CCW331 - BUSINESS ANALYTICS

UNIT I
INTRODUCTION TO BUSINESS ANALYTICS
Analytics and Data Science – Analytics Life Cycle – Types of Analytics – Business Problem
Definition – Data Collection – Data Preparation – Hypothesis Generation – Modeling – Validation and
Evaluation – Interpretation – Deployment and Iteration

ANALYTICS AND DATA SCIENCE

 Relationship: Analytics and data science are closely related fields that involve the use of data
to
a. Gain insights
b. Make informed decisions
c. Solve complex problems
 Analytics: systematic analysis of data to gain insights, makes informed decisions, and
uncovers patterns or trends. (using different types)
o Goal of Analytics in business Intelligence: Transform the raw data into actionable
intelligence, providing organizations with a competitive advantage and helping them
achieve their objectives.
Key Components of Business Analytics:
1. Descriptive Analytics
2. Predictive Analytics
3. Prescriptive Analytics
4. Data Exploration and Visualization
5. Business Intelligence
6. Data Mining
7. Big Data Analytics
8. Decision Support System(DSS)

Applications:
a. Business Intelligence
b. Marketing Analytics
c. Financial Analytics
d. Healthcare Analytics

1
 Data Science: It is multidisciplinary field that combines expertise from statistics, computer
science, and mathematics, and domain knowledge to extract knowledge and insights from data.
o Life cycle: It involves the entire data lifecycle, from data collection to deployment of
predictive models.
o Key Components:
i. Data Collection
ii. Data Cleaning and Preprocessing
iii. Exploratory Data Analysis(EDA)
iv. Modeling
v. Evaluation and Interpretation

o Applications:
1. Predictive Modeling
2. Natural Language Processing (NLP)
3. Image and Video Analysis
4. Recommender Systems

ANALYTICS LIFE CYCLE

 Need for Analytical Life Cycle:
1. Informed decision making
2. Performance improvement
3. Gaining competitive advantages in the market
 Goal: To transform raw data into meaningful insights that drive strategic decisions and
business success. It involves,
1. Understanding the business context
2. Identifying key metrics
3. Continuously improving analytical processes
 Input:
1. Raw Business Data
2. Business problem definition (objectives)
3. Tools
4. Domain knowledge

2
 Output:
1. Insights and Reports (Performance metrics)
2. Data Models
3. Recommendations
4. Continuous improvement for refining models (using performance metrics)
 Analytical Process Model

 Steps involved in Analytical Life Cycle

1. Define Business Objectives

2. Data collection – Gathering data from various data (Databases, spreadsheets or external
sources)
3. Data processing and Cleaning – Removing noisy data, handling duplicate data,
handling incomplete data
4. Data Transformation – translated in useful or unified format (Normalization)
5. Data Integration - combine the data from various sources if necessary
6. Data Exploration – Identify the patterns
7. Interpretation and visualization
8. Decision making
9. Continuous improvement

3
 TYPES OF BUSINESS ANALYTICS
 Purpose of Business Analytics:
1. Informed decision making
2. Enhanced business growth
 Types of Business Analytics:
1. Descriptive Analytics
2. Diagnostic Analytics
3. Predictive Analytics
4. Prescriptive Analytics
5. Text Analytics (or) Text Mining
6. Spatial Mining

1. Descriptive Analytics
o Goal: To summarize historical data and provide insights into what has happened in the
past.
o Need: Understanding past performance, identifying trends, and gaining a historical
perspective.
o Purpose: To describe, aggregate, and interpret historical data for reporting and
decision-making.
o Input: Historical data sets, databases, and structured information.
o Algorithms: Basic statistical methods, data aggregation techniques.
o Outputs:
a. Reporting: Summarizing and presenting historical data in a structured format.
b. Data Visualization: Using charts, graphs, and dashboards to represent data
visually.
c. Key Performance Indicators (KPIs): Identifying and monitoring important
metrics that indicate business performance.
4
 o Advantages:
a. Historical Understanding: Provides a clear understanding of past performance and
trends.
b. KPI Monitoring: Allows organizations to monitor key performance indicators to
assess business health.
c. Ease of Implementation: Generally easier to implement compared to more advanced
analytics.
o Disadvantages:
i. Limited Future Insight: Primarily focuses on the past, offering limited insights into
future trends.
ii. Not Actionable Alone: While informative, descriptive analytics often lacks
actionable recommendations for the future.
2. Diagnostic Analytics
o Goal: To examine data and identify the reasons behind past outcomes or trends.
o Need: Understanding the root causes of specific events or patterns observed in
historical data.
o Purpose: To diagnose problems, identify opportunities for improvement, and
understand the factors influencing outcomes.
o Input: Historical data, relevant context, and domain expertise.
o Output: Root cause analysis, trend identification, and insights into contributing factors.
o Algorithms: Statistical analysis, data correlation, and visualization techniques.
o Methods:
a. Drill-Down Analysis: Investigating data at a more detailed level to understand
specific factors.
b. Root Cause Analysis: Identifying the primary factors contributing to a particular
outcome.
c. Comparative Analysis: Contrasting different data sets to identify patterns or
anomalies.
o Advantages:
i. Root Cause Identification: Helps in identifying the root causes of problems or
successes.
ii. Informed Decision Making: Provides a deeper understanding of why certain
events occurred.
iii. Problem Solving: Supports problem-solving by addressing issues at their source.

5
 o Disadvantages:
a. Time-Consuming: Root cause analysis can be time-consuming and may not
provide immediate solutions.
b. Data Complexity: Identifying true root causes may be challenging in complex
systems.
c. Limited Predictive Power: While it explains the past, it may not necessarily
predict future occurrences.
3. Predictive Analytics
o Goal: To forecast future trends, outcomes, or behaviors based on historical data and
patterns.
o Need: Anticipating future events, making proactive decisions, and identifying potential
opportunities or risks.
o Purpose: To make predictions about future outcomes and optimize decision-making.
o Input: Historical data, variables with predictive value, and machine learning algorithms.
o Output: Predictive models, probability scores, and forecasts.
o Algorithms: Machine learning algorithms (e.g., regression, decision trees, neural
networks).
o Methods:
a. Regression Analysis
b. Time Series Analysis
c. Machine Learning
d. Data Mining
o Advantages:
a. Future Insight
b. Data-Driven Decision Making
c. Risk Mitigation
o Disadvantages:
a. Complexity
b. Data Quality Dependency
c. Overfitting Risk
4. Prescriptive Analytics
o Goal: To recommend actions that can optimize future outcomes or mitigate risks.
o Need: Providing actionable insights, suggesting optimal strategies, and guiding
decision-makers.
o Purpose: To prescribe specific actions to achieve desired outcomes and improve

