Business Analytics

MBA232C6

Unit 1
 RELEVANCE
▪ Increasing complexities associated with businesses in the form of scale of operations and
competition
▪ Demand deeper understanding of the market and customers to serve better and succeed in the
market. One of the main reasons for analytics is the scale of operations.
▪ Competitive Advantage – Nvidia, Google, Amazon
▪ Remove inefficiencies in the system
▪ Provides ability to make better decision
▪ In 2021, the overall amount of data generated in the world was estimated to be around 79
zettabytes. Approximately 402.74 million terabytes of data are created each day. Around 147
zettabytes of data will be generated this year. 181 zettabytes of data will be generated in 2025.
 REASON FOR THE RISE OF ANALYTICS
▪Advanced software techniques are available e.g. advanced data structures, advanced
database systems, cloud computing, etc.
▪Clean data is available, now most organizations have robust software infrastructure
which helps in capturing clean customer, vendors, and sales data
▪Advanced hardwares are available which can store huge data in such a way that it can
be easily available for analysis without any time lag. Also, the cost is quite reasonable
e.g. GPU/TPU processors, distributed networks, etc.
▪Advanced business problem-solving techniques are providing new alternatives to
tackle business problems e.g. Agile and lean six sigma frameworks for business
excellence
 NVIDIA MKT CAP
➢If Nvidia was a country – 7th largest
in the world (Market capitalization)

➢Development of advanced graphics

processing units (GPUs) and its
strategic pivot towards artificial
intelligence (AI) and machine
learning. The company's GPUs, known
for their high performance and
efficiency, have become essential in
various sectors beyond gaming,
including data centers, automotive,
and professional visualization.

➢The human mind lacks the ability to

choose the right decisions due to the
complexity of the problems that the
organizations are facing and the
limited time available for decision
making.
 PROBLEMS FLIPKART IS DEALING WITH
Fraudulent transactions:
➢ Fake items in place of genuine items
➢ Number of stock keeping units (SKUs)
➢ Forecasting demand for products
➢Cancellation of orders placed by customers before their delivery
➢Predicting delivery time
➢Predicting what a customer is likely to buy in future

https://ojdigitalsolutions.com/amazon-sales-data/
SO, WHAT IS ANALYTICS AND WHAT IS ITS SCOPE?

Analytics is a body of knowledge consisting of statistical, mathematical, and operations

research techniques; artificial intelligence techniques such as machine learning and deep
learning algorithms; data collection and storage; data management processes such as data
extraction, transformation, and loading (ETL); and computing and big data technologies such
as Hadoop, Spark, and Hive that create value by developing actionable items from data.
Two primary macro-level objectives of analytics are:
a) Problem solving and
b) Decision making.

Davenport and Harris (2007) and Hopkins et al. (2010) reported that there was a high
correlation between use of analytics and business performance.
BUSINESS ANALYTICS ACROSS DOMAINS (SCOPE)
Scope:
• Data collection and Management
• Data Analysis
• Predictive modelling
• Data visualization
• Decision-making support
• Customer-Making support
• Customer Behavior Analysis
• Market research
• Inventory management
• Financial forecasting
• Operations optimization
• Customer Behaviour analysis
• Sales and Marketing analysis
• Supply chain Optimization
• Financial analysis
• Process improvement
• HRM and analysis
 QUIZ
Question 1: For which company does this problem arise : selling fake items in place of
genuine items?

A. E-commerce platforms
B. Logistics and delivery services
C. Fraudulent sellers
D. Data analytics and machine learning companies
Question 2: Which type of company manages a large inventory with a wide variety of
products, each represented by a unique SKU?

A. E-commerce platforms
B. Logistics and delivery services
C. Fraudulent sellers
D. Data analytics and machine learning companies
Question 3: Which type of company specializes in forecasting demand for products?

A. E-commerce platforms
B. Logistics and delivery services
C. Fraudulent sellers
D. Data analytics and machine learning companies
Question 4: Which type of company experiences a significant number of canceled
orders placed by customers before delivery?

A. E-commerce platforms
B. Logistics and delivery services
C. Fraudulent sellers
D. Data analytics and machine learning companies
Question 5: Which type of company predicts what a customer is likely to buy in the
future?

A. E-commerce platforms
B. Logistics and delivery services
C. Fraudulent sellers
D. Data analytics and machine learning companies
 QUIZ – RECAP OF SESSION 1

Question 1: The two macro-level objectives of analytics are:

A. Describing and Forecasting the data

B. Problem solving and Decision making
C. Analyzing the Data and proving output
D. Enabling critical and out-of-box thinking
Question 2: In which of the following domains is the application of business
analytics not involved? – why?

