SCHOOL OF DATA SCIENCE AND FORECASTING

DEVI AHILYA VISHWAVIDYALAYA

INDORE (M.P)

A PROJECT REPORT
ON
GLOBAL STREAMING INDUSTRY ANALYTICS
In par al fulfillment for the award of the degree of

MASTER OF BUSINESS ADMINISTRATION

in
BUSINESS ANALYTICS
2024-2026

Submi ed by: Guided by:

Dinkar Singh Mrs. Farheen Mam
Enrollment No. - DX2401335
Roll No. – DS5A-2412

1|Page
 SCHOOL OF DATA SCIENCE AND FORECASTING
DEVI AHILYA VISHWAVIDYALAYA
INDORE (M.P)

STATEMENT OF ORIGINALITY

In accordance with the requirements for the Degree of Master of Business

Administra on in BUSINESS ANALYTICS, in SCHOOL OF DATA SCIENCE AND
FORECASTING, I present this report en tled “GLOBAL STREAMING INDUSTRY
ANALYTICS”. This report is completed under the Supervision of Mrs Farheen Mam
(External Supervisor) and Dr. Vandit Hedau (Internal Supervisor).

I declare that the work presented in the report is my/our own work except as
acknowledged in the text and footnotes, and that to my knowledge this material has
not been submitted either in whole or in part, for any other degree at this University or
at any other such Institution.

Signature:
Name: Dinkar Singh
Date: 1-05-2026

2|Page
 SCHOOL OF DATA SCIENCE AND FORECASTING
DEVI AHILYA VISHWAVIDYALAYA
INDORE (M.P)

RECOMMENDATION

This disserta on en tled GLOBAL STREAMING INDUSTRY ANALYTICS submi ed by

DINKAR SINGH towards the par al fulfillment of Degree of Master of Business
Administra on in Business Analy cs of Devi Ahilya Vishwavidyalaya, Indore is a
sa sfactory account of her project work and is recommended for the award of
degree.

INTERNAL SUPERVISOR HEAD OF DEPARTMENT

Dr. Vandit Hedau Dr. Sanjiv Tokekar

3|Page
 SCHOOL OF DATA SCIENCE AND FORECASTING
DEVI AHILYA VISHWAVIDYALAYA
INDORE (M.P)

CERTIFICATE

This is to certify that the dissertation entitled “GLOBAL STREAMING

INDUSTRY ANALYTICS” submitted by DINKAR SINGH is approved for the
award of Master of Business Administration in Business Analytics.

INTERNAL SUPERVISOR HEAD OF DEPARTMENT

Dr. Vandit Hedau Dr. Sanjiv Tokekar

4|Page
 AKNOWLEDGEMENT

The successful completion of any endeavour is seldom the result of solitary effort. It is
a culmination of the dedicated contributions of numerous individuals, each bringing
their unique significance to the table. In this spirit of gratitude, I extend my heartfelt
thanks to all those who, through their implicit or explicit support, have played a
pivotal role in the realization of our project objectives.

I extend my deepest gratitude to our esteemed mentor, Dr. Sanjiv Tokekar (Head of
the Department at the School of Data Science and Forecasting, DAVV, Indore). His
unwavering dedication, invaluable insights, and motivating guidance have been the
cornerstone of our journey. It is with profound appreciation that I acknowledge his
pivotal role in shaping our endeavours towards excellence.

I am sincerely grateful to our guiding light, Mrs Farheen Mam ( Professor at the
School of Data Science and Forecasting, DAVV, Indore). His invaluable insights,
motivating suggestions, and unwavering guidance have been instrumental in steering
this research towards fruition. I am deeply grateful for his constant encouragement and
invaluable counsel throughout this journey.

In every facet of life, positive inspiration and appropriate guidance are indispensable,
particularly in academic pursuits. I acknowledge with deep gratitude all those
individuals who have generously provided me with their continuous guidance,
insightful suggestions, and valuable information. Their assistance, whether direct or
indirect, has been immensely valuable to me, and for that, I am sincerely thankful.

With hear elt apprecia on,

Dinkar Singh

5|Page
 ABSTRACT

The rapid expansion of Over-the-Top (OTT) platforms has generated massive datasets,
yet translating this data into actionable market strategy remains a significant challenge
for business analysts. This project, titled "Global Streaming Industry Analytics,"
presents a comprehensive data pipeline designed to evaluate content performance and
market expansion opportunities. Using MySQL for robust data storage and Python
for cleaning "messy" real-world data, the study identifies key drivers of audience
engagement.

