacturer wants to see where his brand is positioned in relation to
competition.
Benefits of a perceptual map
• The firm accurately know the opinions of your customer base.
• understand how vital the selected attributes are for the customers.
• You have a good idea of what your competition is doing better than you.
• Is the target audience resonating with your brand’s identity?
• Identifying market gaps becomes easy, which help in launching new products
• Changes to marketing and sales strategies are made after findings from these
maps that lead to better campaigns and more conversions.
• With your customer’s perception of 2 different attributes on the map, you get
to see how the ecosystem of your industry is evolving and how to reposition
the company in it.
Uses of perceptual maps

• 1. New product development – By identifying gaps in the


industry
• 2. Evaluating marketing campaigns – basis share of voice and
other attributes
• 3. Develop brand identity – how consumers perceive our brand
over competition
• 4. Sales evaluation- comparison of sales of two brands
Concept and benefit testing
• concept Testing — How people, without prompting, interpret a deliberately sketchy idea for
a new product or service.
• Concept testing is most often used in concept development to test the success of a new
product idea before it is marketed. Concept analysis is often used as one step in the process
of providing “proof of concept.”(Smith and Albaum, 2010) example : a shampoo that lets
you be yourself; a detergent that removes stains but is gentle on fabrics.

• Concept tests provide the direction and guidance necessary to identify and communicate key
product and service benefits and uses, as well as product specific implementations such as
packaging, advertising, sales approaches, product information, distribution, and pricing.

• Concept tests identify the perceptions, wants, and needs of the product and service users and
decision makers. You can integrate each of these dimensions into a concept test survey.
• A variety of concept testing survey formats are available to implement. Each helps to
minimize risk and maximize revenue when introducing all types of goods and services to the
market.
Why concept testing matters
• Concept tests reshape and refine ideas so they have greater
potential for market acceptance. Specifically, concept tests:
• Indicate concentrated segments of the population to which the
product appeals.
• Assess the relative appeal of alternative product
ideas/configurations/positions. Features desirable to the targeted
market segments are highlighted.
• Provides necessary information for developing the product and its
promotion, distribution, and pricing.
Types of concept tests
• A variety of concept test approaches exist. Each has a different
objective and can provide a variety of benefits. Here are the
most important of these approaches.
• New Product Concept Tests
Identify the benefits most important to customers and the
features that are most likely to lead to the fulfillment of that
promise. Features can be categorized into “need to haves” and
“nice to haves.”
• Customer needs must be identified and prioritized for product
development and communicating to the market. This type of
concept test can also test customer initial reactions to the
concept (Reaction and Adoption Test ).
Types of concept tests
• Product Modification-upgrade Tests
Reformulations, modifications, and upgrades can add
new life to existing products and services. Here,
identifying the optimal bundle of features is a priority.
• Differentiating and timing the release of new features
that are “need to haves” vs. “nice to haves” is critical in
creating products and services that are truly “New and
Improved” and are “New Release” and “Upgrade”
worthy.
Various analysis techniques
• Statistical analyses of data using algorithms and data analysis tools
like BARC software or IRS software, wherein desired variables are
entered in the software to generate data.
• Statistical tools :
• 1. Multiple Regression techniques – to evaluate how a value of a
dependent variable changes with respect of any independent variable
e.g. impact of advertising on sales.
• 2. Discrimination analysis- classifying an object in two or more
categories e.g. a retail chain can classify stores as successful or
unsuccessful based on footfall data
• 3. Factor Analysis –Technique used to determine few underlying
dimensions of a larger set of interconnected variables e.g. a TV channel
can classify different types of TV shows based on broadly defined
genre.
• 4. cluster analysis – Techniques used to separate objects into a
specified number of mutually exclusive groups e.g. Metro and Mini
Metro cities.
• 5. Conjoint analysis – A technique to evaluate response to different
offers and inferring the overall utility. E.g.. An airline can determine
the total utility delivered by different combinations of passenger
services
• 6. Multi dimensional scaling – A variety of techniques for
producing perceptual maps of competitive products or brands e.g.
A computer manufacturer wants to see where his brand is
positioned in relation to competition.
Sampling
• Population-total number of people present in a group/market/ demographic
segment
• Sample-smaller representation of a large whole.
• Sampling- is a process of selecting a subset of randomised number of the
members of the population of a study.
• The principle of sampling is to remove bias as far as possible so that the
sample selected is as representative of the population as possible.
• The point to remember is that a lot of valuable social research is based on the
law of statistical regularity, which states that a small group chosen at random
from a large group will share much the same characteristics. This is an
important principle.
Samples and Censuses
• Census :
• ‘Census’ is the process of obtaining responses from/about each of the members of the
population.
• Sometimes it is necessary to contact everyone and not to sample. Most countries have
censuses of their entire populations. Many countries have a census every 10 years. The
main reason for a comprehensive count of everyone within a defined geographical area
is to record reliable information on a given date.
• Selecting a Representative Sample
• A representative sample is a small quantity of something that accurately reflects the
larger entity.
• sampling frame
• is a list of all the units of the population of interest. You can only apply your research
findings to the population defined by the sampling frame.
• Sample design / sample plan-
• is a definite plan for obtaining a sample from a given population.
• Sampling unit-is a geographical one (state,district)
• Sample size-number of items selected for the study
• Sampling Error-is the difference between population value and sample value.
• Sampling distribution-is the relative frequency distribution of samples.
Advantage of sampling
• Saves lot of time
• Provides accuracy
• Reduces cost
• Gives greater speed /helps to complete in stipulated time
• Convenient.
• Helps in wider coverage of population segments
Principles of sampling