6
 decision-making.
o Input: Predictive models, business rules, constraints, and optimization algorithms.
o Output: Prescriptive recommendations, decision support, and scenario analysis.
o Algorithms: Optimization algorithms, simulation models, decision trees.
o Methods:
a. Optimization: Finding the best solution among various possible options.
b. Simulation: Modeling different scenarios to understand their potential outcomes.
c. Decision Analysis: Evaluating decision alternatives and their potential impact.
d. Prescriptive Modeling: Using mathematical models to recommend actions.
o Advantages:
a. Optimized Decision Making
b. Scenario Analysis
c. Continuous Improvement
o Disadvantages:
a. Complexity
b. Dependency on Models
c. Resource Intensive
5. Text Analytics (or) Text Mining
o Goal: To analyze unstructured textual data, extracting meaningful insights and patterns.
o Need: Making sense of large volumes of text data, such as customer reviews, social
media comments, or documents.
o Purpose: To uncover sentiment, identify trends, and extract valuable information from
textual sources.
o Input: Textual data from sources like social media, customer feedback, and documents.
o Output: Sentiment analysis, topic modeling, and key phrase extraction.
o Algorithms: Natural Language Processing (NLP) techniques, sentiment analysis
algorithms.
6. Spatial Mining
o Goal: To analyze geographic or location-based data to discover spatial patterns and
relationships.
o Need: Understanding the spatial distribution of data, identifying clusters, and analyzing
geographic trends.
o Purpose: To make location-specific decisions, optimize routes, and visualize spatial
relationships.
o Input: Geographic data, maps, and location-based information.

7
 o Output: Maps, spatial visualizations, and insights into geographic patterns.
o Algorithms: Spatial clustering algorithms, Geographic Information System (GIS)
techniques.

BUSINESS PROBLEM DEFINITION

 Sources for framing problem definition: It arises from the complexity of organizational
challenges and the vast amount of data generated.
 Proper problem definition in crucial:
1. It provides clarity and specificity to analytical efforts.
2. It ensures that analytics projects align with strategic business objectives.
3. It guides the selection of appropriate analytics techniques and methodologies.
4. It helps in the efficient allocation of resources and prioritization of efforts.
 Parameters justify the effective problem definition:
a. Clarity
b. Alignment
c. Efficiency
d. Effectiveness
 Purpose of effective business definition
1. Strategic Alignment
2. Scope delimitation
3. Resource Optimization
4. Goal Setting
5. Stakeholder communication
 Collaboration Requirement:
a. Domain Experts
b. Stakeholders
c. Analytical professionals
 Condition: Frequent communication and feedback is essential for refining the problem definition.
 Input:
a. Organizational Goals
b. Data
c. Stakeholder Input
d. Market Conditions
 Output:
a. Problem description
b. Scope
8
 c. Objectives
d. Success Criteria

DATA COLLECTION
 Purpose of Data Collection: It gathers the relevant and high-quality data to support the
analysis and decision-making processes.
1. Extract valuable insights
2. Identify patterns
3. Identify the strategic business actions
 Need for Effective Data Collection:
a. Informed Decision making
b. Model development
c. Performance measurement
d. Trend Identification
e. Risk management
 Key Components:
1. Define clear objectives and key metrics
2. Identify the various data sources
3. Ensuring the quality of data
4. Data privacy and compliance
5. Identify the various data collection methods
6. Identify the proper storage management techniques
7. Data integration
8. Ensure data security
9. Documentation
 Different ways of Data Collection
1. Quantitative data collection method
2. Qualitative data collection method
a. Interviews
b. Observations
c. Product reviews
d. Answers to open-ended questions
e. Group Discussions
 Advantages of Effective data Collection
1. Increase ROI
2. Effective model design
9
 3. Reduce errors in predictions
4. Enhance marketing effects

DATA PREPARATION
 Need for Data preparation: It is the process of preparing raw data so that it is suitable for further
processing and analysis.
o It helps organizations make better informed business, improe ROI
 It is the process of gathering, combining, structuring and organizing data so it can be used in
business intelligence (BI), analytics and data visualization applications.
 Key steps Involved:
a. Collecting raw data
b. Cleaning raw data
c. Labeling raw data
 Steps involved in data preparation process:
10. Data collection – Gathering data from various data (Databases, spreadsheets or external
sources)
11. Data Cleaning – Removing noisy data, handling duplicate data, handling incomplete
data
12. Data Transformation – translated in useful or unified format (Normalization)
13. Data Integration - combine the data from various sources if necessary
14. Data Exploration – Identify the patterns
15. Handling Imbalanced data
16. Data splitting
17. Data documentation
18. Data Security and Privacy
19. Automation
 Steps involved to handle the messy data (dirty data):

10
  Benefits:
1. Improved Data Quality
2. Enhanced Data Consistency
3. Effective Handling of Missing Data
4. Increased Accuracy in Analysis
5. Feature Engineering Opportunities
6. Optimized Model Performance
7. Alignment with Business Objectives
8. Reduced Risk of Errors
9. Improved Data Understanding
10. Supports Reproducibility

HYPOTHESIS GENERATION
 Definition: It involves formulating educated guesses or assumptions about the relationships
between variables or factors that might influence a business outcome.
 Structured Approach: It provides a structured approach to investigating relationships between
variables and making informed decisions based on data.
 Need for Hypothesis Generation:
a. Guides Analytics
b. Problem FormulationPrioritization
c. Minimizes Bias
d. Efficient Resource Utilization
e. Measurable Outcomes
 Goal: Formulate the testable hypotheses that can be analyzed using data to gain insights into the
factors influencing a business outcome.
 It is crucial for guiding data analysis and research effort.
 It is critical steps in business analytics process.
 It is an educated guess of the various factors affectingthe business problem to be resolved.
 Purpose of Hypothesis Generation:
a. It helps to identify the articulate potential relationships or patterns in data for guiding
the analysis process.
b. Analysts can test assumptions, identify Key factors, and make informed decisions based
on evidence.
 Precondition:
1. Accessing relevant data
11
 2. Understanding of problem statement and objectives
3. Understanding of various parameters involves in the problem statement
 Input: Knowledge about the problem statement
 Output: Set of tested formulated hypothesis
 Key aspects of Hypothesis Generation
a. Understanding Business Objectives
b. Reviewing Existing Knowledge
c. Identifying Variables
d. Formulating Hypotheses
e. Ensuring Testability
f. Considering Alternative Hypotheses
g. Quantifying Hypotheses
h. Aligning with Data Availability
i. Prioritizing Hypotheses
j. Iterative Process
k. Consulting with Stakeholders
l. Documenting Hypotheses
m. Testing and Analysis
n. Iterative Analysis and Refinement
 Different Techniques:
1. Simple Hypothesis Generation
2. Multiple Hypothesis Generation
3. Quadrant Hypothesis Generation
 Algorithms used:
a. Descriptive Statistics
b. Correlation Analysis
c. Regression Analysis
d. Hypothesis Testing
e. Machine Learning Algorithms
f. Clustering Algorithms
g. Time Series Analysis
h. Text Mining and NLP
i. Simulation and Modeling
j. Optimization Algorithms

