Business Analytics
Lecture 1 — Introduction to Business Analytics
Dr. Muhammad Masroor Ali
Professor
Department of Computer Science and Engineering
Bangladesh University of Engineering and Technology
Batch - 31
Version: 1.1, Last modified: July 7, 2025
What is Business Analytics?
According to Wikipedia:
“Business analytics refers to the skills, technologies,
practices for continuous iterative exploration and
investigation of past business performance to gain
insight and drive business planning.”
In short, BA is a rational, fact-based approach to
decision-making.
Facts come from data.
Thus, BA is about turning data into decisions.
Dr. Muhammad Masroor Ali Business Analytics 2 / 30
Components of Business Analytics
The Science:
Statistics
Artificial intelligence (data mining, machine learning)
Optimization
The Skills:
Computer skills
Communication skills
Project and change management
Dr. Muhammad Masroor Ali Business Analytics 3 / 30
BA is Not a Science
BA by itself is not a science.
BA is the total set of knowledge required to solve business
problems rationally.
Experience in BA projects and knowledge of relevant
business areas (e.g., healthcare, advertising, finance) are
critical.
Dr. Muhammad Masroor Ali Business Analytics 4 / 30
Subdivisions of Business Analytics
Descriptive Analytics:
Analyzes data and identifies patterns.
Predictive Analytics:
Predicts what is likely to happen in the future based on
current insights.
Prescriptive Analytics:
Determines alternative decisions to change the
situation and achieve desirable outcomes.
Dr. Muhammad Masroor Ali Business Analytics 5 / 30
Example 1.1: A Hotel Chain’s Use of Analytics
Analyze reservations to identify patterns:
Busiest days of the week.
Impact of city events.
Seasonal trends.
Use insights to predict upcoming revenue.
Adjust room pricing (e.g., during sports events) to
maximize revenue.
Dr. Muhammad Masroor Ali Business Analytics 6 / 30
Example 1.1: A Hotel Chain’s Use of Analytics
Analyze reservations to identify patterns:
Busiest days of the week.
Impact of city events.
Seasonal trends.
Use insights to predict upcoming revenue.
Adjust room pricing (e.g., during sports events) to
maximize revenue.
Hands on Code.
Dr. Muhammad Masroor Ali Business Analytics 6 / 30
Data as the Foundation of Analytics
Analytics begins with data.
Centralized data warehouses:
Store relevant current and historical data.
Used for reporting and analytics.
Not all companies have a centralized database.
Even when available, it often lacks all necessary
information.
Dr. Muhammad Masroor Ali Business Analytics 7 / 30
Data Pre-Processing in Analytics
Data often needs to be:
Collected.
Cleansed.
Combined with other sources.
Pre-processing is a time-consuming task:
Often takes more time than the analysis itself.
Dr. Muhammad Masroor Ali Business Analytics 8 / 30
Example 1.2: Challenges in Data Collection for
Analytics
In the hotel revenue management example:
Historical data on reservations is needed.
Data on historical and future events in the hotel
surroundings is essential.
Challenges in obtaining data:
Reservation data may be stored at an aggregated
level.
Changes in IT systems may have overridden older
data.
No centrally available list of events in the area.
Dr. Muhammad Masroor Ali Business Analytics 9 / 30
Example 1.2: Challenges in Data Collection for
Analytics
In the hotel revenue management example:
Historical data on reservations is needed.
Data on historical and future events in the hotel
surroundings is essential.
Challenges in obtaining data:
Reservation data may be stored at an aggregated
level.
Changes in IT systems may have overridden older
data.
No centrally available list of events in the area.
Hands on Code.
Dr. Muhammad Masroor Ali Business Analytics 9 / 30
Data Assumptions vs. Reality
Many organizations assume they already have all the
required data.
Practical challenges emerge when:
The data scientist asks for reservation data combined
with booking dates.
An event list from the surrounding area is needed.
Often, the organization realizes that critical data is missing.
Dr. Muhammad Masroor Ali Business Analytics 10 / 30
Full Business Analytics Project Lifecycle
Figure: Data Science Steps in the Full BA Project Lifecycle
Dr. Muhammad Masroor Ali Business Analytics 11 / 30
Full Business Analytics Project Lifecycle
Figure: Data Science Steps in the Full BA Project Lifecycle
The lifecycle includes:
Data collection and pre-processing
Descriptive, predictive, and prescriptive analytics
Implementation of insights
Each step builds toward actionable and optimized
decisions.
Dr. Muhammad Masroor Ali Business Analytics 11 / 30
Skills for a Successful Business Analyst
Figure: Core Skills for Business Analytics Projects
Key skills:
Business intelligence
Cleaning and feature engineering
Data analysis and visualization
Statistics and machine learning
Optimization and reinforcement learning
Soft skills
A combination of technical and interpersonal skills is
essential for success.
Dr. Muhammad Masroor Ali Business Analytics 12 / 30
Example 1.3: Optimizing Debt Collection
A debt collection agency aims to optimize its resource
utilization, primarily calls to debtors.
Data sources include:
Internal data: Payment history.
External data: Household composition and
neighborhood characteristics.
Steps involved:
Data analysis and visualization.
Selecting a predictive method.
Dr. Muhammad Masroor Ali Business Analytics 13 / 30
Example 1.3: Optimizing Debt Collection
A debt collection agency aims to optimize its resource
utilization, primarily calls to debtors.
Data sources include:
Internal data: Payment history.
External data: Household composition and
neighborhood characteristics.
Steps involved:
Data analysis and visualization.
Selecting a predictive method.
Hands on Code.
Dr. Muhammad Masroor Ali Business Analytics 13 / 30
Predictive and Prescriptive Analytics in Debt Collection
Predictive analytics step:
Predicts the probability that a debtor will pay off their
debt.
