72 lines (49 loc) · 1.99 KB

Website-Data-Analysis

🛠️ Tools Used

Python, Pandas, NumPy
Seaborn, Matplotlib
Google Colab

🚀 Objectives

Analyze how users arrive at the website (Channel Group breakdown).
Understand engagement behavior (sessions, bounce, engagement time).
Discover hourly traffic trends.
Compare engaged vs non-engaged sessions by channel.
Recommend data-driven strategies for improving engagement.

🧹 Data Cleaning

Set the correct headers from the first row.
Converted Date-Hour from string (e.g. 2024041723) to datetime.
Coerced numeric columns to proper data types.
Derived Hour feature for temporal analysis.

📊 Analysis & Visualizations

📈 1. Traffic Trends Over Time

Line chart showing Sessions and Users by timestamp.
Identifies peak traffic windows.

📊 2. Top Channels by User Volume

Bar chart grouped by Channel Group.
Reveals which marketing channels drive the most users.

⏱️ 3. Average Engagement Time

Engagement time visualized per channel.
Helps identify channels that attract higher-quality traffic.

🔄 4. Engaged vs Non-Engaged Sessions

Grouped bar chart showing split by channel.
Highlights where traffic is bouncing or interacting.

⏰ 5. Hourly Heatmap

Heatmap of sessions by Hour and Channel Group.
Provides insights for optimal content/campaign scheduling.

🔍 Key Insights

Organic Search and Email show higher engagement rates.
Some channels (e.g., Display) drive traffic but with low engagement.
Email traffic spikes in the morning; Social Media in the evening.
Engagement rate should guide channel optimization, not just traffic volume.

💡 Recommendations

Increase focus on high-engagement channels.
Improve targeting and landing pages for low-engagement channels.
Schedule campaigns based on peak hourly traffic windows.
Prioritize engagement rate over raw traffic for success metrics.