Hello! Please do Question 3) How sensible is the approach with a single model? C

Hello! Please do Question 3) How sensible is the approach with a single model? Can you suggest a better approach?
a)Provide updated estimates of probabilities that customers 672, 354, and 5,203 will leave.
b)What factors contribute the most to the predicted probabilities that these customers will leave?
Below are attached case file, 2 excels, and screenshots for question 2)To start, run a single regression model that best predicts the probability that a customer leaves.What is the predictive probability that customer 672 will leave between December 2011 and February 2012? Is that high or low? Did that customer actually leave?What about customers 354 and 5,203?
Thank you!