The primary deliverable is a two-page strategic dashboard developed in Power BI.

Page one provides a high-level executive overview of global market presence, while
page two utilizes a Market Opportunity Matrix (Heatmap) and ROI Analysis to
pinpoint high-quality, low-volume regions for potential growth. The methodology
adheres to the System Development and Design category, focusing on the practical
application of business analytics tools to solve real-world industry problems.

6|Page
 TABLE OF CONTENTS

Chapter 1 – Introduction
1.1 General
1.2 Problem Statement
1.3 Project Objec ves
1.4 Scope of the project

Chapter 2 – Literature Review

2.1 Introduction to OTT Analytics
2.2 The role of Metadata in Content Performance
2.3 Market Gap Analysis and Visualization
2.4 Technology Stack Review

Chapter 3 – Research Methodology

3.1 Research Framework
3.2 Data Pipeline Architecture
3.3 Data Modeling (Star Schema)

Chapter 4 – Data Collection, Pre-Processing and Analysis

4.1 Advance Data Ingestion and ETL Framework
4.2 Data Sanitization and “Success Score” Engineering
4.3 Distribution and Correlation Analysis
4.4 Relational Schema Design (The Tittle Table)
4.5 Analytical Benchmarking Tables

Chapter 5 – Strategic Results and Business Discussion

5.1 Budget vs. Quality
5.2 Platform Content Strategy
5.3 The “Global Opportunity” Gap Analysis
5.4 Quality vs. Quantity Benchmarking
5.5 Efficiency and ROI (Return on Investment) Analysis
5.6 Executive Strategic Dashboard (Page 1 Analysis)
5.7 Market opportunity and Portfolio Deep-Drive (page 2 Analysis)

Chapter 6 – Conclusion
6.1 Summary of the System Development
6.2 limitation and Ethical Consideration
6.3 Future Scope for Enhancement

Chapter 7 - References

7|Page
 LIST OF FIGURES

Figure 1: Star Schema: Power BI Model View

Figure 2: Python ETL Ingestion & SQL Connection Code

Figure 3: Data Cleaning & Success Score Engineering Code

Figure 4: Global IMDb Rating Distribution Histogram

Figure 5: Metric Correlation Heatmap

Figure 6: SQL Schema: titles Fact Table

Figure 7: SQL Schema: Yearly Trend Dimension

Figure 8: SQL Schema: Country Benchmark Dimension

Figure 9: SQL Schema: Genre Benchmark Dimension

Figure 10: SQL Schema: Platform Benchmark Dimension

Figure 11: Regression Analysis: Budget vs. IMDb Rating

Figure 12: Platform Content Strategy: Movies vs. TV Shows

Figure 13: Global Opportunity Gap Analysis Result Set

Figure 14: Quality vs. Quantity Benchmark Result Set

Figure 15: ROI Efficiency Ratio Analysis Result Set

Figure 16: Strategy Dashboard: Executive Overview (Page 1)

Figure 17: Strategy Dashboard: Strategic Deep-Dive (Page 2)

8|Page
 CHAPTER 1 – INTRODUCTION

1.1 General
In the current digital era, the streaming industry has shifted from a "content is king" model to a
"data-driven content" model. Platforms like Netflix, Amazon Prime, and Disney+ compete in a
saturated market where production costs are rising. Understanding the relationship between critical
reception (IMDb ratings) and production budgets is essential for maintaining a sustainable Return on
Investment (ROI).

1.2 Problem Statement

Streaming platforms often face the "content paradox"—investing millions in high-budget
productions that fail to resonate with local audiences, or missing out on "sleeper hits" in untapped
markets. There is a lack of integrated tools that allow decision-makers to visualize market saturation
alongside content quality simultaneously.

1.3 Project Objectives

The main objectives of this study are:

 Data Integration: To create a seamless pipeline connecting SQL databases to Python for
cleaning and Power BI for visualization.
 Performance Benchmarking: To identify the "Efficiency Frontier" where content achieves
high IMDb ratings with minimal production spend.
 Geographic Gap Analysis: To develop a Market Opportunity Matrix that highlights
countries with high audience sentiment but low existing content volume.
 Portfolio Visualization: To visualize platform-specific distribution strategies using advanced
Treemap analytics

1.4 Scope of the Project

This project covers global streaming data from 2010 to 2026. It focuses specifically on the top 5
global platforms and utilizes quantitative metrics like IMDb scores, budgets (in millions USD), and
release years to drive its conclusions.

