Marketing Mix Modelling

(Digital and Advanced Analytics)

COURSE5 INTELLIGENCE

LIMITED

A Training Report

Submitted in partial fulfillment of the requirements for the award of

degree of

Masters in Business Administration

(Business Analytics)

LOVELY PROFESSIONAL

UNIVERSITY

PHAGWARA, PUNJAB

From 26/06/23 to present

SUBMITTED BY

Name of student: Diotima Bhattacharya

Registration Number: 12206767
Signature of the Student:
 To whom so ever it may concern
I, Diotima Bhattacharya, 12206767 hereby declare that the work done by me on
“Marketing Mix Modelling” from 26th June 2023 to present is a record of original work
for the partial fulfillment of the requirements for the award of the degree, Masters in
Business Administration (Business Analytics).

Diotima Bhattacharya

Signature of the student

Dated: 26th August 2023

INTERNSHIP OFFER LETTER
Confirmation from Reporting Manager
INTERNSHIP SPECIFICATIONS

Profile- Junior Analyst Intern

Department- Digital and Advanced Analytics (DAA)

Reporting Manager/ Director- Mr. Sandeep Kaushik

Location- Ocus Technopolis, Supremework, 2nd Floor, Tower-A, Golf Course Rd, Sector 54,
Gurugram, Haryana 122022

Type/Duration- One-Year Internship (26th June 2023- 24th June 2024)

Stipend- 20,000 INR per month

Payslip of stipend-
ABOUT THE COMPANY

COURSE5 INTELLIGENCE LIMITED

Website link- Course5 Intelligence | Transformative intelligence

In the realm of contemporary business, where data reigns supreme and innovation is the
linchpin of survival, a standout entity emerges, radiating a distinctive brilliance. This entity is
none other than a pure-play data analytics and insights enterprise, an unwavering beacon of
excellence dedicated to elevating organizations within the digital epoch. Delving into their
mission, methodologies, and the transformative footprint they leave upon the business
landscape reveals a narrative of immense significance. At its heart, this company is rooted in
the conviction that data constitutes the quintessence of the digital era. It serves as the catalyst
for growth, a theyllspring of strategic insights, and the bedrock upon which sagacious decisions
are founded. In a landscape flooded with data but often short on clarity, this company offers a
guiding star for organizations seeking to unleash the full potential harbored within their data
repositories.

Distinguishing themselves through a specialization in data analytics and insights, they stand as
an anomaly in the business milieu. Eschewing the one-size-fits-all model, they have
meticulously honed their expertise to fltheirish within the intricate tapestry of data. This fidelity
to specialization ensures that their clientele receives insights and solutions of the utmost
refinement. Central to their ethos is the notion of digital transformation. In an era where
swiftness governs the business ecosystem, maintaining an edge isn’t merely advantageous; it’s
indispensable. This company recognizes this reality and is unflaggingly committed to
equipping organizations with the tools to fltheirish within the digital realm.
To actualize this mission, they harness a trifecta of pioneering technologies: artificial
intelligence (AI), advanced analytics, and insights. These technologies form the bedrock of
their approach and serve as the cornerstones for catalyzing transformation. AI, endotheyd with
learning capabilities, facilitates process automation, predictive analytics, and sound decision-
making. Advanced analytics, meanwhile, offers the instruments to sift through expansive
datasets, unveil patterns, and distill precious insights. Insights, derived from a profound
comprehension of data and its contextual relevance, steer organizations toward decisions that
are both knowledgeable and impactful. They eschew cookie-cutter solutions in favor of
leveraging industry-specific domain proficiency. Be it healthcare, finance, retail, or any
other sector, their repository of knowledge and insights enables them to navigate the
labyrinthine complexities of diverse industries.

Augmenting their domain mastery, their solutions are guided by Intellectual Property (IP). This
denotes a repository of proprietary tools and methodologies, honed through years of
immersion. These IP-led solutions are a testament to their devotion to innovation and their
unwavering commitment to achieving tangible results. Their proficiencies span cloud
computing, data engineering, business intelligence, intelligent process automation, applied AI,
and consulting, allowing us to comprehensively address the entire spectrum of analytics and
insights. This covers the entire jtheirney from data administration and descriptive analytics to
perceptive, anticipatory, and directive analytics. Noteworthy strides in AI and machine learning
have furnished us with the tools to conceive technologies, catalysts, and adaptable frameworks.
These assets are dedicated to ensuring enduring value for their clientele through advanced
digital analytics, marketing analytics, and customer analytics solutions.

They hold the conviction that their adeptness in comprehending the omnichannel customer
experience, digital business paradigms, and the intricacies of their clients' operations,
intertwined with their proficiency spanning analytical domains, deployment of AI technologies,
and consultative protheyss, empotheyrs us to forge a robust competitive edge on behalf of their
clients. By seamlessly integrating their solutions into their clients' crucial business workflows,
their objective is to engender operational efficiency, foster innovation, and consistently
enhance value. This jtheirney propels their clients towards an exciting trajectory, aligned with
what they term the fifth dimension - the future.
APPROACH OF THE COMPANY:

The approach of the company

include:

• Synthesize data from

various stheirces using a
comprehensive 360-
degree approach
• Combine behavioral,
attitudinal, and syndicated
data
• Integrate diverse data types: digital (web, mobile, social), telemetry, and IoT
• Establish connections across multiple channels and touchpoints
• Provide solution toolkits and frameworks tailored to specific business inquiries
• Possess in-depth industry and domain expertise
• Utilize the Digital Suite and ResearchAI Suite to expedite solutions
• Apply cutting-edge AI and next-gen technologies for cognitive automation and
enhanced knowledge discovery
• Emphasize actionable insights over conventional reports
• Deliver swift outcomes and facilitate seamless adoption of digital business models
within client organizations

MAJOR INDUSTRIAL FOCUS:

• Consumer Goods and Services-

To thrive amid continuous innovation, market shifts and increase direct to consumer
offerings.