12
 Example:
 Problem Statement: company wants to increase user engagement on its e-commerce
platform.
 Hypothesis: Increasing the visibility of product recommendations on the homepage
(independent variable) will lead to a significant increase in user click-through rates
(dependent variable).
 Quantified Hypothesis: A 20% increase in the visibility of product recommendations will
result in a 15% increase in user click-through rates.
 Alternative Hypothesis: Factors other than the visibility of product recommendations may
influence user engagement, such as the relevance of recommendations or the overall user
interface.
 Advantages:
1. Guided Analysis
2. Informed Decision making
3. Resource Efficiency
4. Achieving specific business objectives.
 Disadvantages:
1. Biased on few parameters
2. Limited Exploration
3. Risk of overlooking factors

MODELING
 Definition: The process of creating a mathematical or computational representation of a real-
world business problem or system.
 It is defined as the representation of business or solution which included graphical
representation with supports text and relationship to other features.
 Need: Methodologies needed to extract meaningful information from the raw data
 Goal:
1. Gain insights from the historical data
2. Make predictions
3. Optimized decision making
 Diagrammatic Representation:

13
 Purpose of Modeling:
1. Analyze the data
2. Make predictions
3. Optimized decision making
 Precondition:
1. Accessing relevant data
2. Understanding of problem statement and objectives
3. Understanding of various parameters involves in the problem statement
 Input: Historical business information
1. Historical records
2. Transactional data
3. Customer information
4. Other relevant variables depending on the business problem
 Algorithms:
1. Descriptive Statistics
2. Correlation Analysis
3. Regression Analysis
4. Hypothesis Testing
5. Machine Learning Algorithms
6. Clustering Algorithms
7. Time Series Analysis
8. Text Mining and NLP
9. Simulation and Modeling
10. Optimization Algorithms
 Steps involved in modeling
1. Define the Problem
2. Data Collection
3. Data Pre-processing
4. Feature Selection/Engineering
5. Choose Model Type
6. Train the Model
7. Validate and Test
8. Deploy the Model
9. Monitor and Update

14
 Advantages:
1. Informed Decision Making
2. Improved Efficiency
3. Risk Assessment and Mitigation
4. Competitive Advantages
5. Resource Optimization
 Disadvantages:
1. Data Quality and Interpretation
2. Complexity and Expertise
3. Overrealiance on Models
4. Cost and Time Intensive
5. Model Degradation

15
 VALIDATION AND EVALUATION
 Need for Validation and Evaluation of Business Model:
1. Verify the reliability of the model
2. Prevent overfitting and under fitting
3. Model must contributes positive contributions and recommendations to the business
 Goal:
1. Reliability of model
2. Accurate
3. Robust
4. Align with business objectives
 Validation Process: It ensures the modes accurately represents the underlying business
processes and captures relevant patterns in the data.
 Evaluation Process: Evaluate the performance of the model in terms of its predictive power,
reliability, and alignment with business goals.
o Two Methods
a. Hold – Out - dataset is randomly divided to three subsets:
i. Training set
ii. Validation set
iii. Test set
b. Cross Validation – It is a resampling method that uses different portions of the
data to test and train a model on different iterations.
 Four Elements:
a. Conceptual design
b. System validation
c. Data validation and quality assessment
d. Process validation
 Input:
1. Training and testing data set
2. Key Performance indicators (KPIs)
 Output: Validated and well defined business model
 Precondition:
1. Accessing relevant data
2. Understanding of problem statement and objectives
3. Key Performance Indicators (KPIs)

16
 Steps involved:
1. Train the Model (use historical data)
2. Data Splitting (training and testing data)
3. Model Validation
4. Cross-Validation (testing with different subset of data to ensure consistency and
reliability)
5. Performance Metrics – Calculate various performance metrics
6. Business Metrics - evaluate the impact of the model on key business metrics.
7. Model Interpretability - Assess the interpretability of the model's decisions.
8. Sensitivity Analysis: - how changes in the input affect the models performance
 Algorithms:
1. Descriptive Statistics
2. Correlation Analysis
3. Regression Analysis
4. Hypothesis Testing
5. Machine Learning Algorithms
6. Clustering Algorithms
7. Time Series Analysis
8. Text Mining and NLP
9. Simulation and Modeling
10. Optimization Algorithms

DEPLOYMENT AND ITERATION

 Concept: The validated model is deployment in the operational environment, allowing it to
make the real time predictions or recommendation.
 Purpose of Iteration: Iteration involves refining the model based on ongoing feedback and
performance monitoring.
 Need for deployment and integration:
a. Improve the business growth and ROI
b. Model need to adopt with changing business conditions
c. Improving accuracy
d. Maintenance relevance over time
 Input:
a. Validated and well training business model
b. Business Data

17
 Output:
a. Generations of predictions or recommendations
Output for Iteration:
a. Model updation needs
b. Improved performance metrics
c. Enhanced accuracy
 Steps involved:
1. Model Deployment - Implement the trained model into the operational environment.
2. Monitoring and Feedback - continuously monitor the model's performance and collect user
feedback.
3. Performance Evaluation - Evaluate the model's performance in the operational
environment.
4. Iteration Planning - plan for model iteration based on feedback and performance
evaluation.
5. Model Iteration - refine the model based on feedback and identified areas for
improvement.
6. Deployment of Updated Model - Deploy the updated model into the operational
environment.
7. Ongoing Monitoring - continue monitoring the performance of the updated model.
 Algorithms:
a. Descriptive Statistics
b. Correlation Analysis
c. Regression Analysis
d. Hypothesis Testing
e. Machine Learning Algorithms
f. Clustering Algorithms
g. Time Series Analysis
h. Text Mining and NLP
i. Simulation and Modeling
j. Optimization Algorithms

18
 INTERPRETATION
 Need for Interpretation:
a. The impact of business process while using the models recommendations or predictions
b. Effective decision making
 Goal: Actionable information from the model predictions
 Interpretation aids helps to predict various driving factors and implications in the business
strategies
 The interpretation of data is subjective and it varies from business to business.
 Role of various Experts:
1. Ability to analyze the model’s output
2. Clear documentation of predictions with model’s features and variables
 Input:
1. Deployed business model
2. Real world raw data for prediction
3. Relevant contextual information
 Output: Interpretation of useful information
 Various formats of Interpreted Output:
1. bar graphs
2. Line charts
3. Histograms
4. Pie charts
5. Tabular forms
 Two methods of Interpretation:
1. Qualitative method
2. Quantitative method

19
 UNIT II - BUSINESS INTELLIGENCE
Data Warehouses and Data Mart - Knowledge Management –Types of Decisions - Decision
Making Process - Decision Support Systems – Business Intelligence –OLAP – Analytic
functions
DATA WAREHOUSES AND DATA MART
 Definition of Data warehouse: It is a centralized repository for storing and managing
large volumes of data from various sources within an organization.
 Definition of data Mart: It is a subset of a data warehouse that is focused on a specific
business function or department within an organization.
 It is smaller in size
 It is designed to meet the specific analytical needs of a particular group of users.
 Data marts are often created to provide quicker and more localized access to
relevant data for specific business units.
 Three Tier Architecture:

 Various Characteristics of Data warehouse:

1. Subject Oriented  subject-oriented approach allows for a comprehensive and
consolidated view of data related to specific business areas
 2. Integrated  Data integrated from various sources. It ensures consistency and
coherence in the data
3. Time – variant  Data are stored to provide information from an historic perspective
4. Nonvolatile  physically separate store of data transformed from the application
data found in the operational environment
Two operations:
a. Initial loading of data
b. Access of data
 Three Data warehouse Models:
1. Enterprise warehouse
- An enterprise warehouse collects all of the information about subjects
spanning the entire organization
- It provides corporate-wide data integration, usually from one or more
operational systems or external information providers, and is cross-functional
in scope
- Size of Database: GB, TB
2. Data Mart
- It is a subset of a data warehouse that is focused on a specific business
function or department within an organization.
- It is smaller in size
- It is designed to meet the specific analytical needs of a particular group of
users.
- Data marts are often created to provide quicker and more localized access to
relevant data for specific business units.
3. Virtual warehouse - It is a set of views over operational databases.
 Knowledge Management
 Need for Knowledge Management:
1. Informed Decision making
2. Innovation
3. Efficiency - reduces redundancy and streamlining process
4. Adaptability (changing environment and competitive advantages)
 Systematic Process: It involves the systematic process of capturing, organizing, storing,
and sharing information and insights derived from data analysis.
 Goal: It aims to enhance decision-making, foster innovation, and facilitate continuous
learning within an organization.
 Steps involved in extraction of knowledge discovery

 Data Mining: It is the process of discovering interesting patterns and knowledge

from large amounts of data. (Data sources : databases, data warehouses, the Web,
other information repositories, or data that are streamed into the system dynamically)
 TYPES OF DECISIONS
(Types of business analytics)

DECISION MAKING PROCESS

 Need for Decision making process:
1. Informed decision making
2. Performance improvement
3. Gaining competitive advantages in the market
 Goal: To transform raw data into meaningful insights that drive strategic decisions and
business success. It involves,
1. Understanding the business context
2. Identifying key metrics
3. Continuously improving analytical processes
 Input:
1. Raw Business Data
2. Business problem definition (objectives)
3. Tools
4. Domain knowledge
 Output:
1. Insights and Reports (Performance metrics)
2. Data Models
3. Recommendations
4. Continuous improvement for refining models (using performance metrics)
 Analytical Process Model
 Algorithms used:
1. Descriptive Statistics
2. Correlation Analysis
3. Regression Analysis
4. Hypothesis Testing
5. Machine Learning Algorithms
6. Clustering Algorithms
7. Time Series Analysis
8. Text Mining and NLP
9. Simulation and Modeling
10. Optimization Algorithms
 Steps involved in Decision making process
1. Define Business Objectives
2. Data collection – Gathering data from various data (Databases, spreadsheets or
external sources)
3. Data processing and Cleaning – Removing noisy data, handling duplicate data,
handling incomplete data
4. Data Transformation – translated in useful or unified format (Normalization)
5. Data Integration - combine the data from various sources if necessary
6. Data Exploration – Identify the patterns
7. Hypothesis formulation
8. Modelling and predictive Analysis
9. Data Validation
10. Interpretation and visualization
11. Decision making
12. Continuous improvement
 DECISION SUPPORT SYSTEMS (DSS)
 A decision support system is an information system that aids decision-making by
providing interactive tools and information to support managers and other decision-
makers in solving unstructured problems.
 Different parts of DSS:
1. Knowledge base
2. Software
3. User interface
 Components of DSS:
1. Data Management
 Role: Ingests, stores, and manages data from various sources.
 Business Analytics Integration: Interfaces with business analytics tools to
ensure access to relevant data.
2. Model Management
 Role: Manages analytical models and algorithms.
 Business Analytics Integration: Integrates with advanced analytics tools for
predictive modelling and analysis.
3. User Interface
 Role: Provides a user-friendly interface for interacting with the system.
 Business Analytics Integration: Allows users to visualize and explore
analytics results through dashboards and reports.
4. Knowledge base
 Role: Stores domain-specific knowledge and business rules.
 Business Analytics Integration: Incorporates business rules relevant to
analytics processes.
5. Decision making component
 Role: Facilitates decision-making by presenting insights and
recommendations.
 Business Analytics Integration: Utilizes analytics results to provide informed
decision support.
 Types of DSS:
1. Data driven DSS – it makes decisions based on data from internal databases or
external databases
2. Model driven DSS
 3. Communication driven DSS - uses a variety of communication tools -- such as
email, instant messaging or voice chat -- to allow more than one person to work
on the same task
4. Knowledge driven DSS - data that drives the system resides in a knowledge base
that is continuously updated and maintained by a knowledge management system.
5. Document driven DSS – use documents to retrieve data

Online Analytical Processing (OLAP)

 Goal: It allows users to interactively analyze multidimensional data from multiple
perspectives.
 Purpose: It enables users to explore and analyze data to gain insights into business trends,
patterns, and performance.
 Need for OLAP:
a. Gain insights from data for informed decision-making
b. Identify trends, patterns, and outliers in datasets
c. Forecast future outcomes and make predictions
d. Discover hidden relationships and correlations
e. Optimize processes, resource allocation, and strategies
 Goal of OLAP:
a. Provide a multidimensional view of data for analysis
b. Facilitate interactive and exploratory analysis
c. Support complex calculations and aggregations
d. Enable users to make informed and data-driven decisions
 Characteristics of OLAP:
1. Multidimensional data model
2. Interactivity
3. Calculations
4. performance
 OLAP operations:
a. Slice and dice
b. Drill down
c. Roll up
d. Pivot
 ANALYTIC FUNCTIONS
 Purpose: Analytic functions in business intelligence are essential tools and techniques
used to analyze and derive meaningful insights from data.
- It helps organization transform,
a. Gain insights from data for informed decision-making.
b. Identify trends, patterns, and outliers in datasets.
c. Forecast future outcomes and make predictions.
d. Discover hidden relationships and correlations.
e. Optimize processes, resource allocation, and strategies.
 Goal: The primary goal of analytic functions is,
a. Transform raw data into actionable insights
b. Enabling organizations to make informed decisions
c. Improve performance
d. Gain a competitive advantage
 Precondition:
a. Clean and well-structured data
b. Understanding of the business context and objectives
c. Access to relevant analytical tools and technologies
d. Knowledge of statistical and mathematical concepts
 Input:
a. Raw data from various sources (databases, spreadsheets, etc.).
 b. Parameters and configurations for specific analytic methods.
c. Business context and objectives for analysis.
d. Relevant variables and features for analysis.
 Output:
a. Visualizations such as charts and graphs
b. Summary statistics and key performance indicators (KPIs)
c. Predictive models and forecasts
d. Clusters or segments identified in the data
e. Actionable insights and recommendations
 Different types of Descriptive Statistics:
 Goal: Summarize and describe the main features of a dataset.
 Methods: Mean, median, mode, standard deviation, range, percentiles.
 Purpose: Provide a high-level overview of data distribution and central tendencies.
 Regression Analysis
 Goal: Examine relationships between variables.
 Methods: Linear regression, logistic regression.
 Purpose: Identify and quantify the relationships between dependent and independent
variables.
 Time Series Analysis
 Goal: Analyze data points collected over time.
 Methods: Moving averages, trend analysis, and seasonal decomposition.
 Purpose: Uncover patterns, trends, and seasonality in time-dependent data.
 Clustering
 Goal: Group similar data points together.
 Methods: K-means clustering, hierarchical clustering.
 Purpose: Identify natural groupings or clusters within the data.
 Classification
 Goal: Assign data points to predefined categories or classes.
 Methods: Decision trees, support vector machines, Naive Bayes.
 Purpose: Categorize data based on learned patterns and features.
 Predictive Modelling:
 Goal: Forecast future trends or outcomes.
 Methods: Machine learning algorithms (e.g., regression, decision trees).
 Purpose: Provide insights into future scenarios and support proactive decision-
making.
 Data Mining
 Goal: Discover patterns and relationships in large datasets.
 Methods: Association rule mining, anomaly detection.
 Purpose: Uncover hidden insights and trends that may not be apparent through
traditional analysis.
 Sentiment Analysis
 Goal: Determine the sentiment expressed in textual data.
 Methods: Natural Language Processing (NLP), sentiment scoring.
 Purpose: Understand customer opinions, feedback, and social media sentiment.
 Optimization
 Goal: Find the best solution to problem under given constraints.
 Methods: Linear programming, nonlinear optimization.
 Purpose: Optimize resource allocation, production processes, or decision-
making.
 Pattern Recognition
 Goal: Identify and classify patterns within data.
 Methods: Neural networks, deep learning algorithms.
 Purpose: Recognize complex patterns and relationships in large datasets.
 Steps involved
1. Define Objectives
2. Data Preparation
3. Variable Selection
4. Choose Analytic Methods
5. Apply Algorithms
6. Interpret Results
7. Validate and Refine
8. Communicate Findings
 LOGIC AND DATA DRIVEN MODELS

