About this project
Project Background
In this project, I worked with hotel businesses in Homa Bay to analyze customer data and develop strategies to enhance sales and customer retention. The goal was to leverage data-driven insights to understand booking patterns, predict customer churn, and recommend targeted marketing approaches. My role involved cleaning and processing large datasets, building predictive models, and delivering actionable recommendations to help hotels improve their market performance.
Business Problem
The primary objective was to identify factors contributing to customer churn and uncover opportunities for sales growth. Hotel owners wanted to understand customer preferences, peak booking periods, and the impact of promotions on sales. Addressing inconsistencies in customer data and building reliable models to guide marketing strategies were essential to achieving sustainable business growth.
Tools
Python (Pandas, NumPy, Scikit-learn): For data manipulation, analysis, and building machine learning models.
SQL (MyQuery): For querying and managing large datasets.
R (dplyr, ggplot2): For data cleaning, visualization, and statistical analysis.
Google Sheets: For collaborative data organization and quick calculations.
Steps
1. Ask
Collaborated with hotel managers to define key questions:
What factors contribute to customer churn?
Which customer segments are most valuable?
What promotional strategies drive the most bookings?
Are there seasonal trends in bookings and cancellations?
2. Prepare
Data Source: Collected historical booking records, customer profiles, and transaction logs.
Data Sorting: Organized data into subsets (e.g., bookings, cancellations, customer feedback).
Data Credibility: Checked data integrity, identified missing or inconsistent entries, and flagged outliers for review.
3. Process
Cleaned datasets by handling missing values, correcting data types, and standardizing date formats.
Merged customer information with booking histories to create a unified analysis-ready dataset.
Conducted feature engineering to generate variables like booking frequency, average stay duration, and churn indicators.
4. Analyze
Exploratory Data Analysis (EDA): Visualized booking trends, cancellation rates, and customer demographics.
Predictive Modeling: Built churn prediction models using logistic regression, decision trees, and random forests.
Segmentation Analysis: Grouped customers by spending patterns, stay frequency, and feedback sentiment to tailor recommendations.
Key Observations
Repeat customers with longer stays had a lower churn rate.
Cancellations spiked during off-peak seasons, especially for non-refundable bookings.
Customers who received personalized offers were 30% more likely to rebook.
Weekend stays and holiday seasons showed the highest booking volumes.
Hotels that actively responded to negative reviews saw a 15% increase in returning customers.
5. Share
Created dashboards showcasing churn probabilities, booking trends, and customer segments.
Developed visual reports illustrating revenue impacts of different marketing strategies.
6. Act
Recommended targeted marketing campaigns focused on loyal customers and high-churn-risk segments.
Advised on offering flexible cancellation policies during low seasons to reduce churn.
Suggested implementing personalized promotions and loyalty programs to boost retention.
Proposed A/B testing different discount structures to identify the most effective incentive strategies.
Conclusion
This project provided valuable insights into customer behavior, helping hotels make informed decisions to maximize sales and improve guest retention. The predictive models and data-driven recommendations equipped hotel managers with the tools needed to optimize marketing efforts and enhance customer satisfaction.
Further Questions to Explore
How do local events influence booking rates and customer demographics?
What impact do online reviews and ratings have on customer acquisition?
Can dynamic pricing models further optimize revenue during peak and off-peak seasons?