• Life Sciences-
Pharmaceutical analysis and insights thrugh market research.

• Research and Consulting Firms-

Rapidly deliver robust insights to business stakeholders.
 • Retail-
Embracing customer data to form enduring relationships and provide relationship
driven experiences.

• Technology-
Businesses leverage disruptive innovations in AI, cloud, edge, computing, and mobility.

• Telecom and Media- Customer analysis and insights are key to provde deliverable
experiences.

SOLUTIONS PROVIDED:

1. Digital Analytics-
In the contemporary landscape,
consumers are engaging with
brands through digital platforms at
an unprecedented rate, and this
trend is set to persist. Digital
Analytics serves as the conduit for
brands to customize experiences,
amplify revenue, and streamline
expenses. It facilitates the
evaluation and enhancement of interactions within digital channels and customer
journeys. Moreover, it involves an in-depth analysis of Search and Social Media
platforms. The deployment and refinement of your Customer Data Platform (CDP) are
seamlessly integrated. The potential of data is harnessed to propel the expansion of your
digital enterprise through Course5's Digital Analytics.
2. Customer Analytics-
In the realm of customer engagement, a universal approach proves insufficient. The key
to retaining and incentivizing top-tier customers
hinges upon a profound understanding of their
preferences and a foundation of trust. By segmenting
and customizing content and experiences, you
communicate your credibility and reliability. The
acquisition of customers should be rooted in
predictive profitability models. Meanwhile,
sustaining their loyalty is achieved through tailored
lift analysis, concentrating on the retention of high-value clients.

3. Consumer and Market insights-

In an environment that is both
intricate and rapidly evolving,
organizations must possess
thorough and substantiated insights
pertaining to consumers, target
markets, and competitors to ensure
their prosperity. However, the task
of collecting essential data and
swiftly generating these insights is
hindered by the proliferation of data sources and disruptive shifts in the market
panorama. To surmount this challenge, a revolutionary approach founded on the
amalgamation of diverse contemporary data sources and harnessing digital
technologies, advanced analytics, and AI accelerators has emerged. This transformative
strategy is instrumental in fostering potent new capabilities for generating insights.
4. Marketing Analytics-
In an environment that is both
intricate and rapidly evolving,
organizations must possess
thorough and substantiated insights
pertaining to consumers, target
markets, and competitors to ensure
their prosperity. However, the task
of collecting essential data and
swiftly generating these insights is
hindered by the proliferation of data
sources and disruptive shifts in the market panorama. To surmount this challenge, a
revolutionary approach founded on the amalgamation of diverse contemporary data
sources and harnessing digital technologies, advanced analytics, and AI accelerators
has emerged. This transformative strategy is instrumental in fostering potent new
capabilities for generating insights.

5. Industry 5.0 and supply chain Analytics-

NPI Buy Make Move Sell Returns

Supply Chain Visibility / Control Tower

Supply Chain / Inventory Op�miza�on

Customer Analy�cs
Predic�ve Maintenance
3D Prin�ng for
Prototyping Automated Demand Sensing
Energy Management Warehouses

Remote Field Services Management

Tes�ng New Product Data-Driven Contract

Performance Management Dynamic Inventory Track & Trace

Digital Machine Customer Service Process

Engineering Returns/Exchanges
Inspec�on

Digital Twin Telemetry Digital Dealerships

Op�mized Parts Planning Connected Worker

Produc�on Automa�c Transport Precision /

Spend Analy�cs Op�miza�on Planning & Scheduling Dynamic Pricing

Product Quality

Edge Analy�cs Shopper Insight

Connected Products

Enablers Visibility Dashboards Analy tics / Insights Ci

Connected Dev ices Wearables Prototyping / Simulation

Industry 5.0 ushers in a new era where technology-proficient human workers take
center stage in operations. It revolves around empowered individuals driving intelligent
 factories, decentralized production, sophisticated supply chains, and personalized
customer experiences. This framework thrives on a digital and artificial intelligence
(AI)-driven environment, fostering cognitive automation and real-time interconnected
intelligence throughout the organization.

At the heart of Industry 5.0's foundation, the Industrial Internet of Things (IIoT) is
powered by heightened collaboration between humans and intelligent systems that
generate substantial data. This wealth of data can be harnessed through advanced data
analytics and AI-driven insights, coupled with the inventive capabilities and analytical
acumen of human minds. This integration taps into cohesive, instantaneous business
intelligence, resulting in enhancements across revenue, profit margins, efficiency, and
customer satisfaction.

6. Entreprise AI-
In the contemporary landscape,
Artificial Intelligence (AI) has
transcended from a novelty to an
essential prerequisite for enterprises
striving not only to endure but to
excel. Yet, the dynamic evolution of
technology, the abundance of data
resources, and the expanding
expectations of employees, stakeholders, and consumers present challenges in the
endeavor to expand AI initiatives. To navigate this complexity and successfully
implement AI on a broader scale, enterprises must possess the appropriate skill set,
accelerators, domain knowledge, and a strategic design thinking approach. This
convergence equips them to effectively deploy AI initiatives across their operations.
AI PLATFORMS DEVELOPED BY THE COMPANY

Course5 Compete -

Course5 Compete stands as an expansive market and competitive intelligence system

designed to propel brand development and strategic advancement. By enhancing the
understanding of markets, competitors, and customers, this platform facilitates business
growth. Comprising three robust components - Digital Shelf, Brand Experience, and
Market & Competition Sensing - the platform harnesses AI capabilities. This empowers it
to provide swiftly actionable competitive insights, empowering enterprise-wide teams to
center their efforts on executing strategies and adapting to change.

