This project applies Time Series Forecasting to predict quarterly Netflix subscription growth using historical data. Using ARIMA and Python’s data stack, I modeled the future subscription trend to help simulate how platforms like Netflix plan for subscriber growth, content strategy, and financial forecasting. The project demonstrates how data-driven insights can enhance decision-making in the streaming industry.
Forecasting subscriber growth is crucial for OTT platforms like Netflix to:
- Plan production and marketing budgets
- Anticipate server load and infrastructure scaling
- Strategize content development and release
- Ensure business continuity through accurate forecasting
This project was independently conceptualized and built to sharpen my skills in time series modeling, data visualization, and Python-based analytics.
- 📊 Analyzed 10 years of quarterly Netflix data (2013–2023)
- 🧠 Built an ARIMA(1,1,1) model with statistically validated parameters (p, d, q)
- ⏱️ Achieved sub-second forecasting for next 5 quarters with accurate results
- 🔍 Identified non-seasonal trends in subscriber growth using ACF/PACF and differencing
- 📈 Calculated and visualized quarterly & yearly growth rates with color-coded bars
- 🧮 Converted static growth data into a time-series format and made 5 future predictions
- 📊 Used Plotly for interactive, publication-quality visualizations
- Cleaned and transformed subscription data into time series format
- Visualized subscriber growth using Plotly and Matplotlib
- Conducted time series diagnostics: stationarity, ACF, PACF, differencing
- Built and evaluated ARIMA model with optimized hyperparameters
- Predicted future subscription counts and visualized with overlaid graphs
- 🕒 Time Series Forecasting using ARIMA
- 📅 Quarterly and Yearly Growth Rate Calculations
- 📉 Trend Analysis using Differencing
- 📊 Interactive Plotly visualizations with subscriber overlays
- 📦 End-to-end pipeline: data cleaning → modeling → forecasting → visualization
- 🔍 Insightful exploration of non-seasonal trends in subscriber data
netflix Subscriptions.csv– Raw data of quarterly subscription countsnetflix subscription forecast.ipynb– Jupyter notebook with full EDA, model building & predictionsREADME.md– Documentation and explanation of project
- Python
- Jupyter Notebook
- Pandas, NumPy – Data analysis and manipulation
- Plotly, Matplotlib – Visualizations
- Statsmodels – ARIMA model
- Scikit-learn – Metrics and support tools
- Git & GitHub – Version control
- 🔮 Subscription growth planning for streaming platforms
- 🎯 Budget and marketing forecast alignment
- 📆 Quarterly trend analysis for leadership dashboards
- 🔍 Subscriber engagement planning
- 💡 Use-case simulation for academic or business forecasting
- Inspired by a time series forecasting article by Aman Kharwal
- Dataset sourced from a publicly available growth record of Netflix subscribers
