Churn prediction and prevention in Digital Marketing - Practice Problems & Coding Challenges

Challenge - 5 Problems

🧠 Conceptual

intermediate

Understanding Churn Prediction Models

Which of the following best describes the main goal of a churn prediction model in digital marketing?

A To increase the total number of customers regardless of retention

B To segment customers based on their purchase frequency only

C To identify customers who are likely to stop using a service soon

D To predict the exact date when a customer will make their next purchase

💡 Hint

Think about what 'churn' means in customer behavior.

Churn prediction models aim to find customers who might leave soon so companies can act to keep them.

❓ Model Choice

intermediate

Choosing a Model for Churn Prediction

You have a dataset with customer demographics, usage patterns, and past churn labels. Which model is most suitable for predicting churn?

A Principal Component Analysis (PCA)

B Logistic Regression

C K-Means Clustering

D Linear Regression

💡 Hint

Churn prediction is a yes/no problem.

Logistic Regression is used for binary classification problems like churn (yes/no).

❓ Metrics

advanced

Evaluating Churn Prediction Performance

Which metric is most important when you want to minimize false negatives (missing customers who will churn) in a churn prediction model?

A Recall

B Precision

C Accuracy

D F1 Score

💡 Hint

False negatives mean churners predicted as non-churners.

Recall measures how many actual churners are correctly identified, minimizing false negatives.

🔍 Analysis

advanced

Debugging a Churn Prediction Code Snippet

What error will this Python code raise when training a churn prediction model?

from sklearn.linear_model import LogisticRegression
X = [[1, 2], [3, 4]]
y = [0, 1, 0]
model = LogisticRegression()
model.fit(X, y)

A ValueError: Found input variables with inconsistent numbers of samples

B TypeError: 'int' object is not iterable

C NameError: name 'LogisticRegression' is not defined

D No error, model trains successfully

💡 Hint

Check if the number of samples in X and y match.

The feature matrix X has 2 samples but label vector y has 3, causing a mismatch error.

❓ Hyperparameter

expert

Optimizing Hyperparameters for Churn Prediction

You are tuning a Random Forest model for churn prediction. Which hyperparameter adjustment is most likely to reduce overfitting?

A Increase the number of trees (n_estimators)

B Increase the minimum samples per leaf (min_samples_leaf)

C Set bootstrap to False

D Decrease the maximum depth of each tree (max_depth)

💡 Hint

Overfitting happens when trees are too complex.

Reducing max_depth limits tree complexity, helping prevent overfitting.