Course5 Discovery-

A solution infused with AI capabilities, Augmented Analytics offers pertinent, actionable,

and user-friendly insights through diverse consumption methods and user profiles. The
ultimate goal is to foster a culture centered around insights, encouraging data-driven
decision-making.
Incivus-

Incivus represents an AI-driven platform for Creative Effectiveness. It meticulously

assesses your advertisements on a frame-by-frame basis, offering accurate, impartial, and
detailed insights into ad recall, attention, emotions, and other factors. This process aids in
identifying your most impactful creative prior to its launch.

By removing uncertainty from creative decision-making, Incivus empowers forward-

thinking marketing teams to launch campaigns with assurance and certainty.
DEPARTMENT

Digital and Advanced Analytics

Here we harness the power of cutting-edge technologies and data-driven insights to transform
businesses and drive strategic growth. In a rapidly evolving digital landscape, we stand as the
vanguard of innovation, propelling organizations towards data-driven excellence.

Our Mission:

Our mission is to unlock the full potential of digital technologies and advanced analytics to
empower informed decision-making, enhance customer experiences, and achieve operational
efficiency. We are committed to helping our clients navigate the complexities of the digital age
and leverage data as a strategic asset.

Key Focus Areas:

Digital Transformation:

We spearhead the digital transformation journey for our clients, utilizing the latest
technologies to streamline processes, improve customer engagement, and drive revenue
growth. Our expertise spans a wide range of digital solutions, including web and mobile
applications, e-commerce platforms, and interactive customer experiences.

Advanced Analytics:

With an emphasis on data-driven insights, we employ advanced analytical techniques to

extract meaningful patterns and trends from vast datasets. By deciphering complex
information, we enable our clients to make well-informed decisions that lead to competitive
advantages.

Machine Learning and AI:

Our department is at the forefront of harnessing the capabilities of machine learning and
artificial intelligence. We develop predictive models, recommendation systems, and
automation tools that optimize operations, enhance personalization, and unlock hidden
opportunities.
 Customer Experience Enhancement:

Understanding that customer satisfaction is paramount, we work on enhancing customer

journeys across digital touchpoints. By analyzing customer behavior, sentiment, and
preferences, we craft tailored experiences that resonate with the audience.

Data Strategy and Governance:

We establish robust data strategies that ensure data integrity, security, and compliance.
Through meticulous data governance practices, we enable our clients to derive accurate insights
while maintaining regulatory standards.

Our Approach:

Our approach is rooted in collaboration, innovation, and expertise. We collaborate closely with
cross-functional teams to align digital initiatives with overarching business goals. Our
innovative spirit drives us to explore emerging technologies and adapt them to practical
applications. Moreover, our deep domain knowledge ensures that our solutions are tailored to
the unique challenges of each industry.

Impact and Outcomes:

In the Digital and Advanced Analytics Department, our relentless pursuit of excellence
translates into tangible impact and transformative outcomes for the organizations we partner
with. Here's a deeper look at the impact we create and the outcomes we drive:

Optimized Operations:

By infusing data-driven insights into operational processes, we help streamline workflows,

reduce inefficiencies, and eliminate bottlenecks. This leads to enhanced productivity, cost
savings, and improved resource allocation.

Accelerated Growth:

Our digital solutions and advanced analytics empower businesses to identify untapped
growth opportunities. Through predictive models, market trend analysis, and customer
segmentation, we provide actionable insights that guide strategic decisions for expansion and
revenue enhancement.

Enhanced Customer Satisfaction:

With a keen focus on customer experience, we leverage data to personalize interactions and
anticipate customer needs. This results in higher customer satisfaction, increased loyalty, and
positive brand perception.

Informed Decision-Making:

Our advanced analytics capabilities enable leaders and decision-makers to base their choices
on accurate, data-backed insights. This minimizes guesswork, reduces risks, and enhances the
precision of strategic planning.

Innovative Product Development:

By uncovering consumer preferences, market gaps, and emerging trends, we contribute to

the creation of innovative products and services that resonate with target audiences. Our
insights guide the R&D process, leading to products that align with market demand.

Data-Driven Culture:

We foster a culture where data takes center stage. Through our solutions, organizations learn
to value and utilize data as a strategic asset. This cultural shift permeates every level of the
organization, from executives to front-line staff.

Personalized Marketing Campaigns:

Our department empowers marketing teams to create highly targeted and effective
campaigns. By analyzing customer behavior and preferences, we ensure that marketing
messages resonate with the right audience, resulting in improved conversion rates.

Continuous Improvement:

We don't stop at implementation. Through ongoing monitoring, analysis, and optimization,

we ensure that our solutions evolve to meet changing market dynamics and business needs.
This commitment to continuous improvement drives sustained success.
 Competitive Edge:

By leveraging AI, machine learning, and digital strategies, we help organizations gain a
competitive advantage. They stay ahead of the curve by being agile, adaptable, and responsive
to market shifts.

Measurable Results:

Our outcomes are measurable and quantifiable. Whether it's increased website engagement,
higher conversion rates, improved customer retention, or reduced operational costs, our impact
is evident in key performance metrics.
LAYOUT OF THE WORK/TRAINING IN THE COMPANY

What is Marketing Mix?

The 7 Ps of the marketing mix,

also known as the extended
marketing mix, provide a
comprehensive framework for
businesses to plan and execute
their marketing strategies. In
addition to the traditional 4 Ps
(Product, Price, Place, and
Promotion), the extended mix
includes three additional elements
that address the service-based
aspects of marketing. Here's a
detailed breakdown of the 7 Ps:

1. Product-

- This refers to the tangible or intangible offering that a business provides to meet the needs
or desires of its target market.

- Key considerations include product features, quality, design, branding, packaging, and
differentiation from competitors.

2. Price:

- Pricing involves determining the monetary value customers are willing to pay for a product
or service.

