House Rockets Project

House Rockets is a company whose business model is based on buying and selling real estate, and its objective is to find good deals within an available portfolio, which is to find houses with low prices, in great locations and with great potential for reselling at a higher price, making the maximum profit possible.

The objective of this project is to perform an exploratory analysis of the data provided by the company in order to answer some business questions, as well as generate business insights to guide the company in its future decisions.

Business Problems

The business team informed that the main difficulty is to determine which properties should be bought and/or sold and at what price. In this way, this project aims to meet the following demands of the business team:

• What properties should House Rocket buy and at what price?
• Once the property is purchased, what is the best time to sell it and for what price?

Assumptions

• Properties built before 1955 was considered "old_property", after 1955 "new_property"

• Properties without basement have sqft_basement equals 0

• Properties with repeated ID's were excluded from the dataset, leaving only the most recent ID

• If the purchase price is greater than the regional median + seasonality, the sale price will be equal to the purchase price + 10%;

• If the purchase price is less than the region's median price + seasonality, the sale price will be equal to the purchase price + 30%;

• Properties with more than 10 bedrooms will be consider as outlier

• The properties price is affected by the seasons of the year;

Solution Planning

Final product:

Creation of a web application with the availability of two reports:

• Report 1: Real estate purchase recommendation for a specified amount
• Report 2: Recommendation for the sale of the property with a determined date and value

Tools:

To meet this demand, the tools used were:

• Programming language: Python
• IDE for application development: Pycharm
• Data exploration and visualization tool: Jupyter Notebook

Process:

The following steps were part of the project:

• The data collection used as a portfolio for this project was extracted from Kaggle
• Data Cleaning
• Exploratory Data Analysis (EDA)
• Generation of Insights through the creation and validation of business hypotheses
• Creation of a web application to make the results available

Top 10 Business Hypothesis

	Hypothesis	Validation	Business Insight
H1	Waterfront properties are 30% more expensive, on average	True	Investing in water view properties
H2	Properties built before 1955 are 50% cheaper in average	False	Don't invest in houses built before 1955
H3	Properties without basement have, in average, a lot area 50% bigger than properties with basement	False	On average properties without basement have lot are 18% bigger
H4	The average growth in property prices YoY (year over year) is 10%	False	The growth of properties prices didn't have significant variation over year (2014-2015)
H5	Properties with 3 bathrooms have an average price growth month over month (MoM) about 15%	False	There is no constant average growth in the price of 3-bathroom properties
H6	Properties in excellent condition are in average, 25% more expensive than properties in terrible condition	True	Investing in properties with excellent condition
H7	The price of properties with more than 50 years of construction and which have renovation is on average, 20% higher than those that have not been renovated	True	for properties over 50 years old, invest in those that have undergone renovation
H8	In properties with a water view, the price is on average 20% higher for those with an excellent view, compared to those with a regular view	False	Invest in properties with excellent view rather than regular view which are more expensive
H9	The average living room size in high-standard properties is 35% larger than in low-standard properties	True	The average living room size in high-standard properties is 38.05% larger than in low-standard properties
H10	The price of the properties are on average 10% higher in summer than in winter	False	The price of the properties are only 5.22% higher in summer compared to winter

Financial Results

When the solution proposed in this project is applied, the following financial result is expected:

Total Purchase Amount	Total Sale Amount	Profit
US$ 4079586744.00	US$ 4729811252.10	US$ 650224508.10

Conclusion

From the analysis and exploration of the data, it was possible to meet the main needs of the company, indicating the best opportunities for buying and selling real estate in the given region. In addition, important business insights were generated that will guide the company in future decisions.

Next Steps

Collect new data for the company, in order to expand the property portfolio, as well as expand the area of operation.