A. Healthcare
B. Stock Market
C. Gambling
D. Banking
 SESSION-2
BUSINESS ANALYTICS DEFINITION

Business analytics refers to the process of collecting, organizing, analyzing,

and interpreting data to uncover meaningful patterns and insights that can
drive business performance and competitiveness. It involves the application
of statistical analysis, data visualization, and predictive modeling to support
decision-making, solve complex business problems, and identify
opportunities for growth and optimization.
(Source: Adapted from Harvard Business Review)
 BUSINESS ANALYTICS DEFINITION
Business analytics refers to the practice of using data, statistical analysis, and
quantitative methods to gain insights and make informed decisions in a business or
organizational context. It involves the exploration, interpretation, and
communication of data to drive strategic, operational, and tactical decision-making.
(Source: Adapted from International Institute of Business Analysis - IIBA)
Meaning:
-Business Analytics refers to tools, techniques and process for continuous
exploration and investigation of past data to gain insights and help in decision
making.
-It is an integration between data science, technology and business context that assist
data driven decision making.
 OVERVIEW OF DATA ANALYTICS
❑ Encompasses Data, Technology and Business context to extract insights and support
decision making in organizations.
❑ Involves collection, analysis and interpretation of data to help organizations to identify
trends, measure performance and optimize processes.
❑ The goal is to turn data into actionable insights for strategy and improvement in operations.
❑ Business analytics tools and technologies include data warehousing, data mining, machine
learning, and visualization tools, among others.
❑ It requires combination of technical expertise and communication skills.
 BUSINESS ANALYTICS COMPONENTS

Business
context

Data
Technology
Science
Business Context:
- BA applications/ projects starts with business context.
- Ability of the organizations to ask the right questions – helps in selecting relevant
data and right tools – results in a good analytics storyboard.
Technology:
- used for data capture, data storage, data preparation, data analysis, and data share.
- Most data are unstructured – need software for analysis – automation of actionable
items.
Data Science:
- Consists of statistical and operations research techniques, machine learnings and deep
learning algorithms.
- Given the objective or problem, data science component of analytics is to identify the
most appropriate statistical model / machine learning algorithms that can be used.
 DATA DRIVEN DECISION MAKING
A typical data-driven decision-making process uses the following steps:
1. Identify the problem or opportunity for value creation. (Amazon Prime Air)
2. Identify sources of data - primary as well secondary data sources. (Customer feedback, GPS
and telematics data, traffic patterns, Online reviews, etc.)
3. Pre-process the data for issues such as missing and incorrect data. Generate derived variables
and transform the data if necessary (Normalization, standardization, handling skewed data,
less sensitive to outliers, storage efficiency, meeting assumptions of statistical tools). Prepare
the data for analytics model building.
4. Divide the data sets into subsets training and validation data sets.
5. Build analytical models and identify the best model(s) using model performance in validation
data.
6. Implement Solution/Decision/Develop Product
(Refer notes for Applications of typical data-driven decisions in various industries)
ANALYTICS CAN BE USED TO SOLVE VARIOUS KINDS OF PROBLEMS

Walmart’s cost leadership strategy

Amazon sales and

pricing strategy

Toyota Inventory
Management system

Healthcare
 BUSINESS ANALYTICS CAN BE GROUPED
INTO THREE TYPES:
1. Descriptive analytics,
2. Predictive analytics, and
3. Prescriptive analytics
 DESCRIPTIVE ANALYTICS
It involves the exploration and analysis of data to answer questions such as "What
has happened?" and "What is the current state?"
Descriptive analytics often includes tasks such as:
1. Data Exploration (identify patterns, distribution, outliers, Data cleaning, Data
Profiling, and data quality assessment); 2. Data Visualization (Charts, Graphs,
Dashboards); 3. Summary Statistics (mean, median, mode, range, std. deviation); 4.
Data Segmentation (grouping or categorizing data to understand the variation within
data); and 5. Data Reporting (highlighting key observations, presentation)
- It focuses on historical data and identifies the hidden trends, patterns and
relationships using graphs, measures of frequency, measures of central tendency,
measures of Dispersion or variation, and measures of position.
-Simplest form of analytics that mainly use simple descriptive statistics, data
visualization techniques, and business-related queries to understand past data.
- The primary objective is data summarization and understanding the trend in the past
data which can be useful for generating insights.
 PREDICTIVE ANALYTICS
Predictive analytics is a branch of business analytics that uses historical data, statistical
algorithms, and machine learning techniques to make predictions and forecasts about
future events or outcomes. It answers to the question ‘What will happen in the future?’
Here are some examples of predictive analytics applications:
1. Customer Churn Prediction: Customers who are likely to churn or cancel their subscription.
2. Demand Forecasting: Retailers & manufacturers forecast future demand of their product.
3. Fraud Detection: Financial institutions detect fraudulent transactions by analyzing patterns and
anomalies in data.
4. Predictive Maintenance: To optimize equipment maintenance schedules.
5. Credit Risk Assessment: Assess the default risk and determine appropriate credit limits or
interest rates.
6. Healthcare Outcome Prediction: Likelihood of readmission, probability of disease.
 DESCRIPTIVE ANALYTICS VS. PREDICTIVE ANALYTICS

