Guest Intelligence Dashboard 📊

An interactive analytics dashboard built with SQL, Pyhton, Excel, Streamlit, Pandas, and Plotly to visualize guest behavior and sales performance from a public dataset on Kaggle (https://www.kaggle.com/datasets/rajatsurana979/fast-food-sales-report). This project aligns with the role of a Guest Intelligence Analyst by analyzing customer data, identifying trends, and presenting insights in a usable interface.

---

🌐 Live Demo

Try the Dashboard on Streamlit Cloud

---

🔍 Key Features

• SQL Views: Data was pre-processed and aggregated using SQL Server, then exported as CSVs
• Powerful Visuals: Interactive KPIs, bar charts, pie charts, and a monthly heatmap
• Cloud-Deployable: Hosted on Streamlit Cloud
• Modular Codebase: Organized by data view for easy updates or expansion

---

⚙️ Tools & Technologies

• SQL Server:

  • Designed and created 8 SQL views to clean, transform, and aggregate raw transactional data.
  • Used:
    • SELECT, FROM, WHERE, and JOIN statements to merge sales and item dimensions
    • GROUP BY, ORDER BY, and COUNT/SUM for aggregation logic
    • CASE statements for classifying revenue tiers (e.g., Low / Medium / High orders)
    • DATEPART() and CONVERT() functions to break dates into day, month, and weekday
    • CTEs (Common Table Expressions) to handle funnel stages and top-performers logic
    • Window functions like RANK() and ROW_NUMBER() to extract top 3 items by type

• Pandas:

  • Loaded and cleaned CSVs exported from SQL
  • Standardized column names and converted date fields
  • Performed grouping and aggregation (e.g., monthly revenue trend)

• Plotly Express:

  • Bar charts for top items and categories
  • Pie charts for time-of-day and order tier breakdowns
  • Line chart for revenue over time
  • Heatmap for time-of-day vs month revenue performance

• Streamlit:

  • KPIs using st.metric()
  • Interactive layout with st.columns() and st.subheader()
  • Easy deployment and caching with @st.cache_data

---

🧠 Key Insights by Visualization

📊 Monthly Revenue Trend

• Insight: Revenue peaked between April and July 2022, with a sharp drop starting March 2023.
• Actionable Takeaway: Monitor what operational or seasonal factors contributed to the early 2022 success, and investigate the cause of 2023’s decline.

🏅 Top 10 Selling Items

• Insight: Cold Coffee, Sugarcane Juice, and Panipuri are top sellers by volume.
• Actionable Takeaway: Consider upsell combos or promotions featuring these top items.

💰 Revenue by Item Type

• Insight: Beverages outperform Fast Food in total revenue.
• Actionable Takeaway: Focus on expanding beverage offerings or bundling them with food for increased average order value.

⏰ Revenue by Time of Day

• Insight: Sales are highest in the Afternoon and Night.
• Actionable Takeaway: Optimize staffing and inventory during these peak time slots.

🧾 Transaction Type Breakdown

• Insight: Online and cash transactions are evenly split in revenue share.
• Actionable Takeaway: Maintain both payment infrastructures and analyze preferences by customer segment.

🏷️ Order Revenue Tier Distribution

• Insight: 76.7% of orders fall into the High revenue tier.
• Actionable Takeaway: Introduce incentives to increase order value and shift more transactions into the Medium/High tiers.

⭐ Top 3 Items by Type

• Insight: Beverages consistently dominate both volume and ranking across item types.
• Actionable Takeaway: Consider loyalty rewards or featured promotions based on beverage sales patterns.

📊 Revenue Heatmap: Time of Day vs. Month

• Insight: Nights in Q2 and Q4 of 2022 showed consistently high revenue; early 2023 and beyond underperformed.
• Actionable Takeaway: Replicate successful marketing/timing strategies from 2022 to recover post-2023 declines.

---

📈 Project Workflow Overview

1. Data Source & Collection

• Dataset: Fast Food Sales Report
• Raw data downloaded and imported into SQL Server for preprocessing.

2. Data Cleaning & Modeling (SQL Server)

• Created 8 SQL Views for:
  • Aggregating revenue by item, category, transaction type, and time
  • Ranking top products using ROW_NUMBER() and RANK()
  • Categorizing revenue tiers using CASE
  • Parsing datetime into multiple dimensions (month, time of day)

3. Data Export & Wrangling (Pandas)

• Exported SQL views to .csv
• Cleaned column names and ensured datetime formats were standardized
• Grouped, sorted, and restructured data for visualization

4. Data Visualization (Plotly Express)

• Generated:
  • Bar charts (top items, item types)
  • Pie charts (time of day, transaction types, revenue tiers)
  • Line chart (monthly revenue)
  • Heatmap (revenue by month vs. time of day)

5. App Development & Deployment (Streamlit)

• Built dashboard using st.metric(), st.columns(), and st.plotly_chart()
• Deployed using Streamlit Cloud