Project Proposal – Telecom Customer Churn

Project proposal – Telecom customer churn

 1) Justification in terms of why the question is worth asking?

• The target that we are identifying is churn and why telecom customers churn. The attributes that we considered are payment method, gender, whether the customer has a partner, if the customer is a senior citizen, tenure, contract type, tech support, device protection, monthly charges, and total charges.
• This study helps telecom companies better understand the preferences of customers and enact measures to help retain if not increase their clientele.
• With intensifying competition, being able to analyze customer churn data sets a company apart and is a clear-cut competitive advantage over other companies.

2) Describe who would be benefiting from the answers to the question. And identify one main perspective.

• The telecom providers themselves will be the ones who benefit the most from this analysis. Other examples of companies that can benefit from churn analysis include companies in the insurance industry, banks issuing credit cards, etc.
• Analysis of this data would give telecom providers insights into which attributes make their customers churn and which attributes retain customers. Customer retention is much more profitable than the acquisition of new customers, so this data and the analysis would be highly valuable to the telecom providers.
• In the business world this kind of analysis also helps maintain healthy competition within the industry by tapping important information about customers.

3)  Data Source and size:

• All data were collected through Kaggle, a platform where datasets are uploaded by companies and users for predictive modeling and analytics competition. “Telco Customer Churn,” Kaggle, 2017,https://www.kaggle.com/blastchar/telco-customer-churn. The analysis of this data spans across in total of 7043 customers.

4) The target variable is valid and serves the objective of the problem. It is clearly defined and measurable.

• The target variable for this data would be “churn,” which is identified by a 1 and 0, which represents yes and no, respectively. The objective of the problem is to identify the significant reasons why customers in the telecom industry churn. This, in turn, helps the telecom providers to identify new ways to increase retention and positive customer experiences. The target variable is clearly defined and measurable amongst the population.

5) A non-trivial number of attributes should be included. Attributes should also be interesting and potentially have an impact on the target.