Project 2 - Supply and Demand Gap Analysis

UBER Supply and Demand Gap Analysis

Aim

The aim of this is project is to identify and analyze the gap in supply and demand of uber cabs having two pickup points namely airport and city from the dataset downloaded from kaggle.

Introduction

Supply and demand gap analysis is done by every company to know the reasons for the company’s success or failure and to be prepared based on the customer trend and improve the company’s performance.

Uber is a cab service which provides its services through mobile application. The current dataset taken has the details for five days in the month of July of 2016. The main objective of this project is to analyze the dataset and give recommendations based on our findings.

Problem Statement

• To find the performance of UBER
• To find the availability/unavailability of the cabs
• To find the success rate of trips being completed

Methodology

Step 1: Data Exploration

In order to take any necessary steps we must understand the data we have, which is data exploration and based on our understanding we can go ahead with the appropriate data cleaning steps required.

Step 2: Data cleaning

I have used Power BI tool to clean and transform the data as per my understanding and I have replaced values, grouped the data and created a new table where it seemed fit.

Step 3: Data Visualization

I have used Power BI tool to visualize the data available to gain meaningful insights and better understand the dataset. I have used different visualizations like bar graphs, line graphs which are available in the tool.

Analysis

The dataset contains the following columns

Request Id- Gives a unique id when the customer makes a request for the cab.

Pickup Point – gives the point of pickup of the customer (airport or city).

Status – gives the status of the request whether it was completed or not (trip completed, no cars available, cancelled).

Driver id – gives the id of the driver who got the request if cars were available.

Request timestamp –gives the date and time of the request.

Drop timestamp – gives the date and time of the drop off.

The dataset contains null values in the driver id when there are no cars available and contains null values in the drop timestamp when the trip was cancelled or when no cars are available.

I have split the columns into date and time for ease of analysis. Further I created a conditional column called hour of the day from the timestamp so as to know the demand of the cabs based on the hours of the day.

Also created a new column considering only successful trips as supply and considering the no cars available and cancelled trips as gap.

• The dataset contains details about 6,745 requests, out of which only 2.83k trips were completed which 41.97% rides were successful and 58.03% of rides were unsuccessful.
• Out of the unsuccessful rides 1.26k were due to cancellation of the rides and 2.65k rides were due to unavailability of the cars.
• When we see the requests based on the different timings of the day we can see that the demand in the morning is from 5 AM up to 10 AM, and in the evening after 5 PM up to 10:00 PM.
• Further when we plot the graph taking the pickup point into consideration we can see that in the morning peak hours the requests are in the city and in the evening peak hours the requests are in the airport.
• Considering the status of the requests in the city we can see that during the morning the unsuccessful requests are mainly due to the cancellation of the rides.
• Considering the status of requests in the airport peak hours we can see that unsuccessful requests are due to the unavailability of the cars.

When we filter out for the city and the morning peak hours we get the below visualizations

Considering the peak hours 5 AM up to 10 AM in the city we can see that

• During this particular time frame a demand of 1.68k can be seen and out of which 472 requests are successfully completed accounting to 28.15% and there is a gap of 1.21k requests accounting to 71.85%.
• The majority of the gap is due to the cancellations which is about 820.

Now when we consider the peak hours for the airport we get the below visualizations

Considering the peak hours 5 PM up to 10 PM in the city we can see that

• We can see a demand of 1.8k in this period and only 370 requests i.e., 20.72% is being successfully completed and 1.43 k rides are unsuccessful i.e., 79.28%.
• Majority of the unsuccessful rides are due to the unavailability of the cars which is about 1.32k.

Insights

• In the Peak Morning Slot where the demand for City to Airport trips are high, there are maximum number of trip cancellations leading to great Gap in the Demand & Supply.
• A high gap of 71.85% can be seen during the morning peak hours.
• In the Peak Evening Slot where the demand for Airport to City trips are high, there are maximum number of “No Cars Available” leading to great Gap in the Demand & Supply.
• A high gap of 79.28% can be seen during the evening peak hours.

Recommendations

• The increase in number of vehicles will be one solution.
• Giving out rewards based on the number of trips during the peak hours without cancelling them.

Conclusion

Ubers needs to concentrate on these peak hours and add more vehicles to their fleet or increase the incentives of their employees so as to fill the gap they are facing.