Summer 2019 HW Project Assignment — Acquiring new Banking Customers

Assignment:

You will be working from a dataset that contains modified records of a marketing campaign for an international bank. The objective is to identify what determines whether a customer signs up for a new account at the bank. Your answers are due in 2 weeks, on 7/29/19.

The dataset is available on NYU Classes as an Excel spreadsheet, “Big_Bank.xls”, and as a SAS dataset, “Big_Bank.sas7bdat”. The dataset is comprised of 11,162 records, and 17 original variables

Data Variables (listed in order as they appear in the dataset):

Information about the customer:

1 – age (numeric). Age of the customer in years.

2 – job : (categorical). The type of job customer is categorized as having. Values are: (‘admin.’, ‘blue-collar’, ‘entrepreneur’, ‘housemaid’, ‘management’, ‘retired’, ‘self-employed’, ‘services’, ‘student’, ‘technician’, ‘unemployed’, ‘unknown’)

3 – marital : (categorical). Marital status. Values are: (‘divorced’, ‘married’, ‘single’, ‘unknown’; note: ‘divorced’ means divorced or widowed)

4 – education (categorical). Highest level of education completed by customer. Values are:

( ‘primary’, ‘secondary’, tertiary’, ‘unknown’)

5 – default: (categorical). Has the customer defaulted on a loan in the past? Values are: (‘no’, ‘yes’, ‘unknown’)

6 – balance: (numeric). The amount of money in a customers existing account.

7 – housing: (categorical). Does the customer have a home loan? (Values are: ‘no’, ‘yes’, ‘unknown’)

8 – loan: (categorical). Does the customer have a personal loan? (Values are: ‘no’, ‘yes’, ‘unknown’)

Information about current and past marketing efforts

9 – contact: (categorical) How was the customer last contacted? (Values: ‘cellular’, ‘telephone’)

10 – day: (numeric) Day of the month that the customer was last contacted? (Values: 1,2,3,…)

11 – month: (categorical) Month that the customer was last contacted? (Values: ‘jan’, ‘feb’, ‘mar’, …, ‘nov’, ‘dec’)

12 – duration: (numeric). The duration in seconds of the last contact or call with the customer.

13 – campaign: (numeric). The number of contacts performed during this campaign and for this client including the latest contact)

14 – pdays: (numeric). The number of days that have passed since the client was last contacted from a previous campaign (-1 means client was not previously contacted)

15 – previous: (numeric). The number of contacts performed before this campaign and for this client. (Note that if pdays = -1, then previous = 0)

16 – poutcome: (categorical). Outcome of a previous marketing campaign (Values: ‘failure’, ‘nonexistent’, ‘success’)

Dependent variable or target

17 – deposit: (categorical). This is what you want to predict. Has the client signed up for a new deposit-account? (Values: ‘yes’, ‘no’)

Due Date:

The assignment is due on NYU Classes by class time on 7/29/19. Your answer is a written report. Your written report should be no longer than 5 pages maximum for the written text (tables, graphs, charts can be in an appendix). Your answer should cover the 5 points below:

I. Database Marketing HW/Project:

Perform a Multiple Regression analyses on this dataset to arrive at an answer, using predictors you think would be useful and/or derive new ones to use.

1. Justify/explain why you decided to use the predictor variables you selected (20pts)
2. Perform/execute the analysis using SAS, and explain how determined your best and final model? (30pts)
3. Describe/explain your model in “non-technical” terms (15pts)
4. Based on the results of your analysis, what type of customers and/or marketing campaigns would you recommend that the bank use to find customers that would sign up and open new deposit-accounts? (20pts)
5. How would you evaluate the success of the campaign? (15pts)