The Predictive Analytics: The predictive analytics used in the BA process

relate to the application of the advanced statistical, software as well as the
research methods so as to identify the predictive set of variables. The model is
built in order to identify the relations as well as trends that are not generally
seen with descriptive models.

Predictive Modelling:

1. Data driven model 2. Logical driven model

Logical Driven Model:

Definition: A logic-driven model is one based on experience, knowledge, and

logical relationships of variables and constants connected to the desired business
performance outcome situation

Need: The need for logic-driven models arises from their ability to provide
structured and rule-based decision-making.

Concept: Model building requires an understanding of business systems and the

relationships of variables and constants that seek to generate a desirable business
performance outcome.

Key Steps for Logical Driven Model:

1. Define the Problem

2. Identify Decision Rules
3. Collect Relevant Data
4. Pre-process and Prepare Data
5. Design the Inference Engine
6. Knowledge Representation
 7. Implement the Model
8. Testing and Validation
9. Interpretability and Transparency
10. Deployment
11. Monitoring and Maintenance
12. Documentation

Formula to Calculate the Logical Driven Model:

Profit = Revenue − Cost, or
Profit = (Unit Price × Quantity Sold) − [(Fixed Cost) + (Variable Cost ×
Quantity Sold)], or
P = (UP × QS) − [FC + (VC × QS)]
Influence diagrams can be useful to conceptualize the relationships of variables
in the development of models.

It is necessary to be knowledgeable about the business systems being modelled

in order to capture the relevant business behaviour. Cause-and-effect diagrams
and influence diagrams provide tools to conceptualize relationships, variables,
and constants, but it often takes many other methodologies to explore and develop
predictive models.

• Logic driven models are created on the basis of inferences and postulations
which the sample space and existing conditions provide.
• Creating logical models require solid understanding of business functional
areas, logical skills to evaluate the propositions better and knowledge of business
practices and research.

Advantages:

Interpretability: Logic-driven models are often more interpretable, as the

decision-making process is based on explicit rules.

Transparency: The decision-making process is transparent, making it easier to

validate and understand.

Disadvantages:

Limited Adaptability: Logic-driven models may struggle to adapt to complex

and dynamic environments where rules are not easily defined.

Dependency on Experts: The performance of these models heavily depends on

the accuracy and completeness of the predefined rules .

DATA DRIVEN MODEL

Logic-driven modelling is often used as a first step to establish relationships
through data-driven models (using data collected from many sources to
quantitatively establish model relationships).

Definition: Data driven business models are based on the collection, analysis,
and use of data. This data provides the opportunity to gain valuable insights, make
more effective decisions, and identify new business opportunities.

Objectives:
• The main aim of data-driven model concept is to find links between the
state system variables (input and output) without clear knowledge of the physical
attributes and behavior of the system.
• The data driven predictive modelling derives the modelling method based
on the set of existing data and entails a predictive methodology to forecast the
future outcomes.

• It is data-driven only when there is no clear knowledge of the relationships

among variables/system, though there is lot of data. Here, you are simply
predicting the outcomes based on the data.

• The model is not based on hand-picked variables, but may contain

unobserved, hidden combination of variables.

Key Features:

1. Training on Historical Data

2. Predictive Analytics
3. Business Intelligence
4. Optimization and Decision Support
5. Customer Segmentation
6. Fraud Detection
7. Continuous Learning
8. Evaluation Metrics

STEPS for Data Driven Model:

1. Sampling and estimation.

2. Regression Analysis
3. Correlation Analysis
4. Probability Distribution
5. Predictive modelling and analysis
6. Forecasting model

Merits of Data-Driven Models:

1. Predictive Analytics
 2. Personalization
3. Optimization of Processes
4. Competitive Advantage
5. Strategic Decision-Making

Demerits of Data-Driven Models

1. Data Quality Dependency

2. Over fitting
3. Dependency on Historical Data
4. Complexity
5. Cost of Implementation
 MACHINE LEARNING FOR PREDICTIVE ANALYTICS

DEFINITION: Predictive analytics (PA) and machine learning (ML) are powerful tools
for uncovering insights in large volumes of data.

Machine Learning: Machine learning involves training algorithms, neural networks, or

processing computers to analyse data and output findings at scale. This output can include
recommendations, automated text, or flagged outliers.

NEED FOR PREDCTIVE ANALYTICS: Predictive analytics involves advanced statistics,

including descriptive analytics, statistical modelling and large volumes of data. Predictive
analytics can include machine learning to analyse data quickly and efficiently. Like machine
learning, predictive analytics doesn't replace the human element. Instead, PA supports data
teams by reducing errors and uncovering significant insights.

Steps To Perform Predictive Analysis:

Some basic steps should be performed in order to perform predictive analysis.

• Define Problem Statement:

Define the project outcomes, the scope of the effort, objectives, identify the data sets that are
going to be used.
• Data Collection:

Data collection involves gathering the necessary details required for the analysis. It involves
the historical or past data from an authorized source over which predictive analysis is to be
performed.