9|Page
 CHAPTER 2 – LITERATURE REVIEW

2.1 Introduction to OTT Analytics

The Over-the-Top (OTT) industry has witnessed a paradigm shift from traditional broadcasting to
data-driven content delivery. According to recent industry reports, the global streaming market is
projected to grow significantly by 2026, driven by personalization and recommendation engines. The
core of this growth lies in "Data Science and Analytics," where user behavior, content ratings, and
financial metrics are synthesized to drive business decisions

2.2 The Role of Metadata in Content Performance

Literature suggests that metadata, such as IMDb ratings and genre classifications, serves as a primary
indicator of "Content Quality." Research in the field of Business Analytics often utilizes these ratings
as a proxy for consumer satisfaction. Analyzing the correlation between these ratings and production
budgets helps in identifying the "Efficiency Frontier" of content production.

2.3 Market Gap Analysis and Visualization

Data visualization is an essential skill highlighted in the DAVV curriculum. Previous studies have
shown that heatmaps and matrices are the most effective ways to visualize "Market Saturation." By
plotting quality (average rating) against supply (total titles), analysts can identify regions where a
"Supply-Demand Gap" exists. This project builds upon these concepts by implementing a dynamic
Market Opportunity Matrix in Power BI

2.4 Technology Stack Review

 SQL (MySQL): Used for managing relational data and complex querying.
 Python: Preferred for handling "messy data," performing data cleaning, and statistical
analysis using libraries like Pandas and Seaborn.
 Power BI: A cutting-edge tool for creating interactive dashboards that allow for real-time
business context exploration.

10 | P a g e
 CHAPTER 3 - RESEARCH METHODOLOGY

3.1 Research Framework

This project adopts a System Development and Design methodology. The framework is structured
as a multi-stage data pipeline that transforms raw streaming metadata into strategic business insights.
The methodology focuses on three core pillars: Data Integrity, Relational Modeling, and Visual
Analytics.

3.2 Data Pipeline Architecture

The technical workflow follows a "Medallion" approach:

1. Bronze Layer (Ingestion): Raw data is extracted from the MySQL relational database using
optimized SQL queries to handle large-scale streaming catalogs.
2. Silver Layer (Transformation): Python (Pandas) is utilized to clean "messy" data. This
stage involves handling null values in production budgets and correcting data type
inconsistencies where numeric ratings were stored as strings.
3. Gold Layer (Visualization): The refined dataset is loaded into Power BI for final
dashboarding and strategic KPI tracking.

3.3 Data Modeling (Star Schema)

A robust data model was developed to ensure referential integrity, which is a key requirement for
graduate-level analytics projects. The model follows a Star Schema architecture:

 Fact Table: Main_Titles (Contains metrics like IMDb ratings, budgets, and show IDs).
 Dimension Tables: Geography_Lookup (Country data), Platform_Lookup (Service
providers), and Time_Dim (Release years).
 Relationships: Established One-to-Many relationships to allow for seamless cross-filtering
across the dashboard.

11 | P a g e
 CHAPTER 4 - DATA COLLECTION, PRE-PROCESSING AND
ANALYSIS

4.1 Advanced Data Ingestion and ETL Framework

The first stage of the project involved building a robust ETL (Extract, Transform, Load) pipeline.
As shown in the technical snapshots, Python was utilized as the primary processing engine to bridge
the gap between flat CSV files and a centralized MySQL database. The script utilizes the sqlalchemy
library to create a secure connection engine to the local server. A programmatic loop was implemented
to iterate through multiple source files—including titles, country benchmarks, and yearly trends—
ensuring that all data was normalized and uploaded with 100% integrity into the SQL environment.

4.2 Data Sanitization and "Success Score" Engineering

Handling "messy data" is a core requirement of this curriculum. The provided Python cleaning script
demonstrates high-level data manipulation:

 Temporal Normalization: Converting the date_added column using pd.to_datetime

ensures that all subsequent time-series analysis is chronologically accurate.
 Imputation Logic: Rather than deleting records with missing financial data, a Group-wise
Median Imputation was used. Budgets were filled based on the median of their specific
genre, maintaining the statistical distribution of the dataset.

12 | P a g e
  Feature Engineering (Success Score) I authored a custom business metric called the
success_score. This identifies high-performing assets by multiplying critical reception
(IMDb rating) with audience engagement (Hours Watched) and dividing by the production
budget. This represents a true ROI (Return on Investment) metric essential for MBA-level
decision support.