- Factors to consider include cost of production, competitive pricing, perceived value, pricing
strategies (e.g., cost-plus, value-based), and discounts.
3. Place (Distribution):

- Place refers to the channels and methods through which products or services are made
available to customers.

- Distribution decisions include selecting distribution channels (direct, indirect), managing

inventory, determining locations, and ensuring efficient delivery.

4. Promotion:

- Promotion involves the communication strategies used to create awareness, interest, and
desire for the product or service.

- Elements include advertising, public relations, sales promotions, direct marketing, social
media marketing, and other promotional activities.

5. People:

- In service-oriented businesses, people play a critical role in delivering the service

experience.

- This includes interactions between customers and employees, employee training, customer
service quality, and the overall expertise and attitude of staff members.

6. Process:

- The process refers to the procedures, systems, and mechanisms used to deliver the product
or service.

- It encompasses the customer journey, service delivery workflow, efficiency, and ease of
doing business with the company.

7. Physical Evidence:

- Physical evidence pertains to the tangible elements that contribute to the overall customer
experience and service quality.

- Examples include physical facilities, signage, packaging, branding, and any other tangible
cues that influence customer perceptions.
These 7 Ps collectively form a holistic framework that addresses various dimensions of
marketing strategy, ensuring that all aspects of the customer experience are considered and
optimized. By focusing on each element, businesses can tailor their marketing strategies to
align with their target audience's needs, preferences, and expectations, ultimately leading to
more effective and successful marketing campaigns.

Factors affecting brand sales-

Factors impacting brand sales and their measures include-

Incremental factors-

Above the Line (ATL) or pull factor

a. TV- TRP (Television Rating Point), GRP (Gross rating Point)

b. Print (Magazine, newspaper)- Readership, Spends. Eg- an ad on the front page has
higher spend than an ad on the fourth page.
Magnitude of spend is directly proportional to impact of sales.
c. OOH- Out of home advertising counts footfalls, traffic, reach and number of billboards,
Spends
d. Radio- Listenership, measures in Spends

Below the Line (BTL) or push factor

a. Consumer promotions-
1. TPR- temporary price reduction, percentage price change. It is considered when price
is reduced over 5%.
2. Bonus packs/bundle packs- promoted volume (how much volume is on promotion).
3. Feature- Printed material, can either be present or not- yes/no
4. Display- temporary secondary location
b. Trade promotions-
Can be measured in spends and schemes.
c. Sampling- number of samples distributed.
Non-controllable Factors-

GDP

Inflation

Temperature

Seasonality

Festivities

Government policies

Controllable long-term Factors (Base)-

1. Distribution- number of stores covered

It is of two types- Numeric distribution, Weighted distribution

2. Pricing- week on week price trend, average price

3. Competition- Pricing, distribution, promoted volume (promotion), number of items in
store etc.
MARKETING MIX MODELLING (MMM)
MMM takes total sales volume across time and quantifies changes in sales due to individual
executions. It is a statistical analysis technique used by companies to evaluate the effectiveness
and return on investment (ROI) of their marketing activities. MMM seeks to quantify the
impact of different marketing variables on sales.

MMM analyses historical data, typically using regression analysis, to determine the
relationship between marketing inputs (e.g., advertising spend, promotional activities, pricing)
and outputs (e.g., sales, market share, profitability).

Questions that can be answered using MMM are-

a) How effective is each marketing channel in driving sales or other desired

outcomes?
MMM can help identify the most impactful marketing channels, allowing businesses to
optimize their marketing budget allocation accordingly.

b) What is the optimal level of investment in each marketing activity?

MMM can provide insights into the appropriate allocation of resources across various
marketing tactics, such as advertising, promotions, and pricing strategies.

c) How do changes in marketing variables affect sales or other performance metrics?

MMM can help quantify the impact of modifications in marketing inputs, such as increasing
advertising spend or adjusting pricing, on business outcomes, enabling companies to make
data-driven decisions.

Base and Incremental Sales-

In the context of Marketing Mix Modeling (MMM), "base sales" and "incremental sales" are
terms used to describe different components of the total sales for a product or service. These
terms are often used when analyzing the impact of marketing activities on overall sales.

• Base Sales:

Base sales, also known as organic sales or natural sales, refer to the level of sales that would
occur in the absence of any marketing efforts or external factors. In other words, base sales
represent the sales that are driven by factors beyond the scope of the analyzed marketing inputs.
These factors could include the inherent demand for the product, seasonality, economic
conditions, and other variables.

Base sales are an essential concept in Marketing Mix Modeling because they provide a
benchmark against which the impact of marketing activities can be assessed. By separating
base sales from the total sales, marketers can identify the additional sales generated by their
marketing efforts.

• Incremental Sales:

Incremental sales, also referred to as lift or additional sales, are the sales that can be attributed
directly to the impact of specific marketing activities. These sales are above and beyond the
base level of sales and are a result of the marketing inputs being analyzed.

In MMM, the goal is to determine how much of the observed sales increase is due to the
marketing efforts and how much is part of the natural sales trend. By quantifying incremental
sales, marketers gain insights into the effectiveness of their marketing strategies and the return
on investment (ROI) of their marketing activities.

Suppose a company's product had a base sales level of 1,000 units per month before any
marketing efforts were initiated. After launching a marketing campaign, the company observed
total sales of 1,500 units per month.

In this scenario:

- Base Sales: 1,000 units

- Incremental Sales: 1,500 units - 1,000 units = 500 units

The incremental sales of 500 units are attributed to the impact of the marketing campaign,
while the base sales of 1,000 units represent the underlying demand regardless of marketing
efforts.
In a business landscape defined by rapid digital transformation and increasing competition, the
art of effective marketing has evolved into a science. Marketing Mix Modeling (MMM) stands
as a powerful analytical methodology that illuminates the intricate interplay between various
marketing inputs and their corresponding impact on business outcomes. By dissecting the
relationships between marketing activities and performance metrics, MMM empowers
organizations to optimize their strategies, allocate resources judiciously, and amplify their
marketing effectiveness.