Based on characteristics of:
The debt itself.
Actions taken by the agency.
Prescriptive analytics step:
Determines the best action for each debtor to
maximize the likelihood of repayment.
Dr. Muhammad Masroor Ali Business Analytics 14 / 30
Example 1.4: Revisiting Example 1.2
Focus on revenue management for hotels.
Influence of events and intra-week fluctuations on
reservations studied in the descriptive step.
Dr. Muhammad Masroor Ali Business Analytics 15 / 30
Example 1.4: Revisiting Example 1.2
Focus on revenue management for hotels.
Influence of events and intra-week fluctuations on
reservations studied in the descriptive step.
Hands on Code.
Dr. Muhammad Masroor Ali Business Analytics 15 / 30
Key Insights
Integration of descriptive, predictive, and prescriptive
analytics creates a robust revenue management strategy.
Optimized pricing adapts dynamically to factors such as
events and seasonal demand.
Forecasting and optimization tools can significantly
enhance decision-making for pricing and revenue planning.
Dr. Muhammad Masroor Ali Business Analytics 16 / 30
Predictive Step: Demand Forecasting
Key Insights:
Demand per price class is forecasted based on historical
data.
Forecasts consider:
Event proximity.
Day of the week trends.
Dr. Muhammad Masroor Ali Business Analytics 17 / 30
Optimization for Revenue Maximization
Optimization Algorithm:
Uses demand forecasts as input.
Determines daily prices to maximize total revenue.
Objective:
Dynamic pricing strategy to enhance profitability.
Dr. Muhammad Masroor Ali Business Analytics 18 / 30
Step 1: Data Generation
Scenario Selection:
Choose a relevant domain (e.g., sales, healthcare, or
operations).
Example: Predicting product sales or managing resources.
Data Identification:
Identify key metrics (e.g., daily sales, patient visits).
Synthetic Data:
Simulate data if real data is unavailable or incomplete.
Dr. Muhammad Masroor Ali Business Analytics 19 / 30
Step 1: Descriptive Analysis
Analyze Patterns:
Group data by key indicators (e.g., weekdays, events).
Calculate averages or totals.
Visualization:
Use plots (e.g., boxplots, line charts) to identify trends and
outliers.
Dr. Muhammad Masroor Ali Business Analytics 20 / 30
Step 2: Predictive Analysis
Feature Engineering:
Create variables like time-based trends or event indicators.
Model Training:
Select a model (e.g., linear regression, decision trees).
Train using historical data and evaluate performance.
Forecasting:
Predict future values for your target variable (e.g.,
bookings, sales).
Dr. Muhammad Masroor Ali Business Analytics 21 / 30
Step 3: Prescriptive Analysis
Optimization Objective:
Define clear goals (e.g., maximize revenue, minimize cost).
Constraints:
Specify limits (e.g., price range, resource capacity).
Optimization Techniques:
Apply methods like linear programming or gradient-based
optimization.
Dr. Muhammad Masroor Ali Business Analytics 22 / 30
Step 4: Results and Reflection
Present Findings:
Summarize insights using visuals (e.g., charts, tables).
Missing Data:
Apply imputation strategies (e.g., mean fill, forward fill).
Address potential biases due to missing data.
Evaluate Results:
Reflect on alignment with domain-specific challenges and
objectives.
Dr. Muhammad Masroor Ali Business Analytics 23 / 30
Discussion Questions
Consider the following:
How does your scenario align with the workflow?
What data challenges did you face?
What imputation techniques did you apply for missing
data?
Did the predictive model meet expectations?
Were the optimization results actionable?
Dr. Muhammad Masroor Ali Business Analytics 24 / 30
Conclusion
Key Takeaways:
Use descriptive analysis to understand trends and
patterns.
Leverage predictive modeling to forecast future demand.
Apply prescriptive techniques to optimize decisions.
Address data challenges with appropriate imputation
methods.
Dr. Muhammad Masroor Ali Business Analytics 25 / 30
Data Science
Definition:
An interdisciplinary field focused on extracting knowledge
from data.
Combines data mining, statistics, and related fields.
Popularity:
Harvard Business Review called data scientist the “sexiest
job of the 21st century.”
High demand for data scientists in industries.
Comparison with Business Analytics:
Data Scientists: Focus on methods and insights from
data.
Business Analysts: Aim for improved business
performance.
BA includes optimization and soft skills for decision
implementation.
Dr. Muhammad Masroor Ali Business Analytics 26 / 30
Big Data: Overview
Definition:
Differentiates itself by the 3 V’s:
1 Volume: Large datasets exceeding traditional storage.
2 Variety: Unstructured data (e.g., free text).
3 Velocity: Data available in real-time.
Requirements:
Adapted storage systems.
Specialized analysis techniques.
Dr. Muhammad Masroor Ali Business Analytics 27 / 30
Big Data in Practice
Current Trends:
Rapid data collection due to connected devices and
sensors (Internet of Things).
Data volume doubles approximately every three years.
Relevance to BA:
Most BA projects use smaller, structured datasets.
Big data may serve as the origin of extracted, relevant
information.
Dr. Muhammad Masroor Ali Business Analytics 28 / 30
Key Points
Data science focuses on extracting insights from data
using interdisciplinary methods.
Big data is characterized by its volume, variety, and
velocity, requiring specialized tools.
Business analytics primarily aims at improving business
performance using structured data.
Big data often serves as a source for BA but is not always
directly involved.
Dr. Muhammad Masroor Ali Business Analytics 29 / 30
End of
Slides
Dr. Muhammad Masroor Ali Business Analytics 30 / 30