• Two important principles


• 1.Principle of Statistical Regularity-random
• (sufficient representative of the sample),
• when a large number of items is selected at random from the universe, then it
is likely to possess the same characteristics as that of the entire population.
• 2.Principles of Large Numbers-(steadiness ,stability and consistency)
• This principle is based on the notion, that large numbers are more stable in
their characteristics than the small numbers.
A good sample
• The sample should be true representative of universe.
• No bias in selecting sample
• Quality of the sample should be consistent
• Regulating conditions should be same for all individual
• Sampling needs to be adequate
Key concepts of design
• Population The aggregate of all the elements, sharing some common set of
characteristics, that comprise the universe for the purpose of the marketing
research problem.
• Element An object that possesses the information sought by the researcher and
about which inferences are to be made.
• Sampling unit An element, or a unit containing the element, that is available for
selection at some stage of the sampling process.
• Sampling frame A representation of the elements of the target population that
consists of a list or set of directions for identifying the target population.
Probability and non-probability sample
• Sampling methods can be classified into probability sampling and non-probability sampling
methods.
• In ‘probability sampling’, each unit of the population has a probability of being selected as a
unit of the sample. This type of sampling is more rigorous and free from biases.
• In ‘non-probability sampling’, samples are selected based on the subjective judgement of the
researcher, rather than random selection (i.e., probabilistic methods), there may be instances
that certain units of the population will have zero probability of selection, because judgment,
biases and convenience of the interviewers are considered to be the criteria for the selection
of sample units of such sampling.
Methods of sampling
Simple random sampling & systematic sampling

• In this method of sampling, each unit of the population has the equal probability of
being selected as an unit of the sample.
• Thus if a frame has a population N = 100 then each member of the population has a
1/100 chance of being part of the sample = 1% or 0.1
• Systematic sampling
• This is a special kind of random sampling in which the selection of the first unit of the
sample from the population is based on randomization. The remaining units of the
sample are selected from the population at a fixed interval of the sample size
(represented by n)
• For example, if N = 800 and n = 40 then sample size I = N/n = 800/40 = 20 . Now we
can number the population from 1 to 800; select a random sample from 1-20; 20-40
and so on .
• The advantage of this sampling method is that, once the first unit of the sample is
randomly determined all the sampling units can be easily obtained.
Stratified sampling

• Stratified sampling is an improvised sampling over simple random sampling and systematic sampling.
This sampling will have more statistical efficiency. In this sampling method, the population is divided
into a specified set of strata such that the members within each stratum have similar attributes but the
members of different strata have dissimilar attributes. This means that each stratum is homogeneous
when compared to the population.
• If we have a population N = 200 of people living in Three different localities of Goregaon =50,
Borivali 20, and Virar =130 and we wish to target a sample size of 10% that is 20 how do we select
sample from each segment.?
• We divide sample size n(20)/population (200) = 0.1 or 10% , now when we multiply each sub
segment with 0.1 we can get the proportionate sample size.
• So Goregaon 50 x 0.1= 5
• Borivali 20 x 0.1 = 2
• Virar 130 x 0.1 = 13
• Total - 20
Cluster sampling
• Cluster sampling Cluster sampling is a sampling technique in which the population is
divided into different clusters such that the members within each cluster are dissimilar
(heterogeneous) in terms of their attributes, but different clusters are similar to each
other.
• This leads to the inference that each cluster can be treated as a small population which
possess all the attributes of the population. Hence, in cluster sampling, any one of the
clusters is randomly selected and all the units of that cluster are selected (sampled) to
arrive at inference about the population.
• For example segmenting population by cities, like Mumbai, Delhi
• The main difference between cluster sampling and stratified sampling lies with the
inclusion of the cluster or strata.
• In stratified random sampling, all the strata of the population is sampled while in
cluster sampling, the researcher only randomly selects a number of clusters from the
collection of clusters of the entire population.
Multi-stage sampling

• The multi-stage sampling employs more than one stage to sample the population
depending upon the reality. The combination of the types of sampling techniques to be
used in the specified number of stages is unique to the reality.
• Stage 1: In the first stage, the different states of the country are sampled from each
region using stratified sampling. The country can be divided into different regions
(strata), viz., east, west, north,South.
• Stage 2: After selecting some states from each region in stage-1 based on the
particulars of the stratified sampling, again one can use cluster sampling to identify a
district from each selected state by assuming different districts of each state as its
clusters.
• Stage 3: In each selected district, a random sampling may be used to select the
proportionate number of target groups (sampling units) from it.
Non-Probability sampling technique