Descriptive Analytics Predictive Analytics

Objective Descriptive analytics aims to provide a Predictive analytics aims to make predictions and
comprehensive understanding of historical forecasts about future events or outcomes based on
data and summarize past events and trends. historical data and patterns.

Focus It focuses on answering questions such as It focuses on answering questions such as "What is
"What happened?" and "What is the current likely to happen?" and "What is the probability of a
state?" specific event occurring?"

Insights Descriptive analytics provides insights into Predictive analytics uses statistical modeling and
the past by analyzing and summarizing data machine learning algorithms to uncover
through measures like mean, median, mode, relationships and patterns in historical data and
standard deviation, and data visualization applies them to predict future behavior or
techniques. outcomes.

Examples Data profiling, data visualization, summary Customer churn prediction, demand forecasting,
statistics, data segmentation, and reporting are fraud detection, predictive maintenance, credit risk
common techniques used in descriptive assessment, and healthcare outcome prediction are
analytics. some common applications of predictive analytics.
 PRESCRIPTIVE ANALYTICS
Prescriptive analytics is the highest level of business analytics that goes beyond
descriptive and predictive analytics. It focuses on providing recommendations or
prescriptions for actions to optimize decisions and outcomes. It acts as a solution
builder for a problem.
Here are some examples of prescriptive analytics applications:
1. Supply Chain Optimization (Flipkart’s Supply Chain Management: uses BA for
route optimization, Inventory optimization, Warehouse management etc.)
2. Pricing Optimization (Ola Cabs: Dynamic pricing algorithms based on rider
demand, driver availability, traffic conditions, driver incentives)
3. Resource Allocation in Healthcare (Apollo Hospitals – uses BA for optimal
allocation of medical staff (doctors, nurses, technicians), medical equipment (MRI
machines, surgical instruments), and hospital beds based on predicted demand and
resource availability)
4. Portfolio Optimization (Edelweiss Asset Management Limited - uses real-time
monitoring analytics tools, and optimization models to construct portfolios that
maximize returns given a specific level of risk tolerance or minimize risk for a given
level of expected return).

5. Dynamic Pricing in E-commerce (Indigo airlines uses BA for real-time price

adjustments depending on many factors, keeping the revenue and profit in mind).

6. Energy Management (Tata Power uses BA for managing energy resources

efficiently to meet the energy needs of its customers while optimizing operational
costs and reducing environmental impact).
 APPLICATION: COMPANIES USING ALL THE
THREE TYPES OF ANALYTICS
Amazon:

• Descriptive Analytics: Amazon analyzes customer behavior and purchase history to

understand patterns, preferences, and trends. They use this information to personalize
product recommendations and optimize their inventory management.

• Predictive Analytics: Amazon employs predictive analytics to forecast customer demand,

optimize pricing strategies, and anticipate customer preferences. This helps them anticipate
market trends and optimize their supply chain.