4.3 Distribution and Correlation Analysis

Before moving to visualization, the data underwent rigorous statistical validation:

 Rating Distribution: The histogram shows that IMDb ratings follow a near-normal
distribution centered around 6.8. This confirms that the dataset is balanced and not skewed by
extreme outliers.

13 | P a g e
  Performance Heatmap: I u lized a Correla on Heatmap to iden fy rela onships between
variables. The analysis revealed a strong posi ve correla on between imdb_votes and
hours_watched, sugges ng that cri cal popularity is a reliable leading indicator for mass-
market consump on.

14 | P a g e
4.4 Relational Schema Design (The 'Titles' Table)

The core of the system is the titles table, designed as a comprehensive repository for streaming
metadata. As shown in the CREATE TABLE script, I implemented a variety of data types to ensure
storage efficiency and data integrity:

 show_id: Defined as the PRIMARY KEY to ensure every record is unique and indexed for
fast retrieval.
 Quantitative Metrics: Columns like imdb_rating, budget_million_usd, and
hours_watched_million utilize FLOAT and INT types, allowing for the mathematical
aggregations required for ROI analysis.
 Categorical Context: Fields for type, platform, and primary_genre allow for the
"Dimensional Slicing" used in the final Power BI dashboard.

15 | P a g e
4.5 Analytical Benchmarking Tables
To support real-time executive reporting, I developed specialized benchmark tables including
yearly_trends, country_benchmarks, genre_benchmarks, and platform_benchmarks.

 Data Aggregation: These tables serve as "Aggregated Views." By pre-calculating values like
avg_imdb and total_titles at the SQL level, we significantly reduce the processing load
on the visualization engine, ensuring the dashboard remains responsive even as the dataset
grows.
 Competitive Intelligence: The platform_benchmarks table specifically tracks total awards
won and hours watched, providing a 360-degree view of competitor performance.

16 | P a g e
17 | P a g e
 CHAPTER 5: STRATEGIC RESULTS AND BUSINESS DISCUSSION

5.1 Budget vs. Quality:

The Regression Analysis A critical question in the streaming business is: "Does a higher budget
guarantee a higher rating?" To answer this, I implemented a Regression Plot (regplot). The scatter
plot reveals a surprisingly flat red regression line, indicating a very weak correlation between budget
and IMDb rating. From a managerial perspective, this is a profound finding—it suggests that "Creative
Direction" and "Storytelling" are more important for critical success than pure financial expenditure,
allowing smaller platforms to compete effectively with high-budget giants.

5.2 Platform Content Strategy

The stacked bar chart titled "Platform Strategy: Movies vs. TV Shows Volume" provides a direct
comparison of the business models used by different platforms.

 Volume Analysis: We can clearly see which platforms are "Volume Leaders" (high number
of total titles) versus "Niche Players."
 Diversification: The color segmentation allows us to identify platforms that are heavily
invested in Series (TV Shows), which drive long-term subscriber retention (stickiness),
versus those focused on Movies, which drive immediate viral marketing and new sign-ups

18 | P a g e
5.3 The "Global Opportunity" Gap Analysis
One of the most strategic queries developed in this project is the Global Opportunity Gap
Analysis. As seen in the SQL script:

 The Logic: This query filters for countries with fewer than 500 titles but an average IMDb
rating above 6.8.
 Business Impact: This identifies "High-Engagement, Low-Saturation" markets. For a
platform manager, these results represent the "Next Growth Frontier"—regions where the
audience is clearly satisfied with existing content, but there is a significant lack of volume.
This is a prescriptive insight for market-entry strategy.

19 | P a g e
5.4 Quality vs. Quantity Benchmarking
The results of the Platform Quality vs. Quantity query reveal fascinating market dynamics.

 Finding: As shown in the result set, Apple TV+ leads in average rating (7.12), despite
having a much lower title count than Netflix (4,502 titles).
 Strategic Interpretation: This confirms a "Boutique vs. Superstore" strategy. Apple TV+
focuses on high-quality, high-budget "Anchor Content" to win awards (270), whereas Netflix
focuses on "Market Ubiquity" to maximize global subscriber reach.

5.5 Efficiency and ROI (Return on Investment) Analysis

The final technical milestone was the Efficiency Query, which calculates the efficiency_ratio
($Hours Watched \div Budget$).

 The "Documentary" Phenomenon: The query results show that the Documentary genre
has the highest efficiency ratio (13.38) due to its extremely low average budget ($3.64M)
compared to its high viewership ($48.67M).
 Managerial Recommendation: This suggests that platforms should diversify their portfolio
with non-scripted content to balance the high financial risks of $20M+ big-budget Dramas
and Crime thrillers.

