15-Minute Delivery Performance Analysis — Zepto & Instamart

A data analysis project evaluating quick-commerce delivery performance for Zepto and Instamart, focused on identifying why orders miss the 15-minute SLA (Service Level Agreement) and what operational factors drive delays.

📌 Project Overview

Quick-commerce platforms promise ultra-fast (15-minute) delivery, but operational bottlenecks — like slow picking, late order acceptance, traffic, and weather — often cause delays. This project analyzes 1,000 delivery orders across both platforms to measure SLA performance and pinpoint the root causes of delay, with actionable recommendations to improve delivery efficiency.

Type: Junior Data Analyst Project Tools Used: Microsoft Excel (Data Cleaning, Pivot Tables, Dashboard/Visualizations, KPI Calculations) Dataset Source: Kaggle

🎯 Objective

• Evaluate 15-minute delivery performance across Zepto and Instamart
• Identify the operational bottlenecks causing delayed deliveries
• Analyze the impact of weather, traffic, distance, and rider experience on delivery time
• Provide data-backed recommendations to improve SLA compliance and customer satisfaction

🗂️ Dataset Information

Item	Value
Source	Kaggle
Raw Records	1,041 rows (including header)
Columns	41
Platforms	Zepto & Instamart
Missing Values	Filled using Median (numeric) and "Unknown" (categorical)
Duplicates Removed	40 duplicate ORDER_ID records
Final Records (after cleaning)	1,001 (including header)

Key columns: Platform, City, Zone, Dark Store ID, Rider ID, Product Category, Vehicle Type, Weather Condition, Traffic Index, Order/Accept/Pack/Delivery timestamps, Total Delivery Minutes, SLA Status, SLA Breach Reason, Distance to Customer, Rider Experience, Customer Rating, and more.

🧹 Data Cleaning Summary

• Removed 40 duplicate ORDER_ID records to ensure each order was uniquely represented
• Standardized text values (Platform, City, Weather, etc.)
• Replaced missing numerical values using the median
• Replaced missing categorical values with "Unknown"
• Standardized Rider IDs and platform names (Instamart, Zepto)
• Corrected inconsistent city names and text formatting

🔍 Key Findings

• A large proportion of orders were delivered after the 15-minute SLA — highlighting a major operational improvement opportunity
• Late Acceptance was the #1 reason for delivery delays
• Slow Picking was the second most common delay reason
• Weather and traffic conditions had a measurable impact on delivery delays
• Longer delivery distances generally resulted in higher average delivery times
• Delivery performance varied noticeably between Zepto and Instamart

📊 Headline Numbers

• 1,000 total orders analyzed
• 125 orders met the 15-minute SLA
• 875 orders exceeded the SLA target
• SLA achievement rate: ~12.5%

📈 Analysis Performed

• Pivot Tables: SLA Status breakdown, Platform vs. SLA, Delay Reasons, Weather Impact, Dark Store Performance, Traffic Impact, City-wise Performance, Vehicle Type Analysis, Rider Experience Analysis, Distance Analysis
• KPI Calculations: SLA achievement rate, average delivery time, delay breakdowns by platform
• Visualizations: Interactive dashboard summarizing delivery performance across all key dimensions

✅ Conclusion & Recommendations

Summary: Late Acceptance and Slow Picking were the most significant operational bottlenecks, with traffic, weather, and delivery distance further influencing performance.

Recommendations:

• Reduce order acceptance time through faster rider assignment
• Optimize warehouse/dark store picking operations
• Increase rider availability during peak demand
• Improve route optimization using real-time traffic information
• Monitor underperforming dark stores and provide operational support
• Increase staffing during adverse weather and peak hours

🚀 Future Opportunities

• Build a live Power BI dashboard connected to real-time delivery data
• Use machine learning to predict delivery delays
• Forecast order demand for better rider allocation
• Develop rider performance scorecards
• Analyze customer satisfaction alongside delivery performance
• Create real-time operational monitoring dashboards

📁 Repository Structure

├── zepto_instamart_RAW_DATA.xlsx          # Original, uncleaned dataset (1041 rows)
├── zepto_instamart_analysis_report.xlsx   # Cleaned data + pivot tables + dashboard + KPIs + report
└── README.md                              # Project documentation (this file)

🛠️ How to Use

1. Download both Excel files
2. Open zepto_instamart_analysis_report.xlsx to explore:
   • EXECUTIVE SUMMARY — objective, dataset info, key findings, conclusion
   • Cleaned_Data — the cleaned dataset used for analysis
   • PIVOT TABLES — breakdown by SLA, platform, delay reason, weather, city, etc.
   • VISUALIZATIONS — dashboard view of performance
   • KPI_CALCULATIONS — core performance metrics
   • CONCLUSION — findings and recommendations

👩‍💻 About Me

Aspiring Data Analyst learning SQL, Excel, Python, and Power BI. This project was built to practice real-world data cleaning, analysis, and reporting on a quick-commerce delivery dataset.