To identify patterns and root causes behind ride cancellations to help Uber reduce their frequency and improve operational efficiency.
Uber Dataset from Kaggle https://www.kaggle.com/datasets/yashdevladdha/uber-ride-analytics-dashboard/data
MS SQL Server / Excel / python
● Clean the ride bookings dataset to handle missing values, inconsistencies, and errors in records related to cancellations.
● Standardize data formats and representations for uniformity.
● Perform Exploratory Data Analysis (EDA) to uncover the root causes, patterns, and trends behind ride cancellations.
● Derive actionable insights to help Uber reduce cancellation rates, improve driver-rider matching, and enhance overall platform efficiency.
● What is the overall cancellation rate, and what is the split between driver and rider cancellations?
● How does the cancellation rate vary by time?
● By Month/Season: Are there seasonal trends (e.g., more cancellations in rainy season)?
● Are there geographical patterns to cancellations?
● Which specific reasons are most commonly cited for cancellations?
● Use the dataset to analyze the impact of policy changes, such as introducing a cancellation fee or a new driver incentive program.
● Analyze the impact of cancellation chains (e.g., a driver cancelling leading to a rider cancelling another ride).
● Interactive Dashboard can be Develop for real-time dashboard for operations managers to monitor cancellation hotspots and trends as they happen, enabling proactive measures.
● Most rides are completed, but there's a significant number of cancellations by both drivers and customers.
● Go Mini and Auto are the most commonly used vehicle types for completed rides.
● Customers mostly cancel because "Driver is not moving towards pickup location"
● Drivers mostly cancel due to "Customer related issues" and "Personal & Car related issues"
● Peak booking hours are in the evening (around 6 PM)
● Weekdays see more bookings than weekends
● UPI is the most popular payment method, followed by Cash.
● There's a strong positive correlation between ride distance and booking value.