• Prescriptive Analytics: Amazon uses prescriptive analytics to optimize its logistics and
delivery operations. They analyze various factors such as transportation routes, weather
conditions, and customer preferences to determine the most efficient delivery options.
 APPLICATION: COMPANIES USING ALL THE
THREE TYPES OF ANALYTICS
Netflix:

• Descriptive Analytics:
• Predictive Analytics:
• Prescriptive Analytics:
 APPLICATION: COMPANIES USING ALL
THE THREE TYPES OF ANALYTICS
Uber:

• Descriptive Analytics:
• Predictive Analytics:
• Prescriptive Analytics:
 APPLICATION: COMPANIES USING ALL
THE THREE TYPES OF ANALYTICS

Procter & Gamble:

• Descriptive Analytics:
• Predictive Analytics:
• Prescriptive Analytics:
 APPLICATION: COMPANIES USING ALL THE
THREE TYPES OF ANALYTICS
Walmart:

• Descriptive Analytics:
• Predictive Analytics:
• Prescriptive Analytics:
TECHNIQUES: DESCRIPTIVE, PREDICTIVE AND
PRESCRIPTIVE;
Descriptive Analytics Predictive Analytics: Prescriptive Analytics:

Data Visualization: Regression Analysis: Optimization:

Descriptive Statistics: Time Series Analysis: Genetic Algorithms:
Data Mining: Machine Learning:
Cluster Analysis: Decision Trees: Simulation:
Text Mining: Neural Networks: Prescriptive
Cohort Analysis: Ensemble Methods: Simulation:

Decision Analysis:
Multi-Criteria Decision
Analysis (MCDA):
 BIG DATA ANALYTICS
Big data refers to extremely large and complex data sets that exceed the capabilities of
traditional data processing tools and techniques to capture, store, manage, and analyze.
Big data is characterized by the 5 V's: volume (large amount of data), velocity (High speed
data generation and processing), variety (diverse data type and different sources) , veracity
(data quality and accuracy - biased, incomplete, noise and abnormality in data) and value
(offers benefits).
Key characteristics of big data analytics include:
1. Advanced Analytics:
2. Scalability:
3. Real-time or Near Real-time Analysis:
4. Data Variety:
5. Data Integration:
6. Business Value:
 WEB AND SOCIAL MEDIA ANALYTICS
Web and social media analytics refers to the process of collecting, analyzing, and
interpreting data from websites and social media platforms to gain insights and make
data-driven decisions.

General steps followed in conducting web and social media analytics is as follows:
1. Data Collection: done through various methods such as web scraping, APIs
(Application Programming Interfaces), or 3rd party analytical tools.
2. Data Cleaning and Preparation: (Removing duplicates, handling missing data etc.)
3. Data Analysis: (includes statistical, network, sentimental analysis, text mining etc.)
4. Visualization and Reporting: (Charts, graphs, dashboards, reports etc.)
5. Interpretation and Action: (helps in making business decisions, optimizing
strategies, solving problems, improve customer experiences).
 FROM A BUSINESS PERSPECTIVE, WEB AND SOCIAL
MEDIA ANALYTICS IS FOUND TO HIGHLY RELEVANT:
1. Customer Insights: To create more targeted marketing campaigns, tailor products or
services to customer needs, and improve overall customer experiences
2. Marketing Optimization: To identify effective marketing channels, track the success
of campaigns
3. Brand Monitoring and Reputation Management: To proactively manage the brand
image, address any negative feedback
4. Competitor Analysis: To understand competitor strengths and weaknesses, and
adapting strategies accordingly.
5. Market Research and Trend Analysis: To help in staying ahead of industry trends,
identifying new market opportunities
6. Customer Service and Support: To provide timely and personalized responses, and
enhancing overall customer support experience.
INTRODUCTION TO MACHINE LEARNING
❑ Google: "Machine learning algorithms are a set of computational methods and models
that allow computers to automatically learn from data and make predictions or decisions
without explicit programming.“

❑ Microsoft: "Machine learning algorithms are mathematical models and statistical

techniques that enable computers to analyze data, learn from patterns and relationships,
and make predictions or decisions.“

❑ Machine learning is a branch of artificial intelligence (AI)

❑ It focuses on developing algorithms and statistical models that enable computers to

learn from and make predictions or decisions based on data.
INTRODUCTION TO MACHINE LEARNING
❑It is designed to learn from examples and experiences, improving their performance over
time.

❑ It requires dataset to train the machine learning model, consisting of input features
(attributes) and corresponding output labels or targets.

❑ ML models are evaluated based on their performance using metrics such as accuracy,
precision, recall, and F1-score (F1-score is a measure of a model's accuracy that combines
both precision and recall metrics into a single value. It is particularly useful when dealing
with imbalanced datasets where one class is more frequent than the other).

❑ Challenges in machine learning where a model may perform too well on training data
but poorly on new, unseen data (overfitting) or fail to capture the underlying patterns in
the data (underfitting).
 TYPES OF MACHINE LEARNING
1. Supervised Learning: A type of machine learning where the model is trained on labeled
data, allowing it to learn relationships between input features and output labels.