Understanding the Core Elements of Marketing Mix Modeling:

At its essence, MMM revolves around the meticulous examination of the marketing mix - a
concoction of marketing elements that collectively shape a brand's strategy. These elements
include product pricing, advertising expenditures, promotional activities, distribution channels,
and external variables like economic conditions and competitive actions. The primary objective
of MMM is to decode the impact of each of these elements on business outcomes such as sales,
revenue, market share, and profitability.

Data Integration and Analysis:

Central to the process is the integration of data from diverse sources, including sales data,
marketing expenditures, customer demographics, market trends, and more. This
comprehensive dataset serves as the bedrock for statistical analysis. Employing advanced
statistical techniques such as regression analysis, time-series analysis, and machine learning
algorithms, MMM unearths the patterns, correlations, and causal relationships between
marketing inputs and outcomes. It isolates the individual contributions of each element,
revealing their influence on the overall business performance.

Attribution and Effectiveness Measurement:

One of the key outputs of MMM is the attribution of revenue or sales fluctuations to specific
marketing inputs. This enables businesses to discern which marketing activities are driving the
most significant impact on the desired metrics. For instance, MMM might reveal that a specific
advertising campaign has a substantial effect on sales growth, while changes in pricing
strategies might have a more nuanced impact. This information forms the foundation for
informed decision-making, guiding resource allocation towards the most impactful strategies.
Forecasting and Optimization:

Beyond retrospective analysis, MMM holds the power of foresight. Leveraging historical data
and observed relationships, the methodology forecasts the potential outcomes of various
marketing strategies. Organizations can simulate scenarios to understand the implications of
different approaches before implementation. This empowers decision-makers to make strategic
choices that align with their goals and mitigate potential risks.

Benefits and Applications of Marketing Mix Modeling:

 Resource Allocation:

Perhaps one of the most significant benefits of MMM lies in its ability to optimize resource
allocation. By quantifying the impact of each marketing element, businesses can make
informed decisions about how to distribute their marketing budgets effectively. This ensures
that resources are directed towards activities that yield the highest return on investment,
resulting in a more efficient use of funds.

 Strategic Decision-Making:

Marketing Mix Modeling offers a data-driven compass for strategic decision-making.

Organizations can determine optimal pricing strategies, allocate budgets to specific advertising
channels, and tailor promotional campaigns to resonate with target audiences. Such decisions
are no longer based on intuition but grounded in empirical evidence.

 Performance Evaluation:

MMM enables real-time performance evaluation of ongoing marketing campaigns. Businesses

can measure the effectiveness of their strategies as they unfold and make adjustments based on
data-driven insights. This agility enhances the potential for campaign success and minimizes
the risk of costly missteps.

 Scenario Analysis:

Organizations can employ MMM to conduct scenario analyses. By simulating different

scenarios, they can anticipate potential outcomes and assess the risks and benefits associated
with various strategies. This pre-emptive analysis empowers businesses to make proactive
decisions that align with their objectives.

 Market Insights:
MMM not only empowers organizations to make better marketing decisions but also provides
insights into market trends and dynamics. Through the lens of data, businesses can uncover
shifts in consumer behavior, identify emerging market trends, and adjust their strategies
accordingly.
Challenges and Considerations

While Marketing Mix Modeling offers a plethora of benefits, its implementation is not devoid
of challenges. These challenges include data quality and availability, attribution complexity,
model selection, and accounting for external factors. The accuracy of MMM hinges on the
availability of accurate and comprehensive data. Inconsistent or incomplete data can skew
results and compromise the reliability of insights. Additionally, attributing changes in business
outcomes solely to marketing inputs might be intricate due to the influence of external
variables. The selection of appropriate statistical models is crucial, as improper model selection
can lead to biased or inaccurate results.

Future Prospects:

As organizations increasingly recognize the importance of data-driven decision-making, the

significance of Marketing Mix Modeling is poised to grow. The proliferation of digital
channels, the advent of Big Data, and advancements in analytical tools will further enhance the
accuracy and scope of MMM. The integration of machine learning and AI techniques will
enable more sophisticated analysis and prediction capabilities. Moreover, the evolution of
MMM will extend its applications beyond traditional marketing, into areas such as supply chain
management, operations optimization, and risk assessment.
MMM Project Life Cycle-
The project life cycle of Marketing Mix Modeling (MMM) typically consists of several
stages, from the initial kick-off and data collection to the final implementation and
measurement of the model's impact.

1. Kick Off and Data Collection:

In this stage, the project team establishes goals, objectives, and timelines for the MMM
project. Data requirements are identified, and data collection methodologies are determined.
The team collaborates with data providers, ensuring the necessary data is collected and
prepared for analysis.

2. Data Review:

During the data review stage, the collected data is examined and assessed for quality,
accuracy, and completeness. Data validation and cleaning processes take place to ensure that
the data is suitable for analysis. Any missing or inconsistent data is addressed and corrected.

3. Modeling and Analysis:

In this stage, statistical modeling techniques are applied to the data to build the MMM
model. The model incorporates the relationships between marketing inputs (e.g.,
advertising, promotions) and the dependent variable (e.g., sales). The model is refined, and
various statistical analyses, such as regression analysis, are conducted to evaluate the
significance and impact of the marketing variables.

4. Developing Recommendations:

Based on the insights generated from the MMM analysis, recommendations are developed.
These recommendations may include optimal marketing strategies, budget allocation across
marketing channels, and suggested changes to marketing tactics. The recommendations
align with the business goals and objectives identified at the project kick-off stage.