• Non-probability sampling relies on the personal judgement of the researcher rather than
on chance to select sample elements.
• The researcher can arbitrarily or consciously decide which elements to include in the
sample.
• Non-probability samples may yield good estimates of the population characteristics, but
they do not allow for objective evaluation of the precision of the sample results.
• Because there is no way of determining the probability of selecting any particular
element for inclusion in the sample, the estimates obtained are not statistically
projectable to the population.
• Commonly used non-probability sampling techniques include convenience sampling,
judgemental sampling, quota sampling and snowball sampling.
Convenience sampling
• Convenience sampling A non-probability sampling technique that attempts to obtain a
sample of convenient elements. The selection of sampling units is left primarily to the
interviewer.
• Often, participants are selected because they happen to be in the right place at the right
time. Examples of convenience sampling include: (1) use of students, religious groups
and members of social organizations;
• Convenience sampling is the least expensive and least time-consuming of all sampling
techniques. The sampling units are accessible, easy to measure and cooperative. Despite
these advantages, this form of sampling has serious limitations.
• . Convenience samples are not representative of any definable population. Hence, it is
not theoretically meaningful to generalise any population from a convenience sample,
and convenience samples are not appropriate for marketing research projects involving
population inferences.
Judgmental sampling
• A form of convenience sampling in which the population elements are
purposely selected based on the judgement of the researcher.
• The researcher, exercising judgement or expertise, chooses the elements to be
included in the sample because it is believed that they are representative of the
population of interest, or are otherwise appropriate,
• For example a researcher may choose a test markets selected to determine the
potential of a new product on judgement basis convenience rather than
randomizing.
• Judgmental sampling is inexpensive, convenient and quick, but will have
challenges because the population is not defined explicitly.
Quota sampling
• A non-probability sampling technique that is a two-stage restricted judgmental sampling.
The first stage consists of developing control categories or quotas of population elements.
In the second stage, sample elements are selected based on convenience or judgement.
• Quota sampling may be viewed as two-stage restricted judgmental sampling that has
traditionally been associated with street interviewing. It is now used extensively, and with
much debate, in drawing samples from access panels.
• The first stage consists of developing control characteristics, or quotas, of population
elements such as age or gender. To develop these quotas, the researcher lists relevant
control characteristics and determines the distribution of these characteristics in the target
population, such as Males 48%, Females 52% (resulting in 480 men and 520 women
being selected in a sample of 1,000 participants).
• In the second stage, sample elements are selected based on convenience or judgement.
Once the quotas have been assigned, there is considerable freedom in selecting the
elements to be included in the sample. The only requirement is that the elements selected
fit the control characteristics.
Snowball sampling
• A non-probability sampling technique in which an initial group of participants is
selected randomly. Subsequent participants are selected based on the referrals or
information provided by the initial participants. By obtaining referrals from referrals,
this process may be carried out in waves.
• In snowball sampling, an initial group of participants is selected, sometimes on a
random basis but more typically targeted at a few individuals who are known to possess
the desired characteristics of the target population. After being interviewed, these
participants are asked to identify others who also belong to the target population of
interest. Subsequent participants are selected based on the referrals. By obtaining
referrals from referrals, this process may be carried out in waves, thus leading to a
snowballing effect.
• Even though probability sampling can be used to select the initial participants, the final
sample is a non-probability sample.
• Although this system is advantageous and cost effective but the researcher is dependent
on the judgement of the sample.
Experiments in Marketing
• Experiments are quite widely used in marketing. For example, experiments may be used by
advertisers in order to check the specific version of the advertisement that is likely to give
the best results.

• Say a marketer of a car brand wishes to decide on the version of the advertisement on TV
that will result in making people visit dealerships for undertaking test drive on a newly-
launched model of the car brand.

• To decide on the version of the advertisement to air on TV, the advertiser may develop four
versions of the advertisement – each of the four versions is a separate ‘treatment’.
• Each of these versions is shown to a set of potential consumers, and they are asked to
indicate whether they would undertake a test-drive of the newly-launched car.

• The four sets of responses obtained from each treatment are compared to determine which
version of the advertisement yields the highest rate of test-drive.
A/B testing
• A/B testing is a randomized experiment carried out in the online context.
• A/B testing is a way to compare two versions of something to figure out which performs
better.
• While it’s most often associated with websites and apps, but the method is almost 100
years old.
• How Does A/B Testing Work?
• The researcher divides the two sets of samples in group A and Group B. Now both sets are
shown two versions of the same attribute, like content to answer a basic question which
attribute is better?
• For example a website wants to decide how big should be the subscribe button?
• One set is shown a page which has a smaller size and another set is shown a different size.
The experiment can further be done on colour, dimensions and other attributes.
• The researcher must estimate the sample size you need to achieve a statistical significance,
which will help them get the right amount of data to take a decision
Benefits of A/B testing
• Multiple groups could be used to test multiple attributes for example how big
should be the buy button.

• What colour should it have Blue, Red? Which font should be use?