• Data Cleaning:

Data Cleaning is the process in which we refine our data sets. In the process of data cleaning,
we remove un-necessary and erroneous data. It involves removing the redundant data and
duplicate data from our data sets.

• Data Analysis:

It involves the exploration of data. We explore the data and analyze it thoroughly in order to
identify some patterns or new outcomes from the data set. In this stage, we discover useful
information and conclude by identifying some patterns or trends.

• Build Predictive Model:

In this stage of predictive analysis, we use various algorithms to build predictive models based
on the patterns observed. It requires knowledge of python, R, Statistics and MATLAB and so
on. We also test our hypothesis using standard statistic models.

• Validation:

It is a very important step in predictive analysis. In this step, we check the efficiency of our
model by performing various tests. Here we provide sample input sets to check the validity of
our model. The model needs to be evaluated for its accuracy in this stage.

• Deployment:

In deployment we make our model work in a real environment and it helps in everyday
discussion making and make it available to use.

• Model Monitoring:

Regularly monitor your models to check performance and ensure that we have proper results.
It is seeing how model predictions are performing against actual data sets.
IMPROVE PREDICTIVE ANALYSIS WITH MACHINE LEARING:

E-commerce/retail

Predictive analytics achieved through machine learning helps retailers understand customers’
preferences. It works by analysing users’ browsing patterns and how frequently a product is
clicked on in a website.

Customer service

Customer segmentation is performed based on insights by predictive analytics. Customers are

placed into different segments depending on their purchase patterns.

Medical diagnosis

Machine learning models that are trained on large and varied datasets can study patient
symptoms comprehensively to provide faster and more accurate diagnoses. Performing
predictive analytics on the reasons behind past hospital readmissions can also improve care.

Sales and marketing

Predictive analytics of historical data of customer behaviour and market trends can help
businesses understand the demands of prospective customers. Demand forecasting also helps
businesses estimate the demand for certain products in the future.

Financial services

Predictive analytics using machine learning helps detect fraudulent activities in the financial
sector. Fraudulent transactions are identified by training machine learning algorithms with past
datasets. The models find risky patterns in these datasets and learn to predict and deter fraud.
Cyber security

Machine learning algorithms can analyse web traffic in real-time. When an unusual pattern is
observed, advanced statistical methods of predictive analytics foresee and prevent cyber-
attacks. They also automatically collect attack-related data and generate useful reports on a
cyber-attack, thereby reducing the need for manpower.

Manufacturing

Machine learning and predictive analytics help manufacturers monitor machines and notify
them when crucial components need to be repaired or replaced. They can also predict market
fluctuations, reduce the number of accidents, improve key performance indicators (KPIs), and
enhance overall production quality.

Merits of Machine Learning for Predictive Analysis:

1. Automation and Efficiency

2. Accurate Predictions
3. Scalability
4. Continuous Learning
Demerits of Machine Learning for Predictive Analysis:

1. Data Quality Dependency

2. Interpretability Challenges
3. Over fitting
4. Ethical Concerns
5. Resource Intensiveness
Applying HR Analytics to make a prediction of the demand
for hourly employees for a year

HR Analytics :

 Involves the use of data and statistical methods to gain

insights into human resources-related processes and improve
organizational decision-making.

 Also known as talent analytics or workforce analytics.

Goal :

 Optimizing Workforce Planning

 Enhancing Talent Acquisition
 Employee Turnover
 Improving Employee Performance
 Elevating Employee Engagement
 Ensuring Leadership Succession
 Enhancing Remote Work Strategies

Requirements :

 Historical data
 Statistical models
 Machine learning algorithms

Procedure:

Let's consider an example of a retail business that needs to predict the

demand for hourly employees over the course of a year. In this scenario,
the business aims to optimize staffing levels based on anticipated
customer traffic.
Step1: Objective Definition

The objective is to predict the hourly employee demand for each

day to ensure the right staffing levels, minimizing both overstaffing and
understaffing.

Step 2: Data Collection

Gather historical data, including daily sales, foot traffic, and the
number of hourly employees scheduled for each shift. Additional data
might include promotions, holidays, and special events that could
impact customer demand.

Step 3: Data Cleaning and Preprocessing

Clean the data by handling missing values and outliers. Convert

timestamps into appropriate formats and ensure consistency across the
dataset.

Step 4: Feature Engineering

Create features such as day of the week, month, holidays, and

special events. Calculate average sales and foot traffic for each day to
capture trends.

Step 5: Exploratory Data Analysis (EDA)

Analyze the data to identify patterns. For example, you might

observe increased demand during weekends, holiday seasons, or
promotional events.

Step 6: Model Selection

Choose a suitable model for time series forecasting. In this case,

let's say you opt for a SARIMA (Seasonal AutoRegressive Integrated
Moving Average) model due to the presence of seasonality in the data.
Step 7: Training the Model

Split the data into training and testing sets. Train the SARIMA
model on the historical data to learn patterns and relationships.

Step 8: Validation and Hyperparameter Tuning

Validate the model using the testing set and fine-tune its
hyperparameters to optimize performance, ensuring that the model
accurately predicts hourly employee demand.

Step 9: Evaluation

Evaluate the model's performance using metrics like Mean

Absolute Error (MAE) or Root Mean Squared Error (RMSE) to quantify
the accuracy of predictions.

Step 10: Deployment

Deploy the trained SARIMA model to make daily predictions for

hourly employee demand. These predictions can be used to adjust
staffing levels accordingly.

Step 11: Continuous Monitoring and Updating

Periodically monitor the model's performance and update it as

necessary. For instance, if there are changes in customer behavior or
external factors affecting demand, the model should be adjusted
accordingly.

Step 12: Feedback Loop

Establish a feedback loop with HR professionals and store

managers. Their insights and feedback can help improve the model over
time, ensuring that it aligns with business goals and adapts to changing
conditions.
Benefits :

 Optimized Staffing Levels

 Improved Customer Experience
 Enhanced Employee Productivity and Satisfaction
 Cost Savings
 Data-Driven Decision-Making
 Resource Allocation
 Competitive Advantages
 Strategic Workforce Planning

Limitations :

 Data Quality
 External factors
 Human Factors
 Complexity of Human Behaviors
 Lack of Causation
 Overfitting
 Privacy and Ethical Concerns
 Analytics in HR and Supply Chain

Analytics :

Definition : The systematic analysis of data to gain insights,

make informed decisions, and uncover patterns or trends.

Goal : To transform raw data into actionable intelligence,

providing organizations with a competitive advantage and helping
them achieve their objectives.

Two Domains :

1. Human Resource (HR) Analytics

2. Supply Chain Analytics

Analytics in HR and Supply Chain :

Role : Providing insights that can enhance decision-making,

optimize processes, and improve overall efficiency.

HR Analytics:

Definition :
 Also known as talent analytics or workforce analytics.

