franklinen / happy-customers

Customer Happiness Prediction Project

Overview

This project aims to predict customer happiness based on survey responses collected from a cohort of customers. The survey includes questions about various aspects of the customer's experience, such as delivery times, order accuracy, pricing, courier service, and ease of use of the ordering app. The goal is to build a predictive model that can classify whether a customer is happy or unhappy based on their survey responses. Achieving this goal will enable the company to take proactive measures to enhance customer satisfaction and improve overall operations.

Background and Motivation

Business Problem

As a rapidly growing startup in the logistics and delivery domain, maintaining high levels of customer satisfaction is crucial to our success. The COVID-19 pandemic has introduced several challenges that have impacted our operations. Despite these challenges, we remain committed to making our customers happy. Understanding what contributes to customer happiness is key to driving improvements across all levels of our business.

Why This Matters

Customer feedback is invaluable in helping us identify areas for improvement. However, collecting and analyzing this feedback is challenging, particularly in a high-growth, global expansion context. By predicting what makes customers happy or unhappy, we can take targeted actions to address any issues and enhance the overall customer experience. This project is a critical step toward building a data-driven approach to customer satisfaction.

Datasets

The dataset provided for this project includes survey responses from customers. The target variable (Y) indicates whether a customer is happy (1) or unhappy (0). The feature variables (X1 to X6) correspond to specific questions in the survey, with responses on a scale from 1 to 5. Here’s a brief description of the dataset:

• Y: Target attribute indicating customer happiness (0 = unhappy, 1 = happy).
• X1: My order was delivered on time.
• X2: Contents of my order were as I expected.
• X3: I ordered everything I wanted to order.
• X4: I paid a good price for my order.
• X5: I am satisfied with my courier.
• X6: The app makes ordering easy for me.

Each feature (X1 to X6) is measured on a scale from 1 (strongly disagree) to 5 (strongly agree).

Success Metrics

The primary success metric for this project is achieving an accuracy score of 73% or above in predicting customer happiness. This threshold has been set based on the business requirements and the need for actionable insights from the model.

Milestones

• 1. Data Exploration and Preprocessing Objective: Understand the dataset, identify missing values, and perform any necessary data cleaning. Tasks: Conduct exploratory data analysis (EDA) to identify patterns and correlations. Handle missing values and outliers. Normalize or standardize the data if required.
• 2. Feature Engineering Objective: Enhance the predictive power of the model by creating new features or modifying existing ones. Tasks: Generate interaction features or polynomial features. Test different feature selection techniques to identify the most important predictors.
• 3. Model Development Objective: Develop a predictive model that can accurately classify customer happiness. Tasks: Experiment with various machine learning algorithms (e.g., logistic regression, decision trees, random forest, XGBoost). Optimize model parameters using cross-validation. Evaluate model performance on the training dataset.
• 4. Model Evaluation and Tuning Objective: Assess the model's performance on a validation set and fine-tune it for better accuracy. Tasks: Split the dataset into training and validation sets. Fine-tune the model parameters to improve accuracy. Ensure the model generalizes well and avoids overfitting.
• 5. Final Model Testing Objective: Test the final model on a private test set and prepare it for deployment. Tasks: Evaluate the model on the private test set provided by the company. Prepare a detailed report summarizing model performance.
• 6. Documentation and Reporting Objective: Document the entire project and provide insights based on the model's predictions. Tasks: Prepare a comprehensive report detailing the methodology, findings, and recommendations. Create visualizations to illustrate key insights. Ensure all code is well-documented and follows best practices.

This project is critical to our mission of enhancing customer satisfaction in the logistics and delivery domain. By accurately predicting customer happiness, we can take proactive measures to address any issues and continue to deliver exceptional service to our customers.