• A/B testing is a great way to


• 1. gain quick information about a specific question you have, particularly in an
online setting.
• 2. Run multiple tests simultaneously called Multivariate A/B testing.
• 3. Useful in digital content analysis
Process of A/B test
• Define the objective- what do we want the A/B test to achieve
• Create two versions – Create two versions of the same attribute
• Randomize – Randomly assign each version to a group of
participants
• Collect feedback – collect feedback from participants
• Analyse the results – Analyse the feedback
• Implement the change- implement the change
Regression Model
• A regression model is a quantitative forecasting method that explores
relationships between dependent and independent variables. It helps in
understanding how changes in one variable impact another.
• Common regression models include linear regression(2), multiple regression
(more than 2) logistic regression( probability) polynomial
regression( extends linear to unrelated factors)

73
CLV
• Customer lifetime value (CLV) is a prediction of the net value or the total
worth attributed to the entire future relationship with a customer.

• CLV is one of the key metrics that needs to be tracked as part of a customer
relationship management program.

• this metric is usually a measure of how valuable a customer is to the company.


• At any point of time, CLV is the sum of all previous contributions by a
customer adjusted for time value of money (i.e., compounded), as well as that
of all future anticipated contributions adjusted for time value of money (i.e.,
discounted).
CLV
• CLV is usually calculated for an individual customer.
• The CLV obtained for every customer is then aggregated in order to
obtain the total value obtained by the firm from all customers – this
aggregated total value is also referred to as Customer Equity.

• The individual-level calculations are discussed in this section. CLV is the


present-day value of all net margins earned from a relationship with a
customer,
CLV
• To effectively manage its customer relationships, a
company must be able to measure the value that each of
its customers is likely to bring to the company over the
course of that customer’s tenure with the company.

• Yet, although many companies measure customer


satisfaction, few measure individual customer
profitability.
Calculating CLV
•Scenario: SOBA coffee shop Now, let’s calculate the Customer Lifetime Value
(CLV):
•Imagine owning a cozy coffee shop called “ SOBA Coffee Shop.”
1. AOV =
You serve delicious coffee, pastries, and sandwiches. Your regular
customers love your warm atmosphere and friendly service. 2. Visits per year =
3. {Annual spending per customer} = A x B=
1. Average Order Value (AOV):
4. Multiply the annual spending by the average
o Your average customer spends around Rs 500 per visit.
customer lifetime:
This includes their coffee, pastry, or sandwich.
Result: The average lifetime value of a customer at is
2. Purchase Frequency Rate: approximately 78000
o Let’s say your loyal customers visit your coffee shop once
a week on average. That’s about 52 visits per year
(assuming they keep coming back consistently).
3. Average Customer Lifetime:
o Most of your customers stay loyal for a long time. On
average, they continue visiting your coffee shop for about
3 years.
Calculate CLV
• You are running a restaurant in juhu your average revenue per customer each
month is Rs 100
• Number of years that they are a customer of your restaurant 36 months
• Per customer the average advertising, promotion and other retention costs are
Rs 1000 for the 3 year period.
• Calculate CLV.
• CLV = 100x 36 -1000
TOOLS FOR EDA
• 1. Python and R: These programming languages enable analysts to analyze
and manipulate data using libraries and packages such as Plotly, Seaborn, or
Matplotlib.

• 2. S plus and R : These statistical programming languages come bundled


with a plethora of tools that help perform specific statistical functions like
classification and dimension reduction techniques.

• 3. BI Tools Business Intelligence tools like Qlik Sense, IBM Cognos, and
Tableau are also used for effective EDA.
What are R and Python
•R is a language and environment for statistical programming which includes statistical
computing and graphics. t was designed by Ross Ihaka and Robert Gentleman in 1993.
•R is mainly used for statistical analysis and representation. It has many features which
are useful for statistical analysis and representation. It has many easy-to-use packages
for performing tasks.

•Python, is a general-purpose, high-level programming language. It was created by
Guido van Rossum in 1991 and further developed by the Python Software Foundation.
•It was designed with an emphasis on code readability, and its syntax allows
programmers to express their concepts in fewer lines of code¹.

•Python is widely used in various software domains, including data science, web
development, and gaming.
•t is backed by a vast community of users and developers who ensure the smooth
growth and improvement of the language, as well as the continuous release of new
libraries designed for all kinds of purposes².
EDA VS DATA MINING
• Exploratory Data Analysis (EDA) is a technique used to understand the main
characteristics of a dataset, discover patterns, find outliers, and identify
relationships between variables. It’s typically used when there are no complete
expectations about the nature of those relations.
• Data Mining, on the other hand, is the process of extracting valuable
information from a large dataset. It uses sophisticated data search capabilities
and statistical algorithms to discover patterns and correlations in large
databases. The goal is to extract patterns and knowledge, not the data itself.
• In short, EDA is for understanding and exploring data, while data mining is for
extracting valuable insights and patterns from large datasets. Both are crucial
in the data analysis process.
EDA tools in advertising
•1.Google Analytics A platform that collects data from your websites and apps to create
reports that provide insights into your business.
•2. MixPanel An analytics tool that captures data on how users interact with your digital
product. It allows you to analyze this data with simple, interactive reports.
•3. Whatagraph An all-in-one intuitive marketing data platform that takes manual tasks
and hassle away from your marketing data management process.
•4. AdWords Performance Grader A tool that automates spending analysis and offers a
quick sanity check by comparing accounts to others in their “weight range” based on
monthly spending.
•5. Heap Analytics: A web analytics platform that automatically captures and organizes
every customer touchpoint to help in making faster
EDA tools in advertising
•6. Cyfe A business dashboard software that uses dashboards to monitor and analyze
data coming from websites, multiple departments, and other locations.

