Hotel Booking Analysis

Business problem:

Hotels in cities and resorts have experienced significant cancellation rates recently. As a result, each hotel is currently dealing with a variety of problems, such as decreased revenues and less-than-optimum hotel room usage. As a result, our main objective is to provide a comprehensive business guide to handle this issue and minimize cancellation rates in order to boost the hotel's efficiency in producing income.

The major subjects of this research are an investigation of hotel booking cancellations as well as other reasons that affect their business or yearly income creation.

Research Question:

1. What are the more important influencing variables for canceled hotel reservations?
2. How can we reduce the number of hotel reservations that get canceled?
3. How will hotels be supported in choosing their prices and marketing strategies?

Hypothesis:

1. When prices are higher, there are more cancellations.
2. Customers tend to cancel more frequently when there is a long waiting list.
3. The majority of customers use traditional travel agencies to book their travel.

Motivation

Have you ever wondered what if there was a way you could predict which guests were likely to cancel and adjust the overbook rate accordingly? That would be great right?

Luckily, by using Machine learning with Python, we can predict the guests who are likely to cancel the reservation and this could help produce better forecasts and reduce uncertainty in business decisions.

In this article, I will apply Exploratory Data Analysis (EDA) to get insights from the data set to know which features have contributed more in predicting cancellations by performing Data visualization with Matplotlib & Seaborn. It is always a good practice to understand the data first and try to gather as many insights from it.

Exploratory Data Analysis (EDA) with Data Visualization

To better understand the dataset, we have to come up with a list of questions.

1. What are the Top 10 Countries of Origin of Hotel visitors (Guests)

• Around 40% of all bookings were booked from Portugal followed by Great Britain(10%) & France(8%).

2. Which Month is the Most Occupied with Bookings at the Hotel?

• According to the graph, August is the most occupied (busiest) month with 11.62% bookings and January is the most unoccupied month with 4.96% bookings.

3. How many Bookings were Cancelled at the Hotel?

• According to the pie chart, 63% of bookings were not canceled and 37% of the bookings were canceled at the Hotel.

4. Which Month Has Highest Number of Cancellations By Hotel Type?

• For the City hotel, the number of cancelations per month is around 40 % throughout the year and for the Resort hotel, the cancellations are highest in June, July & August and lowest during November, December & January.

5. How many Bookings were Cancelled by Hotel Type?

• For the Resort Hotel, a total of 25.14% Bookings were canceled and for the City Hotel, a total of 74.85% Bookings were canceled.

6. Relationship between Average Daily Rate(ADR) and Arrival Month by Booking Cancellation Status

• The highest Average Daily Rate (ADR) has occurred in August and due to the highest ADR in August, maybe it could be one of the reasons for more cancelations in August.

7. Total Number of Bookings by Market Segment

• Around 47% of bookings are made via Online Travel Agents, almost 20% of bookings are made via Offline Travel Agents and less than 20% are Direct bookings without any other agents.

8. Arrival Date Year vs Lead Time By Booking Cancellation Status

• For all the 3 years, bookings with a lead time less than 100 days have fewer chances of getting canceled, and lead time more than 100 days have more chances of getting canceled.

9. Relationship between Special Requests and Cancellations

• Around 28% of bookings were canceled with no special requests from the guests followed by 6% bookings were canceled with one special request from the guests.