Examples:

❑ Linear regression: Predicting house prices based on features like area, number of
bedrooms, and location.

❑ Classification: Identifying whether an email is spam or not based on its content and
attributes.

❑Support Vector Machines (SVM): Classifying images into different categories, such as
recognizing handwritten digits.
2. Unsupervised Learning:

It is a type of machine learning that learns from data without human supervision.
Unsupervised learning involves training a model on unlabeled data, where the input data
does not have corresponding target labels. The model learns to identify patterns,
structures, or relationships in the data without specific guidance.

Examples:

❑ Clustering: Grouping similar customers based on their purchasing behavior.

❑ Dimensionality Reduction: Reducing the number of variables while retaining important

information.

❑ Anomaly Detection: Identifying unusual or abnormal data points in a dataset.

3. Reinforcement Learning:
It is a machine learning technique that trains software to make decisions to achieve the
most optimal result. Reinforcement learning involves an agent learning to make
sequential decisions in an environment to maximize a reward signal. The agent interacts
with the environment, takes actions, and learns through a process of trial and error.
Examples:
❑ Game Playing: Training an AI agent to play games like chess or other video games.

❑Robotics: Teaching a robot to perform tasks like walking, grasping objects, or

navigating in a dynamic environment.

❑Autonomous Vehicles: Training self-driving cars to make decisions in real-world

scenarios.
4. Evolutionary learning algorithms:
These are a category of machine learning algorithms inspired by the process of natural
evolution. They mimic the principles of natural selection, genetic inheritance, and
mutation to optimize solutions to complex problems. They have been applied in various
domains, including engineering design, scheduling, finance and data mining.
Examples:
Genetic Algorithms (GA): Genetic algorithms use techniques such as selection,
crossover, and mutation to search for optimal solutions in a large solution space. They are
commonly used for optimization problems, where the goal is to find the best combination
of parameters or variables that maximize or minimize a given objective function. (It was
initially conceived by Holland as a means of studying adaptive behavior)
Genetic Programming (GP): Genetic programming applies the principles of genetic
algorithms to evolve computer programs or mathematical expressions. It starts with a
population of randomly generated programs and uses genetic operations to evolve them
over generations to improve their fitness.
 FRAMEWORK FOR DECISION MAKING
The framework for data driven decision making or problem identification consists of
five integrated stages: These stages may often require iteration and refinement as new
challenges and insights arise.
1. Problem and Opportunity identification:
❑ It involves understanding the business context, gathering the requirements, and
clearly defining the objectives.
❑ It is crucial to have a clear understanding of what problem needs to be solved or
what opportunity needs to be pursued. (Example)

2. Collection of relevant data:

❑ Identify the data that is used for analysis. It includes identifying the data source
(internal and external) and ensuring data quality. (Example)
3. Data Pre-processing:
❑ To make data suitable for analysis. It involves data cleaning, data integration, handling
missing values, removing outliers, and transforming the variables if necessary. Data needs to
be complete, accurate, and relevant to the problem. (Example)
4. Analytics model building:
❑ To extract insights and make predictions from the data. It involves selecting appropriate
analytical techniques and algorithms based on the problem and type of data available.
❑ Model building include tasks such as feature selection, model training, parameter tuning,
and validation to ensure model’s performance and accuracy. (Example)
5. Model Deployment:
❑ The final stage of the framework is to deploy the analytics model to generate actionable
insights and support decision making.
❑It involves creating dashboards, reports that communicate the findings to stakeholders.
❑ The inferences and recommendations derived from the model are used to make informed
business decisions, drive improvements, and capitalize on the opportunities. (Example)
 ANALYTICS CAPABILITY BUILDING;
ROADMAP AND CHALLENGES
To build analytics capabilities, organizations can follow a roadmap that outlines the steps and
milestones for development. The roadmap typically includes the following components:
1. Strategy Alignment: Aligning analytics with business strategy; identify key areas for analytics
to create value.
2. Data Infrastructure: create robust data management systems, data warehouses, and data
lakes.
3. Talent Acquisition and Development: acquire / train employees on skills in data analytics,
statistics, machine learning and domain expertise.
4. Tools and Technology: Software for visualization, machine learning; big data processing.
5. Process Integration: It refers to integrating analytics into existing business processes. It
includes workflows, data governance policies, and performance metrics etc.
6. Culture and Change Management: Fostering data-driven culture. May require to incentivize
for adopting analytics practices.