5. Client Implementation:

After the recommendations are finalized, they are shared with the client and convince them
to utilize those techniques. The client's marketing teams or stakeholders implement the
recommended strategies and tactics based on their feasibility, resources, and priorities.
Implementation may involve adjusting marketing budgets, optimizing media plans, refining
promotional activities, or other actions based on the MMM findings.

6. Date of Impact:

The date of impact refers to the point in time when the implemented changes based on the
MMM recommendations start influencing the business outcomes. It is the period from which
the impact of the MMM model is measured. It could be several weeks or months after the
client implementation stage, depending on the lag effects and the nature of the business
being analyzed.

Throughout the project life cycle, continuous monitoring, evaluation, and feedback loops
are essential to assess the performance of the MMM model, validate its effectiveness, and
adjustments as needed. This iterative approach ensures that the MMM insights and
recommendations remain aligned with the evolving business environment and objectives.
Outputs of Marketing Mix Modelling-

1. Sales Contribution:

Sales contribution refers to the percentage contribution of each marketing input or factor to the
total sales. It helps businesses understand which specific marketing activities are driving the
most significant portion of their overall sales. Sales contribution analysis allows for the
allocation of credit to different marketing inputs based on their proportional impact.

For example, if the analysis reveals that a particular advertising campaign contributes to 30%
of total sales, this information helps in evaluating the effectiveness of that campaign in driving
revenue.

2. Sales Due To:

Sales due to specific marketing inputs

indicate the amount of sales that can be
attributed to each individual marketing input.
It quantifies the direct impact of each input on
the sales outcome. Sales due to analysis is
important for understanding the specific
return on investment (ROI) of different
marketing activities.
For instance, if the analysis shows that a specific advertising channel led to $500,000 in sales,
it provides a clear understanding of the direct revenue generated from that channel.

3. Return on Investment (ROI):

ROI is a critical metric in MMM that measures the efficiency of each marketing input in
generating revenue compared to the cost of that input. It helps marketers determine the financial
impact of their marketing efforts and allocate resources to strategies that yield the highest ROI.

These outputs are crucial in evaluating the effectiveness of marketing strategies, optimizing
resource allocation, and making data-driven decisions. They allow businesses to identify high-
impact marketing activities, understand the relationships between inputs and outcomes, and
determine which activities deliver the best return on investment.
REGRESSION-
Regression analysis is a
statistical technique used to
examine and quantify the
relationships between one
or more independent
variables and a dependent
variable. It is employed to
understand how changes in
the independent variables
are associated with changes
in the dependent variable.
In other words, regression
analysis helps us predict or estimate the value of the dependent variable based on the values of
one or more independent variables.

Key Concepts in Regression:

1. Dependent Variable (Response Variable): This is the variable that you are trying to predict
or explain. It's affected by changes in the independent variables.

2. Independent Variables (Predictor Variables): These are the variables that you believe have
an impact on the dependent variable. The goal is to determine how these variables influence
the dependent variable.

3. Regression Line (or Hyperplane): It's the best-fitting line or plane that represents the
relationship between the independent and dependent variables. In simple linear regression, it's
a straight line; in multiple linear regression, it's a hyperplane in higher-dimensional spaces.

4. Coefficients: These are the parameters of the regression equation that quantify the
relationship between the independent variables and the dependent variable. They indicate the
change in the dependent variable for a unit change in the independent variable, all else being
equal.

5. Residuals: Residuals are the differences between the observed values and the predicted
values from the regression model. Minimizing the sum of squared residuals is the objective of
regression analysis.
Statistical Regression Equation

The Marketing Mix Modeling (MMM) regression equation involves estimating the relationship
between the dependent variable (Y) and the independent variables (X) to analyze the impact of
marketing inputs on business outcomes. While the specific variables and their weights may
vary depending on the context, here is a general form of the MMM regression equation:

Y = β0 + β1X1 + β2X2 + β3X3 + ... + βnXn + ɛ

Where:

Y represents the dependent variable, typically sales volume or revenue.

X1, X2, X3, ..., Xn represent the independent variables, such as marketing inputs (e.g.,
advertising spend, promotions, pricing).

β0 represents the intercept or constant term, which captures the baseline level of the dependent
variable when all independent variables are zero.

β1, β2, β3, ..., βn represent the coefficients or weights associated with each independent variable,
indicating the magnitude and direction of their impact on the dependent variable.

ɛ represents the error term, accounting for unexplained or random factors that affect the
dependent variable but are not captured by the independent variables.

Key Performance metrics

1. R-Square (Coefficient of Determination):

R-Square measures the proportion of the dependent variable's (Y) variation that can be
explained by the independent variables (X) in a regression model. It indicates the goodness-of-
fit of the model.

- Acceptable Range: R-Square ranges between 0 and 1. Higher values indicate a better fit,
with 1 representing a perfect fit. Acceptable ranges can vary depending on the context, but a
commonly accepted range is 0.6 to 0.9.

2. Adjusted R-Square:
Adjusted R-Square adjusts the R-Square value by considering the number of independent
variables and the sample size. It penalizes excessive inclusion of variables to prevent
overfitting.

- Acceptable Range: Similar to R-Square, higher values indicate a better fit. Acceptable ranges
vary depending on the context, but a commonly accepted range is similar to R-Square, around
0.6 to 0.9.

3. Durbin-Watson:

Durbin-Watson is a measure of autocorrelation in the residuals of a regression model. It

assesses the presence of serial correlation, which indicates whether there is a relationship
between the residuals at different time points.

- Acceptable Range: Durbin-Watson ranges between 0 and 4. A value around 2 indicates no

significant autocorrelation. Acceptable ranges vary, but values between 1.5 and 2.5 are often
considered acceptable, with values closer to 2 indicating no significant autocorrelation.