•7. Klipfolio: A cloud-based web app that helps you understand, visualize, and track
the KPIs and metrics that matter most. It allows you to gather, share, display, and
learn from your data in real-time

•8. Optimizely A digital experience platform that helps enterprises deliver engaging
web experiences and success through data-driven decisions, continuous
experimentation, and constant innovation
Xl for eda
Dataset Creation: You can manually input data or import it from external
sources.

Data Summarization: Excel provides functions to calculate descriptive


statistics for each variable.

Data Visualization: Excel’s charting capabilities allow you to visualize the


distribution of values.

Missing Values Identification: Excel can help identify and count missing
values in your dataset.
Marketing Analytics and KPI
• KPIs play a crucial role in your digital marketing strategy. They serve as quantifiable
measures for evaluating whether you meet your marketing objectives. By setting clear KPIs,
you can assess progress toward specific goals.

• (KPIs) are used to check you are on track. KPIs are specific metrics that are used to track
performance to make sure you are on track to meet specific objectives. They are sometimes
known as performance drivers or critical success factors (CSFs) for this reason.

• Key performance indicators (KPIs) Metrics used to assess the performance of a process
and/or whether goals set are achieved.

• The marketing funnel consists of three stages:


1. Awareness: Introducing potential customers to your brand or product.
2. Consideration: Making customers aware of your brand while they evaluate it against
alternatives.
3. Decision: Influencing consumers' purchasing decisions using information gathered during
the previous stages.
Marketing KPIs to Measure Success
1. Impressions: The number of times your ad or organic content is displayed
or viewed, regardless of whether it garners clicks. Tracking impressions helps
you understand how well you attract potential customers 1.
2. Website Traffic: Monitoring the volume of visitors to your website provides
insights into your brand's visibility and reach. You can measure this through
tools like Google Analytics.
3. Conversion Rate: Calculated as the percentage of website visitors who take a
desired action (such as making a purchase or filling out a form). A higher conversion
rate indicates effective marketing efforts
Reach x Frequency
4. Customer Engagement: Metrics like time spent on your website, pages viewed per
visit, and social media interactions reflect how well you connect with your target
audience. Engaged customers are more likely to convert.
5. Click-Through Rate (CTR): The percentage of people who click on your ad after
viewing it. A higher CTR suggests compelling ad content.
6. Cost Per Acquisition (CPA): The cost incurred to acquire a new customer. It helps
evaluate the efficiency of your marketing spend.
7. Return on Investment (ROI): Calculated by comparing the revenue generated from
marketing efforts to the cost of those efforts. A positive ROI indicates successful
campaigns.
Levels of key performance indicators (KPIs):
• • Business-level KPIs to measure contribution from media These KPIs
include contribution to revenue through direct sales attributed to social /
other media.
• Softer measures include reputation and customer satisfaction (CSAT).
• • Reach and influence KPIs to review reach, share-of-voice and sentiment.
These show the relative comparison of a brand’s reach.
• • Engagement KPIs to manage media.
• These are the easiest measures to collect, but the least valuable since they
don’t directly show contribution to business value.
• Although easy to collect, data on interaction with social sites is often
supplied separately by the owners of the different social presence and tools
for managing social interaction.
(KPIs) for performance review of an online presence include:
• • unique visitors: the number of separate, individual visitors to the site;
• • total numbers: of sessions or visits to the website;
• • repeat visits: average number of visits per individual;
• • duration: average length of time visitors spend on the site;
• • conversion rates: the percentage of visitors converting to subscribers
(or becoming customers);
• • attrition rates: through the online buying process;
• • churn rates: percentage of subscribers lapsing or unsubscribing.
• Objectives:
• 1 Acquisition objective Acquire 50,000 new online customers (based on current sales) this financial
year at an average cost-per acquisition (CPA) of Rs30 with an average profitability of 5%
• Substantiation (informed by situation analysis or insight or research)
• Based on growth forecast 40,000 sales per year, but with incremental sales arising from new
affiliate programme and SEO development
• Strategies to achieve goals
• Start affiliate marketing programme and improve SEO. Existing media mix based on pay-per-click
and display advertising supported by offline media Use social media engagement to generate leads
• Key performance indicators (critical success factors)
• Overall CPA for online sales Incremental number and % of sales from affiliate marketing
programme Number of strategic keywords ranked in top positions in natural search results page
• Number and % of existing customers registering to use online service Number and % of customers
actively using online services at different points after initially registering
• % of site visitors responding to merchandising/crossselling messages
What Is Market Basket Analysis?
• Market basket analysis pinpoints commodities that people usually buy together. It’s like observing which items
tend to keep each other company in a shopping cart.
• Retailers use this analytical approach to find associations between items purchased together. For instance, if
customers often buy bread and butter together, market basket analysis reveals this connection.
• The insights from market basket analysis have practical applications:
• Optimizing product placement on store shelves.
• Crafting effective cross-selling strategies.
• Designing targeted promotional offers
• Benefits:
• Increased Sales: By strategically placing related items, retailers boost sales.
• Enhanced Customer Satisfaction: Relevant product recommendations lead to happier customers.
• Informed Decision-Making: Insights guide decisions on product placements, deals, and inventory management.
• Drawbacks:
• Data Complexity: Handling large transaction datasets can be resource-intensive.
• Assumptions: Association rules assume independence, which may not always hold true.
Attribution modelling
• Attribution modelling is the influence of different referring media in the
‘path to purchase’ is assessed to see which influenced conversion to sale.
This compares to the simplistic approach of the last-click model.
• What Is Attribution Modeling?
• Attribution modeling allows marketers to analyze the customer journey and
attribute credit to various touchpoints that lead to conversions. It answers
questions like:

• Which marketing channels or interactions influenced a user’s decision to


convert?
• How did different touchpoints work together to drive a conversion?
Purpose of Attribution Modeling
• Identify Improvement Areas: By assigning credit to marketing channels and
touchpoints, businesses can identify areas in the buyer’s journey that need
improvement.

• Determine ROI: Attribution models help determine the return on investment


(ROI) for each channel or touchpoint.

• Optimize Marketing Budget: Marketers can allocate resources more


effectively by understanding which channels drive the most conversions.

• Tailor Campaigns: Attribution insights allow customization of marketing


campaigns and content based on unique personas.
Types of Attribution Modeling:
• Multi-Touch Attribution Modeling:
• Considers every channel and touchpoint a customer interacts with throughout the buyer’s
journey. Provides insights into the most influential channels and how they work together.
• Cross-Channel Attribution Modeling:
• Similar to multi-touch attribution but focuses on overall channel performance (paid,
organic, social media) without diving into specific touchpoints.
• Linear Attribution Modeling:
• Gives equal credit to all channels and touchpoints in the customer journey.
• First-Touch Attribution Modeling:
• Credits the first channel or touchpoint a lead interacts with.
• Last-Touch Attribution Modeling:
• Attributes all credit to the last touchpoint before conversion.
Imagine a customer journey:
• The user sees a Facebook ad (first touch).

• Later, they read a blog post on your website (middle touch).

• Finally, they click an email link and make a purchase (last touch).

• First-Touch Model: Facebook ad gets all the credit.


• Last-Touch Model: Email link gets all the credit.
• Linear Model: Each touchpoint gets equal credit (33% each).
• Multi-Touch Model: Considers all three touchpoints and their combined impact.
Why define TG on the first place ?
• Without being clear on TG creating relevant content is not
possible.
• Targeted communications is cost efficient.
• Targeted communications can be created.
• Creating word of mouth and viral marketing is easier.
• Creates USP.
• Gives competitive advantage.
What's a consumer persona?

• A fictional profile that represents a segment of the target audience


• Provides a visual and contextual representation of the target
audience's demographics, psychographics and media/technology usage
Caleb: The Thrill Seeker
• He's 24 years-old
• He graduated college two years ago with an English
degree
• He is currently the assistant manager at a ski & board
shop
• He makes $30,000 a year
• Whether it's jumping over cliffs into powder on a
snowboard, racing down treacherous trails on a
mountain bike, or hitting the half-pipe skateboard, he's
there because he's a daredevil. He welcomes new
experiences and loves taking risks with adventurous
activities like skydiving and windsurfing. He likes to
live a fast-paced existence and he's always on the go.
No matter how daring the challenge, he will always be
there with a Red Bull in hand, on high energy, and
ready to jump.
CREATE A BRAND PERSONA
• Personality Traits:
• . Consider adjectives like innovative, friendly, bold, or sophisticated. These traits will shape your target
consumers personality and its alignment with the brand’s character.
• Personality Colors:
• choose colors associated with your brand. Colors evoke emotions and play a significant role in brand
perception. Think about what each color represents (e.g., blue for trust, red for passion).
• Personality Mascot/Role Models:
• It could be a fictional character, a historical figure, or even an animal. What qualities does this mascot
possess which may appeal to your Target customer.
• Personality Tone:
• Is it conversational, formal, playful, or authoritative? What kind of conversations does the target consumer
have?
• Personality Voice:
• Imagine how your brand and the target consumer would converse with each other consider customer
inquiries, social media posts, or marketing campaigns.
Attribution; Weighting the Credit Display
Contributes to Conversion
• Attribution Modeling is the science of determining the value of each
customer touch point leading to a conversion. It helps you
understand the customer journey and justify your marketing spend.
The Benefits & Impact of Attribution
Modeling
Planning Attribution Modeling
• Start by identifying your marketing goals. Are you focused on branding and
awareness, lead generation, developing new business, or repeat business?
• Identify the channels that you want to track
• Map out the consumer conversion path. Develop a basic outline for your
customer journey, including path length, time to conversion, and the
relevant marketing channels. You can find this information in the Multi-
Channel Funnels.
• Determine how much ‘value’ to assign against each touch point. Define the
role and expected impact of each campaign element.
• Plan your next steps. If you learn that a certain campaign or source is
performing differently than expected, you will need to take action.
Standard Attribution Models on Google
Analytics
Two More Default Models
5 keys of attribution modeling success
•1. Understand Customers’ Purchase Journey
 Interview your client, customer support, and salespeople to gain insights into the target audience.
 Understand the decision-making process, desires, expectations, and objections.
 Explore the concept of ZMOT (Zero Moment of Truth), where consumers research products online before making a
purchase.
 Leverage data from Consumer Barometer with Google and conduct proper market research.
•2. Concept of Missing Touchpoints
 Recognize that not all interactions are tracked or visible.
 Some touchpoints (e.g., offline interactions, word-of-mouth) may not be captured in analytics tools.
 Consider indirect influences on the customer journey.
•3. Fix Data Integration Issues
 Ensure seamless data flow across different platforms and channels.
 Address discrepancies in tracking and reporting.
 Integrate data from various sources (e.g., CRM, web analytics, advertising platforms).
5 keys to attribution modelling success
•4. Not All Touchpoints Are Equally Valuable
 Understand that different touchpoints play varying roles in the customer journey.
 Some touchpoints (e.g., brand awareness, educational content) contribute indirectly.
 Prioritize touchpoints based on their impact on conversions.