20 | P a g e
5.6 Executive Strategic Dashboard (Page 1 Analysis)

The first page of the system is designed for "C-Suite" decision-makers, offering a comprehensive
overview of the global streaming landscape as of 2026.

 Strategic KPIs: The top ribbon features four critical metrics. The Total Content Catalog
(15.00K) highlights the scale of the data processed. The Global Quality Benchmark (6.81)
provides a baseline for evaluating all new acquisitions. Most importantly, the Capital
Investment ($0.24M) card provides the financial context necessary for ROI discussions.
 Market Share Dynamics: The doughnut chart illustrates a highly competitive "Blue Ocean"
market. While leaders like Netflix and Amazon Prime Video dominate volume, the presence
of niche players like Crunchyroll and JioCinema indicates successful market segmentation.
 Interactive Slicers: On the far left, I implemented high-resolution slicers for Release
Timeline and Content Type. This enables "Temporal Analysis," allowing a manager to see
how market share shifted from the 1980s to the present day—a feature that transforms static
data into a dynamic history of industry growth.

21 | P a g e
5.7 Market Opportunity & Portfolio Deep-Dive (Page 2 Analysis)
The second page shifts from "Overview" to "Actionable Intelligence," focusing on where the
platform should invest next.

 The Expansion Matrix (Prescriptive Analytics): This matrix is the crown jewel of the
analysis. It utilizes conditional formatting to highlight Denmark and Brazil as high-
engagement, low-supply markets. For an MBA-level project, this demonstrates the ability to
use data for "Market Gap Analysis".
 Portfolio Tree Map (Competitive Strategy): This hierarchical visual provides a unique
look at "Content Diversification." For example, the visual clearly shows that Netflix
maintains a significant lead in Movies (2.66K) compared to TV Shows (1.84K), whereas
Hulu and HBO Max show a tighter ratio. This informs us about the different "Subscriber
Stickiness" strategies employed by each competitor.
 Strategic ROI Visualization: The scatter plot (Top Right) maps Production Budget against
Critical Reception. By plotting the "Efficiency of Spend," we can identify which platforms
are achieving the highest IMDb ratings per dollar invested, a core competency in Business
Analytics.

22 | P a g e
23 | P a g e
 CHAPTER 6: CONCLUSION

6.1 Summary of the System Development

This major project successfully designed and implemented a business analytics system tailored for
the OTT industry. By integrating SQL, Python, and Power BI, the project demonstrated that
complex, "messy" data can be transformed into a clear strategic roadmap.

6.2 Limitations and Ethical Considerations

In accordance with DAVV research ethics, it is noted that IMDb ratings are subjective and subject to
"review bombing" or bias. Furthermore, production budgets are often estimates and may not reflect
the total marketing spend. These limitations suggest that while the dashboard is a powerful tool, it
should be used alongside qualitative market research.

6.3 Future Scope for Enhancement

The current system can be enhanced by integrating Predictive Analytics. Future work could involve
building a Machine Learning model in Python to predict the success of a script before production
begins. Additionally, integrating "Social Listening" data from platforms like X (formerly Twitter)
would allow for real-time sentiment tracking to complement the static IMDb ratings.

24 | P a g e
 CHAPTER 7 – REFERENCES

 Amazon Web Services (2024). 'Understanding OTT: Evolution of Streaming Video Technology'.
Available at: [Link]

 IMDb (2026). 'IMDb Datasets for Educational and Research Purposes'. Available at:
[Link]

 McKinney, W. (2022). 'Python for Data Analysis: Data Wrangling with Pandas, NumPy, and
Jupyter', 3rd Edition, O'Reilly Media.

 Microsoft Power BI (2025). 'Power BI Documentation: Best Practices for Data Visualization and
Star Schema Modeling'. Available at: [Link]

 MySQL (2025). 'MySQL 8.0 Reference Manual: Relational Database Design and Optimization'.
Available at: [Link]

 Netflix Technology Blog (2023). 'Using Big Data to Drive Global Content Strategy'. Available at:
[Link]

 Pandas Development Team (2025). 'Pandas Documentation: Handling Missing Data and Type
Conversion'. Available at: [Link]

 Statista (2026). 'Global OTT Video Market Revenue Forecast and Subscriber Growth Trends'.
Available at: [Link]

 Wickham, H. (2021). 'Visualization of Large Datasets: Principles of Heatmaps and Treemaps in

Business Intelligence'.

25 | P a g e