4. MAPE (Mean Absolute Percentage Error):

MAPE measures the average percentage difference between predicted and actual values. It
quantifies the accuracy of a forecasting model.

- Acceptable Range: Acceptable ranges depend on the specific industry and context. Lower
values indicate better accuracy. It is essential to compare MAPE values to the average error
rates in the industry or domain to assess acceptability.

5. VIF (Variance Inflation Factor):

VIF measures the multicollinearity among independent variables in a regression model. It

quantifies how much the variance of the estimated coefficients is inflated due to collinearity.

- Acceptable Range: VIF values below 5 or 10 are generally considered acceptable, with lower
values indicating less multicollinearity.
6. T-Stat (T-Statistic):

T-Stat measures the statistical significance of an independent variable's coefficient in a

regression model. It assesses whether the coefficient is significantly different from zero.

- Acceptable Range: T-Stat values greater than 1.6 (in absolute value) are often considered
statistically significant. However, acceptable ranges can vary based on the level of significance
desired (e.g., 95% or 99%).

7. P-Value:

P-Value represents the probability of obtaining a result as extreme as, or more extreme than,
the observed result if the null hypothesis is true. It indicates the statistical significance of a
coefficient in a regression model.

- Acceptable Range: P-Value values less than a predefined significance level (e.g., 0.05) are
often considered statistically significant. However, acceptable ranges can vary based on the
level of significance desired.
REGRESSION IN MMM-

The dependent variable (Sales Volume)-

The Y variable i.e. dependent variable is taken as sales volume and not sales unit or sales value
as sales volume is the true and continuous measure of sales change.

Sales volume, measured in terms of the quantity or number of products sold, provides a simpler
and more standardized measure compared to sales value or sales units. It eliminates the need
to consider variations in product prices or packaging configurations, which can introduce
additional complexity in the analysis. It allows businesses to evaluate the effectiveness of
marketing strategies and tactics without being influenced by variations in product pricing or
package sizes.

The independent Variables-

Independent variables are the factors or inputs that are believed to influence the dependent
variable, which is typically a business outcome like sales, revenue, or market share.
Independent variables represent the various marketing elements that a business can control or
influence as part of its marketing strategy.

Advertising Spend:

This includes expenditures on advertising campaigns across various channels such as TV, radio,
digital, print, and outdoor. The analysis aims to understand how different advertising spends
affect business outcomes.

Pricing:

The price of a product or service can significantly impact consumer behavior and sales. MMM
analyzes the relationship between pricing strategies and sales volume or revenue.

Promotions:

Promotional activities such as discounts, coupons, special offers, and loyalty programs are
taken into account to assess their impact on sales and customer behavior.

Distribution Channels:

The choice of distribution channels (e.g., online, retail, wholesale) can influence product
availability and consumer reach, which can affect sales.
Seasonality and Time Factors:

Time-related variables like day of the week, month, season, and holidays are considered to
capture the influence of temporal patterns on sales.

Economic Indicators:

Macroeconomic factors such as GDP growth, consumer confidence, and unemployment rates
can impact consumer spending patterns and sales.

Competitive Activity:

Information about competitors' marketing activities, pricing strategies, and promotions can be
used as independent variables to understand how they affect a business's sales.

Geographic Variables:

Variables related to geography, such as region, urban/rural classification, or store location, can
be included to analyze regional differences in sales performance.
OLS Regression
OLS is a type of linear regression that aims to find the best-fitting line (or hyperplane in the
case of multiple variables) that minimizes the sum of squared differences between the observed
and predicted values. It assumes a linear relationship between the independent variables and
the dependent variable.

Method for Performing OLS Regression in MMM:

1. Data Collection and Preparation:

Gather historical data on marketing inputs and business outcomes. Ensure the data is clean,
complete, and appropriately formatted.

2. Choose Variables:

Select the independent variables (marketing inputs) and the dependent variable (business
outcome) you want to analyze. For this example, let's consider advertising spend and pricing
as independent variables and sales as the dependent variable.

3. Build the Model:

Construct the OLS regression equation:

4. Estimate Coefficients:

5. Interpret Coefficients:
Interpret the coefficients to understand their impact on the dependent variable:

6. Validate the Model:

Assess the model's goodness of fit using metrics like R-squared. A higher R-squared indicates
a better fit of the model to the data.

7. Make Predictions and Analysis:

Use the model for prediction and scenario analysis. For example, predict the sales for a specific
combination of advertising spend and pricing.

8. Business Insights:

Extract insights from the model to guide marketing decisions. For instance, you can determine
the optimal advertising spend and pricing strategy that maximizes sales.

9. Continuous Improvement:

Update and refine the model as new data becomes available. This ensures that the model
remains accurate and relevant in changing market conditions.

Benefits of OLS Regression in MMM:

- OLS provides interpretable coefficients that quantify the impact of each marketing input.

- It's a well-established method that is easy to understand and implement.

- OLS can handle multiple independent variables and is suitable for analyzing the effects of
various marketing inputs simultaneously.
PROJECT PERFORMED
MMM of Sazerac Fireball Whiskey
"Sazerac Fireball Whiskey" refers to a popular brand of
cinnamon-flavored whiskey produced by the Sazerac
Company.

The company is headquartered in New Orleans, Louisiana, and

has a rich history dating back to the 19th century. Sazerac is
known for its commitment to quality, craftsmanship, and
preserving traditional recipes while also embracing innovation
in the spirits industry.

Problem statement- What all media variables impacted the

sales of the Fireball brand and what changes should be done in
the investments to increase sales in the next financial year.

Steps included/ Tools used-

1. Data collection from various sources.

2. MS-Excel:
 Defining the dependent variable (sales volume) and figuring out the several
independent variables.
 Calculating the adstock (retention) and lag of media variables.