•5. Use Data-Driven Attribution Models


 Consider machine learning models that allocate credit based on actual data patterns.
 Experiment with different attribution models to find the most accurate fit for your business.
Attribution Modeling;
Campaign Optimization Examples
• Reallocate Budget
Strengthen campaigns along the most profitable position in the purchase funnel.
• Revise CPA (cost-per-acquisition)
Better reflect the true contribution of your marketing activities to the whole consumer
journey.
• Reduce Time-to-Conversion
Look for opportunities to improve the efficiency of your conversion path and reduce
the number of paid clicks required to drive a purchase. For example, provide price
guarantees so customers don’t have to price shop, quick coupon codes, or more
detailed product information so they don’t have to look elsewhere.
• Reschedule campaigns
Change the timing of particular campaign types, such as email promotions.
.
Programmatic ad buying
• A major change in online advertising is through the use of new programmatic
ad buying techniques, known as Demand Side Platforms (DSPs), which use
an approach called real-time bidding (RTB). The purpose of these is to exploit
efficiencies through using technology to automatically bid on the most cost-
effective ad inventory in an auction.
• Since individuals are tracked across different sites they can be targeted
according to their interests shown by content view using a technique known
as behavioural targeting.
Programmatic Buying Benefits

Programmatic buying today provides an opportunity where the advertiser can not only attach
different value to each ad-impression based on 100+ parameters but also optimize for the media
buying on a real-time basis (RTB). The benefits of RTB vary based on audience segments as well
as the intelligence of the algorithm optimizing for the bidding and creative based on it but can be
huge when done right.
Display Lumascape(ecosystem).www.lumpapartners.com
Definition:
Ad Networks

Ad networks connect advertisers to publishers. They aggregate ad


inventory and offer it to advertisers.

Networks provide a way for media buyers to coordinate ad campaigns


across multiple sites (ranging from dozens to thousands) efficiently. Ad
networks vary in size and focus: large ad networks may require premium
brands and millions of impressions per month, while small ad networks
may accept unbranded sites with thousands of impressions per month.
SSP: Supply Side Platform
(For Publishers)

• SSPs allow publishers to jump into ad exchanges via DSPs to make their inventory
available and optimize selling of their online media space. Through SSPs, publishers
can gain the highest eCPM for their inventory rather than selling remnant space at
lower costs.
Ad Exchanges

Display space that’s unsold by either sites or networks is usually collected by an ad exchange, where it is
auctioned off to the highest bidder among advertisers, networks and agencies. It’s a very simple way to buy ad
space, and for publishers to squeeze value from their unused inventory. Exchanges let buyers purchase very
specific audiences, especially when using real-time bidding technology. Advertisers and agencies typically
use DSPs to buy display
DSP: Demand Side Platform

A demand side platform (DSP) is a system that allows digital advertisers to manage multiple ad
exchange and data exchange accounts through one interface. Real time bidding for display online
ads takes place within the ad exchanges, and by utilizing a DSP, marketers can manage their bids
for the banners and the pricing for the data that they are layering on to target their audiences.
DSPs, Exchanges and SSPs Together
• DSPs are used by marketers to buy ad impressions from exchanges as cheaply and as
efficiently as possible, SSPs are designed by publishers to do the opposite: to maximize
the prices their impressions sell at.

• SSPs allow publishers to connect their inventory to multiple ad exchanges, DSPs, and
networks at once. This in turn allows a huge range of potential buyers to purchase ad
space and for publishers to get the highest possible rates.

• When an SSP throws impressions into ad exchanges, DSPs analyze and purchase them
on behalf of marketers depending on certain attributes such as where they’re served,
and which specific users they’re being served to. By opening up impressions to as many
potential buyers as possible via real-time auctions, publishers can maximize the
revenues they receive for their inventory.

• This process takes place in milliseconds!, as a user’s computer loads a webpage.


