This project focuses on analyzing Netflix's content library using both SQL and Tableau. It includes solutions to 15 business problems related to Netflix's movies and TV shows dataset using SQL and provides insightful visualizations through an interactive Tableau dashboard. The analysis aims to uncover patterns in Netflix's catalog, optimize content strategy, and explore key trends such as genre popularity, content ratings, regional availability, and release years.
- SQL Analysis: Solutions to 15 business problems, including the count of movies vs. TV shows, most common ratings, top content by country, and more.
- Tableau Dashboards: Interactive visualizations showcasing genre popularity, content distribution across regions, release year trends, and ratings analysis.
- Business Problem Insights: Answers to business questions using SQL queries and data exploration to help optimize Netflix's content strategy.
Dataset Link: Netflix Movies and TV Shows Dataset on Kaggle
The dataset used for this analysis includes publicly available information about Netflix’s content library, including:
- Show ID
- Content type
- Titles
- Director
- Casts
- Date added
- Release years
- Content ratings
- Content duration
- Country availability
- Genres
- Description of content
- SQL: Used for solving business problems and extracting key insights from the Netflix dataset. Advanced SQL techniques were employed, including:
- Aggregation Functions:
COUNT(),SUM(),ROUND(), etc. - String Manipulation:
STRING_TO_ARRAY(),UNNEST(),SPLIT_PART(),TRIM(). - Date Manipulation:
TO_DATE(),EXTRACT(),CURRENT_DATE. - Window Functions:
RANK()for ranking data. - Pattern Matching:
ILIKEfor case-insensitive searches. - Conditional Logic:
CASEstatements for categorizing content. - Subqueries: To calculate average releases, filter specific years, etc.
- Filtering & NULL Handling: Using
WHEREandIS NULLto filter and clean the data.
- Aggregation Functions:
- Tableau: Used to create interactive dashboards, which offer deeper insights into Netflix's content. Key features include:
- Calculated Fields for deriving new metrics.
- Filters for dynamic content exploration.
- Interactive Charts to explore genre, ratings, and regional content distribution.
- Clone the repository to your local machine:
git clone https://github.com/PriyanshiNegi01/Netflix-Data-Analysis.git
- Open the Tableau workbook (
NetflixDataAnalysis_PriyanshiNegi.twb) using Tableau Desktop. - Explore the interactive visualizations and insights on the dashboard.
- Use the filters to explore different genres, ratings, or regions.
- Hover over the charts for detailed insights and data points.
- Navigate through the multiple sheets to view different aspects of the analysis.
This project provides a comprehensive analysis of Netflix's content catalog using both SQL and Tableau. The SQL queries tackle a range of business problems, while the Tableau visualizations help present these insights interactively. Together, they offer a deep understanding of Netflix's content trends, user preferences, and strategic opportunities for content optimization.
This project is licensed under the MIT License. See the LICENSE file for more details.