3. Python:
 Preprocessing and cleaning the data.
 Feature Engineering- creating new variables from existing variables,
transforming existing variables etc.
 Checking Correlation of the independent variables with the dependent variable
through correlation matrix and heatmap.
 Running OLS Regression iterations and create the model.
 Calculate Variation Inflation Factor (VIF).
 Check the R-square, Adjusted R-square and other metrics.
  Checking the signs of the variables, constant should be positive and price
should be negative, rest should be positive if stated otherwise.
 Calculate Mean Absolute Percentage Error (MAPE) which should be less
than 5.
 For a good model, Adjusted R-square>80 and MAPE<3, VIF<5 for each
variable and correct directionality.
 Calculate elasticity, contribution, ROI, due-to, and predicted sales.
 Exporting the results to an Excel file.
 Create Actual v/s Predicted Sales line chart and analyse.

4. MS- PowerPoint:
 Compiling the results from python and creating a presentation describing the
gist and providing the analysis to the higher stakeholders and clients.

A sample correlation matrix looks like-

A sample Actual v/s Predicted Sales chart looks like-
SKILLS ACQUIRED
1. Data Integration and Cleaning:

- Combining data from different sources requires skills in data integration and cleaning to
ensure data consistency and accuracy.

2. Data Transformation:

- Transforming variables for adstock, lags, and creating new features enhanced my ability to
prepare data for analysis effectively.

3. Feature Engineering:

- Creating meaningful variables from existing ones requires a creative mindset and the ability
to extract valuable insights from raw data.

4. Exploratory Data Analysis (EDA):

- EDA skills help uncover trends, outliers, and relationships in the data, allowing for more
informed decisions during modelling.

5. Data Visualization:

- Creating informative charts and graphs aids in understanding complex relationships and
presenting insights clearly.

6. Statistical Assumptions:

- Understanding assumptions of regression analysis, like homoscedasticity and normality, and

addressing violations contributes to accurate model results.

7. Model Evaluation:

- The ability to assess model performance using various metrics like Adjusted R-squared,
MAPE, and VIF demonstrates your analytical skills.

8. Interpretation of Results:

- Translating complex statistical outputs into actionable insights for non-technical stakeholders
showcases your communication skills.
9. Business Acumen:

- Recognizing which variables are most relevant to the business problem and framing the
analysis in a business context is a valuable skill.

10. Problem Solving:

- Addressing challenges that arise during data preprocessing, model building, or interpretation
demonstrates my analytical and critical thinking abilities.

11. Attention to Detail:

- Precise data collection, cleaning, and model implementation require meticulous attention to
detail to avoid errors.

12. Time Management:

- Successfully completing each phase of the project within deadlines shows my ability to
manage time effectively.

13. Stakeholder Management:

- Communicating findings and recommendations to clients or stakeholders involves

interpersonal skills and understanding their needs.

14. Continuous Learning:

- The project likely exposed you to new tools and techniques, fostering a willingness to learn
and adapt to evolving analytical methods.

15. Ethical Considerations:

- Ensuring the ethical use of data, especially if customer data is involved, demonstrates a
commitment to responsible data analysis.

16. Real-World Application:

- Applying analytical skills to a practical business scenario hones my ability to bridge the gap
between theory and real-world applications.

17. Presentation Skills:

Ability to convey findings in a clear, structured, and engaging manner during presentations.
18. Creativity:

Thinking creatively to generate unique approaches to problem-solving and analysis.

19. Attention to User Experience (UX):

Considering how insights will be used by stakeholders and tailoring your analysis accordingly.

20. Learning Agility:

Willingness and ability to learn new tools, techniques, and industry trends to stay relevant.
CONCLUSION
Learning Marketing Mix Modelling (MMM) as an ongoing intern in the Department of
Digital and Advanced Analytics at Course5 Intelligence has been an invaluable experience
that has enriched my analytical skillset, deepened my understanding of data-driven decision-
making, and provided practical insights into the world of marketing analytics. Throughout this
internship, I have gained a holistic perspective on how businesses use MMM to optimize their
marketing strategies and drive growth.

Engaging in this internship has enabled me to develop a comprehensive understanding of the

multifaceted process involved in MMM. From data collection and preprocessing to model
building, evaluation, and interpretation, I have been exposed to each crucial step of the analysis.
This hands-on experience has not only solidified my technical skills in Python and data analysis
but has also honed my ability to approach complex problems systematically.

Working in the Department of Digital and Advanced Analytics at Course5 Intelligence has
provided me with the opportunity to work alongside experienced professionals who are experts
in the field. Moreover, I have had the chance to witness the significance of effective
communication, as conveying complex analytical findings to non-technical stakeholders
requires a balanced blend of technical expertise and clear presentation.

Being a Courser has also underscored the importance of ethical considerations in data analysis.
Understanding the ethical implications of handling data, especially in marketing analytics, is
crucial for maintaining integrity and building trust with clients and customers. As an ongoing
intern, I recognize that learning is an ongoing journey, and my experience with MMM has laid
the foundation for continuous growth. The exposure to real projects, the application of
analytical techniques, and the guidance from seasoned professionals have collectively provided
me with an enriched learning environment that extends beyond theoretical knowledge.

It has been an exceptional opportunity to dive into the world of Marketing Mix Modelling. This
experience has expanded my skillset, deepened my understanding of marketing analytics, and
fueled my passion for data-driven decision-making. I am confident that the insights gained
from this internship will continue to shape my professional journey and contribute to my
growth as an aspiring data analyst.
REFERENCES
Fireball (sazerac.com)

Course5 Intelligence | Transformative intelligence

How Marketing Mix Modeling Helps Businesses to Make Better Decisions

(expressanalytics.com)

Understanding Ordinary Least Squares (OLS) Regression | Built In

Understanding Marketing Mix Modeling | Nielsen