 Involves the use of data and statistical methods to gain

insights into human resources-related processes and improve
organizational decision-making.
Specifications :
 Talent Acquisition and Recruitment
 Employee Performance and Engagement
 Workspace Planning
 Employee Retention
 Learning and Development
 Diversity and Inclusion
 Employee Well-Being
 HR Metrics and Reporting

1. Talent Acquisition and Recruitment:

Predictive Analytics: Identifying potential candidates based on

historical data, improving the accuracy of recruitment efforts.
Application Tracking System (ATS) Analytics: Tracking and
optimizing the recruitment process to reduce time-to-fill and
improve the quality of hires.

2. Employee Performance and Engagement:

Performance Analytics: Evaluating employee performance based
on key metrics, helping in performance appraisals and identifying
areas for improvement.
Employee Engagement Surveys: Analyzing survey data to
understand employee satisfaction, identify trends, and address
potential issues.
3.Workforce Planning:
Predictive Modeling: Forecasting workforce needs based on
business goals and market trends.
Succession Planning: Identifying and developing talent for future
leadership roles within the organization.

4.Employee Retention:
Churn Analytics: Identifying factors leading to employee turnover
and implementing strategies to improve retention.
Predictive Modeling: Predicting which employees are at risk of
leaving based on various factors.

5.Learning and Development:

Training Effectiveness Analytics: Assessing the impact of training
programs on employee performance.
Skill Gap Analysis: Identifying gaps in employee skills and
developing targeted training programs.

Advantages :

 Informed Decision-Making
 Talent Optimization
 Cost efficiency
 Employee Engagement
 Strategic Planning

Drawbacks :

 Data Quality and Privacy Concerns

 Bias in Decision-Making
 Resistence to Change
 Overreliance on Technology
Supply Chain Analytics:

Definition :
 Involves the use of data analysis and statistical methods
to gain insights into the various aspects of the supply
chain.
 It helps organizations optimize their supply chain
processes, enhance visibility, and make data-driven
decisions.
Specifications :

 Demand Forecasting
 Inventory Management
 Supplier Performance
 Logistics Optimization
 Cost Analytics

1.Demand Forecasting:
Predictive Analytics: Forecasting demand based on historical data,
market trends, and external factors.
Seasonal Trend Analysis: Identifying patterns in demand during
specific seasons or periods.

2.Inventory Management:
Inventory Turnover Analysis: Optimizing inventory levels to
reduce carrying costs while ensuring product availability.
Stockout Analysis: Analyzing instances of stockouts to improve
inventory forecasting and replenishment.

3.Supplier Performance:
Supplier Scorecards: Evaluating supplier performance based on
metrics such as on-time delivery, quality, and cost.
Risk Management: Identifying potential risks in the supply chain
and developing strategies to mitigate them.

4.Logistics Optimization:
Route Optimization: Analyzing transportation data to optimize
delivery routes and reduce transportation costs.
Warehouse Efficiency: Monitoring and optimizing warehouse
operations to improve overall supply chain efficiency.

5.Cost Analytics:
Cost-to-Serve Analysis: Understanding the cost associated with
serving different customers or product lines.
Total Cost of Ownership (TCO): Analyzing the full cost of
procuring and managing products throughout their lifecycle.

Advantages :

 Optimized Inventory Management

 Operational Efficiency
 Cost Reduction
 Risk Mitigation
 Strategic Decision Support
 Customer Satisfaction

Drawbacks :
 Disruption
 Resource Intensive
 Scalability Challenges
 Complex Implementation

Applications of Analytics :

 Finance
 Healthcare
 Marketing
 E-Commerce
 Manufacturing
 Human Resources

Limitations of Analytics :

 Data Quality
 Data Security and Privacy
 Model Complexity
 Bias and Fairness
 Dynamic Environment
 User resistence
 Unit – 5
Marketing Strategy, Marketing Mix, Customer Behaviour –selling Process – Sales Planning –
Analytics applications in Marketing and Sales - predictive analytics for customers' behaviour in
marketing and sales.

MARKETING STRATEGY
 Strategy: It is a carefully devised plan of action designed to achieve a specific goal or set of
goals.
 Purpose of Business Analytics for Managers: Delivering the right decision support to the right
people at the right time.
 Customer Analytics : It helps in Delivering customer - centric decision support to the right
people at the right time
 Three Types of Decision Makers:
1. Strategic Decision Makers
2. Sales and marketing decision makers
3. Operation decision makers
 Information Wheel:

 Need for effective Marketing Strategy: (Goals)

1. Need data driven based decision making
2. Fulfil customer needs
3. Competitor challenges (competitor advantages)
4. Increasing customer engagement
5. Improving conventional rates
6. Create more revenue (increase ROI)
7. Increase overall business growth
 Inputs: Relevant and quality of historic data

1
 - Industries / companies must spend time and cost for data collection, storage and analytics of
data
- Organization needs a practice for the values and prioritize the values of data in decision
making.
 Different Area of Analytics:
1. Customer Behaviour
2. Market trends
3. Performance of marketing initiatives
 Marketing Strategy helps
Input : Data and analytics tools
Output: Marketing plans and initiatives (Develop, implement and evaluate)
 Steps involved:
1. Collection of relevant business information (Date warehouse) – Accurate and timely data
2. Analyzing the data to identify trends and patterns
3. Interpretation of information
 Key Components of Marketing Strategy:
1. Data Collection and Integration
o Identification of relevant data sources (internal and external – Data warehouse)
o Seamless integration of data from various sources.
2. Customer Segmentation
3. Predictive Analytics
4. Personalization
5. Marketing Mix Optimization
6. Competitor Analysis
7. Campaign Performance Measurement
8. Customer Journey Mapping
9. Real-time Analytics
10. ROI Analysis
11. Data Security and Compliance
12. Continuous Improvement
 Different Types of Marketing Strategies:
1. Content Optimization
2. Social Media Analytics
3. Email marketing Analytics
4. Search engine Optimization analytics
5. Customer Management Relationship(CRM) analytics

2
 MARKETING MIX
 Need for Marketing Mix: It addresses the fundamental need for businesses to
strategically manage their marketing efforts to meet customer needs and achieve
organizational objectives.
 Goal:
a. Create a balanced and effective marketing strategy
b. Increase customer satisfaction
c. Brand awareness
d. Increase ROI
 It is also referred as 4P’s
 Framework: It is a framework that businesses use to plan and execute marketing
strategies.
 Flexible framework: It is a flexible framework that can be adapted to various industries
and market conditions. (Flexibility is crucial to adjust strategies based on changing
conditions.)
 External Conditions: Adopt to external conditions
a. Market trends
b. Economic factors, competition
c. Consumer behaviour
 Essential Inputs:
a. Market Research
b. Consumer Feedback
c. Data
d. Customer preferences
e. Market Trends
f. Competitive landscape
 4 Ps:
1. Product - the goods or services that a company offers to meet the needs of its
target market.
- Decision related to product design, features, branding and packaging
2. Price - determining the right price for the product or service
Factors:
a. Production cost
b. Competitor pricing
c. Perceived value

3
 d. Target market's willingness to pay.
3. Place (distribution) - channels through which the product or service is made
available to customers.
Factors:
a. Selecting appropriate distribution channels
b. Managing inventory
c. Ensuring that the product is accessible to the target audience
4. Promotion - The methods used to communicate and promote the product or
service to the target market
Methods:
a. Advertising
b. Public relations
c. Sales promotions
d. Personal selling
e. Other promotional activities
 Modern Marketing Management Four Ps
a. People
b. Processes
c. Programs
d. Performance
 Extended Marketing Mix (Variation in 4 Ps):
1. Product
2. Price
3. Place
4. Promotion
5. People - The employees and customer service representatives who directly or
indirectly impact the customer experience
6. Process – It encompasses the procedures, mechanisms, and systems involved in
delivering the product or service, ensuring a smooth and efficient customer
experience.
7. Physical evidence – It involves the tangible elements that contribute to the
overall customer experience.
Parameters:
a. Physical environment
b. Packaging
c. Any other sensory aspects.

