Analyzing Bank Term Deposit Subscriptions: Two Tasks

Task 1: Prediction of Subscribers

Introductions

There has been a revenue decline for a bank and they would like to know what actions to take. After investigation, they found out that the root cause is that their clients are not depositing as frequently as before. Term deposits allow banks to hold onto a deposit for a specific amount of time so they can invest in higher gain financial products to make a profit. In addition, banks also hold better chance to persuade term deposit clients into buying other products such as funds or insurance to further increase their revenues. As a result, the bank would like to identify existing clients that have higher chance to subscribe for a term deposit and focus marketing efforts on such clients. 

 

The assessment consists of two analytical tasks: 
Task 1: Based on the analysis from a number of marketing campaigns the bank performed, predict which of the existing clients are most likely to subscribe to a term deposit. The marketing campaigns were based on phone calls. Often, more than one contact to the same client was required. For this task you are provided with two data files. The first one, 

MarketingCampaignsData.csv, includes information about: 
Various demographic information about customers – their age, job, marital status, and education. 
Customer account information – more specifically, status of their current credit (default), information about their past mortgage (housing) and/or personal loan applications (loan), and the preferred mode of contact (mobile or landline). 
Details about specific marketing campaign – marketing campaign (there were five in total, 1-5), month and day of week the customer was contacted, duration of call (in seconds), and the contact outcome (success or failure, poutcome column). 

The last column (called y) tells if the customer has eventually subscribed to a term deposit ('yes') or not ('no'). 

The second file, ExistingCustomers.csv, holds the information about bank’s current customers and includes everything listed above except column y. Building on the knowledge obtained from the first file, try to predict which of the bank’s current customers are most likely to subscribe to a term deposit. To build prediction model, choose the best prediction technique, explain which variables are most important for predicting client subscription to a term deposit, and establish who will subscribe. Provide a detailed explanation of your analytical approach – i.e. workflow model and its elements. Explain step by step what was done to get the best possible predictions. Provide detailed overview of the outputs and elaborate your findings. Provide links to your workflows and files (be explicit in which file you have saved predicted subscribers).

Task 2: Segmentation Analysis and Recommendations

 

Task 2: Once you have established who is most likely to subscribe to a term deposit, explore if different types of subscribers exist. More specifically, establish how many segments of subscribers there are, profile and name the identified segments, and recommend marketing strategy for each segment. As in the previous task, elaborate your approach and your findings. Provide link to your workflow model and files (be explicit in which file you have saved your clustering results). 

 

Once you have completed the two tasks, you need to provide recommendations on what actions the bank should take to improve subscription to its bank term deposit service. While you have creative freedom, your recommendations have to be linked with the insights you have created and other external sources (from research and/or practice). Please note that there will be two audiences reading this report. First audience is the bank’s board of directors. They do not like technical and lenghty reports, which require hours of reading. Instead, they prefer intuitive info graphs with very little text that can be easily understood. They are willing to read up to 3 infographic slides. Put these in the executive summary so they do not have to struggle finding them in your report. The second audience is the bank’s data analysis team. They will be interested to thoroughly read and inspect your analytical approach and findings. Therefore, be elaborate and precise and provide links to all your models so they can inspect your workflows, working files, and outputs. Make sure you justify decisions in building models. Bear in mind you do all of these analyses to help the bank improve subscription to its term deposit product.

 

Further instructions and expectation 

All analyses must be performed in Orange Data Mining software. If necessary, install add- ons that will assisst you in completing the two tasks. In explaining your analytical approach you are allowed to insert workflow screenshots. It is vital that your workflows are well annotated and made available. Therefore, save all your workflows and data files in a cloud (Dropbox, Google Drive, etc.) and provide links to those so that they can be inspected when marking. Failing to do so will have negative impact on your final mark. When making infographics you can use any software you want. Producing data analysis report is a very demanding task. Start working on it as soon as possible.

 

Suggested report structure 
The assessment is an individual report (4,000 words) where part of that is written report proper (2,000 - 2,500 words). The rest will be data analysis workflows and datasets that you have to produce by using Orange Data Mining software. The latter must be included into your submission (via link to a cloud storage, e.g. Dropbox, OneDrive, Google Drive, etc.) and will be counted as equivalent of 1,500 - 2,000 words.