Content Analytics Dashboard

Business problem: A broadcast network producing 10 hours of daily live TV needs to know which genres, time slots, and content acquisitions are actually driving audience growth — and which are burning budget for minimal return.

This project simulates three years of audience data for a regional broadcast + OTT operation and produces a full analytics dashboard: growth trends, scheduling optimization, platform breakdown, and content ROI.

Inspired by real audience analytics work at a Mexican broadcast network (UNIFE México, 2022–2026), where data-informed programming decisions drove 300% audience growth over three years, verified by independent INRA ratings.

---

What it does

Output	Insight
01_audience_growth.png	Year-over-year and cumulative growth curve
02_monthly_trend.png	36-month audience trend with seasonal patterns
03_genre_performance.png	Audience share and average viewers by genre
04_scheduling_heatmap.png	Optimal time slots per genre (scheduling matrix)
05_content_roi.png	Cost-per-viewer ranking across 80 titles and genres
06_platform_breakdown.png	Monthly viewer growth by platform (Linear, iOS, Android, CTV, Web)

---

Quick start

git clone https://github.com/Neckr0ik/content-analytics-dashboard.git
cd content-analytics-dashboard
pip install -r requirements.txt
python main.py

Charts are saved to output/. Runtime: ~10 seconds.

---

Project structure

content-analytics-dashboard/
├── main.py                  # Entry point — orchestrates full pipeline
├── src/
│   ├── data_generator.py    # Synthetic audience data (ratings, app metrics, content ROI)
│   ├── analyzer.py          # AudienceAnalyzer + ROIAnalyzer classes
│   └── visualizer.py        # All chart generation (matplotlib + seaborn)
├── output/                  # Generated charts (git-ignored)
└── requirements.txt

---

Key design decisions

• Synthetic data with realistic patterns — seeded RNG ensures reproducibility; growth curve is calibrated to 3× over 36 months with seasonal variation and prime-time multipliers
• Separation of concerns — data generation, analysis logic, and visualization are decoupled so each layer can be swapped independently
• Business-first output — every chart answers a real programming or acquisition decision, not just a technical exercise

---

Skills demonstrated

Python · Pandas · NumPy · Matplotlib · Seaborn · Data Modeling · Business Analytics · Content ROI · Audience Segmentation

---

License

MIT © 2026 Giovanni Oliveira