DMP: Data Management Platform- 3rd Party Data Overlay for
More-Precise Targeting

• A data management platform (DMP) is a centralized data management platform that allows advertisers to create target
audiences based on a combination of in-depth first-party and third-party audience data. DMPs enable advertisers to
consolidate online and offline customer data from various sources into a single location, then use it to create demographic
and behavioral segments that can be used to target online advertising. Performance data from each campaign is then fed back
into the DMP, creating a feedback loop that improves optimization efforts and can be used for related reporting and analysis
• Companies use DMPs to collect and analyze huge amounts of data from many different sources. DMPs are now so powerful
that companies can track users and customers who visit from banners, Facebook pages, Tweets, mobile, video and even
offline applications. They collect and analyze data from cookies, small files that keep website settings and also record user
behavior. For example, DMPs can allow e-commerce sites, publishers and advertisers to find out how many users who
bought a big screen TV online also searched for high-end digital cameras in the past week .
DSPs and DMPs Together

• DMPs can be used to store and manage any form of information, but for
marketers, they’re most often used to manage cookie IDs and to
generate audience segments, which are subsequently used to target
specific users with online ads.
• Advertisers buy media across a huge range of different sites and through
various middlemen, including DSPs, ad networks and exchanges as you
read. DMPs tie all this activity together in one, centralized location and
use it to help optimize future media buys and ad creative.
• So in summary, a DMP is used to store and analyze data, while a DSP
is used to actually buy advertising based on that information.
RTB: Real Time Bidding
• Real-time bidding (RTB) is a
digital ad buying process that
allows advertisers to evaluate and
bid on individual impressions.
• Component of a DSP, ad exchange
or network, RTB lets buyers use
their own data and targeting
options to bid for each ad
impression.
Thanks to real-time bidding, ad buyers no longer need to work
directly with publishers or ad networks to negotiate ad prices and to • Advertisers can take factors such
traffic ads. Using exchanges and other ad tech, they can access a as site, placement, price, and user
huge range of inventory across a wide range of sites and cherry-pick data into account when bidding on
only the impressions they deem most valuable to them. That cuts each impression. The winning
down the number of impressions wasted on the wrong users.
bidder gets to serve the ad, which
is often customized on the fly to
better tailor the message to the
audience. The entire bidding
process for each impression takes
less than 25 milliseconds
How Bidding Works
• The actual bidding process which takes less than 100 milliseconds looks
like this:
• 1. The Exchange makes a call to the DSP with an available impression.
2. DSP checks to see if they want this impression – it could be someone in
their retargeting pool, or in a desired audience segment according to a third
party data vendor. If yes …
3. DSP makes a bid for it based on how much they think it’s worth or can
afford to pay
4. Exchange sells the impression to the highest bidder.
5. Ad is delivered by the winning bidder.
Goals of Online Display Advertising
• There are two primary goals of Display Advertising
• first, using display adverts as a form of marketing communication used to raise brand awareness; second, as
a direct response medium focused on generating a response.
• Additional Goals of Online Display Adverts are as under :
• Delivering content. This is the typical case where a click-through on a banner advertisement leads to a
destination site giving more detailed information on an offer. This is where a direct response is sought.
Today ads often embed videos or whitepapers to deliver content directly within the ad.
• ● Enabling transaction. If a click-through leads to a merchant such as a travel site or an online bookstore
this may lead directly to a sale. A direct response is also sought here.
• ● Shaping attitudes. An advertisement that is consistent with a company brand can help build brand
awareness.
• ● Soliciting response. An advertisement may be intended to identify new leads or as a start for two-way
communication. In these cases an interactive advertisement may encourage a user to type in an email
address or other information.
• ● Encouraging retention. The advertisement may be placed as a reminder about the company and its service
and may link through to on-site sales promotions such as a prize draw.
Measurement of display effectiveness
• Different terms are used for measuring banner ad effectiveness. Each time an advertisement is viewed is referred
to as an advertisement or ad impression – also known as page impressions (page views).
• Since some people may view the advertisement more than once, marketers are also interested in the reach, which
is the number of unique individuals who view the advertisement.
• This will naturally be a smaller figure than that for ad impressions.
• Cost of ads is typically based on CPM or cost-per-thousand (mille) ad impressions as with other media.
However, the popularity of CPC search advertising and CPA affiliate deals mean that these are options too.
• As with other digital media, direct response to ads is measured through click-through rate.
• Interaction rate (IR) is a form of measurement that is unique to display ads. It refers to the many ads which
encourage the site visitor to interact through a prompt to ‘rollover’ and another Flash creative will be loaded
which may offer a clear brand message rendered in large font, a response form such as an insurance quote or a
request to obtain a SIM or a game or poll.
• When a user clicks on the advertisement, he or she will normally be directed to further information, viewing of
which will result in a marketing outcome. Usually the user will be directed through to part of the corporate
website that will have been set up especially to deal with the response from the advertisement. When a user
clicks on an advertisement immediately this is known as a click-through, but adserving systems (using cookies)
also measure view-through, which indicates when a user views an ad and subsequently visits a website within a
defined period.
Programmatic advertising
• https://www.youtube.com/watch?v=cPVMGM0oEKQ
Programmatic Advertising with AWS
• cale and optimize real-time-bidding worldwide at low latency & low cost
• Scale bidding and auctions across the cloud.
• Then use machine learning to optimize bid decision-making, analyze
supply path, and improve ad serving with millisecond latency, agility and
scalability at low cost.
• https://aws.amazon.com/advertising-marketing/programmatic-advertising/

You might also like