4
  Steps to create marketing mix for overall business success:
1. Research - Understand the market, competitors, and customer needs.
2. Define Objectives - Clearly outline what the marketing mix aims to achieve.
3. Develop the Product - Create a product that meets customer needs and differentiates
from competitors.
4. Set the Price - Determine a pricing strategy that aligns with market expectations and
business goals.
5. Choose Distribution Channels - Decide how and where the product will be
available to customers.
6. Promote the Product - Develop and execute marketing and promotional strategies.
7. Monitor and Adjust - Regularly assess the effectiveness of the marketing mix and
make adjustments based on feedback and changing market conditions.

CUSTOMER BEHAVIOR
 Need for Customer behaviour Analysis: It helps to meet the needs and expectations of
their target audience.
 It helps in (align with business strategies)
a. Tailoring products
b. Services
c. Marketing strategies
 Goal: It enhances the customer satisfaction and loyalty
 Definition: It refers to the actions, decisions, and patterns of consumers as they search
for, purchase, use, and evaluate products and services.
 Role of Business Analytics in understanding customer Bahaviour:
1. Data collection and Utilization
2. Data Analysis and pattern recognition
3. Segmentation
4. Predictive Analytics
5. Personalization
6. Performance Measurement
7. Feedback Analysis
 Input: Information collected from various sources (Customer feedback, surveys, market
research, data analytics)
 Output:
1. Customer Preferences
2. Buying patterns

5
 3. Factors influencing decisions
 The customer behaviour analysis helps in strategic decisions related to product
development, marketing, and customer engagement.
 Three types of Consumer Behaviour:
1. Purchase behaviour (Demand, retention, recall)
2. Usage behaviour (Attrition)
3. Consumer Loyalty (post – purchase behaviour, pre – purchase behavior)
 Different Mythologies:
a. Surveys and Questionnaires - Gathering direct feedback from customers.
b. Observational Studies - Analysing customer behaviour in real-world settings.
c. Interviews - Conducting one-on-one conversations to understand motivations and
preferences.
d. Data Analysis - Utilizing data from customer interactions, transactions, and online
behaviour.
 Elements of Customer Behaviour:
a. Cultural Factors - Values, beliefs, and customs influencing behaviour.
b. Social Factors - Family, reference groups, and social roles impacting decisions.
c. Personal Factors - Age, lifestyle, occupation, and personality traits.
d. Psychological Factors - Motivations, perceptions, and learning affecting choices.
 Steps involved for effective customer Behaviour Analysis:
a. Define Objectives
b. Select Research Methods
c. Collect Data
d. Analyse Data
e. Draw conclusions
f. Implement changes
g. Monitor and Iterate
 Technologies integrated with Business Intelligence:
1. Dashboard and reporting
2. Data Visualization
3. Data warehousing
4. Strategic decision support
5. Dynamic Querying

6
 ANALYTICS APPLICATIONS IN MARKETING AND SALES
 Need for Analytics applications:
1. Informed Decisions
2. Optimized strategies
3. Enhance overall performance
 Goal: Improve
1. Optimize marketing ROI
2. Enhancing customer engagement
3. Revenue growth
 It plays a crucial role in enhancing the effectiveness of marketing and sales strategies.
 Input:
1. Customer demographics
2. Purchase history
3. Website interactions
4. Social media engagement
5. Marketing campaign performance metrics
 Output: insights into visualization
Reports:
1. Customer segment reports
2. Sales forecasts
3. Campaign performance dashboards
4. Recommendations for strategy optimization
 Customer Lifetime value (CLV):
CLV= (Average Purchase Value × Purchase Frequency) / Churn Rate
ChurnRate = (Number of Customers Lost during a Period / Total Number of Customers at
the Start of the Period) * 100
Different Analytics:
1. Descriptive Analytics (summarizing historical data)
2. Diagnostic Analytics (identifying trends and patterns)
3. Predictive analysis (forecasting)
4. Prescriptive analytics (recommendations)
 Algorithms:
1. Regression analysis for sales forecasting
2. Clustering algorithms for customer segmentation
3. Classification algorithms for lead scoring

7
 4. Natural Language Processing (NLP) for sentiment analysis in social media
5. Machine Learning Models
 Types:
1. Web analytics for online activities
2. Social media analytics for monitoring social platforms
3. Customer analytics for understanding behaviour
4. Sales analytics for tracking performance and forecasts
5. Marketing analytics for evaluating campaign effectiveness
 Various Tools available:
1. Microsoft Power BI
2. Tableau (Data Visualization tools)
3. Elastic Search (powerful search and analytics engine)
4. Google cloud data flow
5. Spark Streaming
6. Google Analytics Real time
7. Apache Storm
8. Amazon Kinesis

PREDICTIVE ANALYTICS FOR CUSTOMERS'

BEHAVIOUR IN MARKETING AND SALES
 Goal: It helps to forecast the customer behaviour based on historical data and patterns. It
enables
a. Informed Decisions
b. Personalize marketing campaigns
c. Identify potential leads
d. Increase the efficiency of marketing and sales effort
 Statistical Analysis: It is not suitable in customer behaviour analysis
 Input:
a. Historical Data
b. Demographic information (Age, location, income and other demographic data)
c. External factors (economic conditions, industry trends, seasonality)
 Output:
a. Predictive Scores
b. Segmentation
c. Recommendations

8
 Types of Analytics:
a. Descriptive
b. Predictive
c. Prescriptive

 Predictive Analysis:

 Predictive Analytics involves in predicting future customer behaviour based on historical

data.
 Predictive Analysis in Marketting:

 Algorithms used:
1. Regression Analysis
2. Classification Algorithm
3. Clustering Algorithms
4. Machine Learning Models
5. NLP
 Stages of predictive marketing analytics implementation:
1. Project definition

9
 2. Data Collection
3. Data processing
4. Modelling
5. Interpretation
 Steps involved for effective implementation predictive analysis:
1. Data collection
2. Data cleaning and preparation
3. Feature Selection (identify the most relevant features for prediction of customer
behaviour)
4. Model Selection
5. Training the model
6. Model evaluation and validation
7. Implementation and integration
8. Monitoring and optimization
9. Actionable insights
 Use of predictive analytics in marketing:
1. Segmentation
2. Forecasting
3. Demand pricing
4. Improve customer satisfaction

10