Hoor Ismail

TELECOM CUSTOMER CHURN ANALYSIS PRESENTED BY : HOOR ISMAIL INSTRUCTED BY : ILSA QAZI DIGITAL HUSTLERS : Data Analytics COHORT 3 ABOUT PROJECT The telecom company provides telecom services to many clients. But due to many reasons. Customers switch from one service provider to other. DATA SET ABOUT DATA SET • Each row represents a customer and each column contains the customer attributes described in the data. • The dataset shows the customer’s account details like customer id, gender, senior citizen, partner and whether dependent on someone or not . • It shows various services that each customer has signed up like phone service, tenure, Multiple lines, internet service, online security & backup, device protection, streaming TV & movies and tech support. • It also shows customer account information like a contract, payment method, paperless billing, monthly charges and total charges. • Customer who left their service is called Churn. INSIGHTS TO BE FOUND 1. What is the distribution of customers by churn status, and how many customers have churned compared to those who have not? 2. What percentage of customers have churned? 3. Does churn vary significantly between genders? 4. How does senior citizen status impact the churn rate? 5. Which tenure duration has the highest likelihood of churn? How does tenure length correlate with customer retention? 6. Which type of contract has the highest churn rate? 7. How does churn differ between month-to-month contracts and longer-term contracts? 8. How do different services (Phone Service, Internet Service, Online Security, etc.) impact customer churn rates? 9. Which payment method is associated with a higher likelihood of churn? And How does the payment method affect customer churn? VISUAL ANALYTICS AND FINDINGS • Import Libraries • numpy and pandas are imported for data manipulation and analysis. • matplotlib.pyplot and seaborn are imported for creating visualizations. • Read Data • pd. read_ csv("Telco-Customer-Churn.csv") loads a dataset from a CSV file called "TelcoCustomer-Churn.csv" into a Data Frame named df. • A Data Frame is like a table where rows represent individual data points (like customers), and columns represent features (like age, gender, or churn status). • Show Data • df. head() displays the first 5 rows of the dataset. DATA INFORMATION DATA DESCRIPTION converted 0 and 1 values of senior citizen to yes/no to make it easier to understand INSIGHTS TO BE FOUND The majority of customers, 5,174 have not churned, while a smaller portion, 1,869 have churned. This indicates that customer retention is relatively strong overall. INSIGHTS TO BE FOUND From the given pie chart we can conclude that 26.54% of our customers have churned out. INSIGHTS TO BE FOUND From this graph we can observe. The dataset has an almost equal distribution of male and female customers. INSIGHTS TO BE FOUND The chart shows that the majority of customers are not senior citizens, with a count of 5,901 compared to only 1,142 senior citizens. This indicates that non-senior citizens constitute the dominant segment of the customer base. This means most of the customers are younger. INSIGHTS TO BE FOUND From this graph we can conclude. Comparative a greater percentage of people in senior citizen category have churned INSIGHTS TO BE FOUND From this graph we conclude that people who have used our services for a long time have stayed and people who have used our services 1 or 2 months have churned. INSIGHTS TO BE FOUND From this graph we conclude that people who have month to month contract are likely to churn then from those who have 1 or 2 years or contract. INSIGHTS TO BE FOUND The majority of customers who do not churn tend to have services like Phone Service, Internet Service(particularly DSL),and online security enabled. For services like Online Backup, Tech Support, and Streaming TV, churn rates are noticeably higher when these services are not used or are unavailable. INSIGHTS TO BE FOUND The chart shows that customers using "Electronic Check" as a payment method have the highest churn rate (1,071), while other payment methods like "Mailed Check," "Bank Transfer (automatic)," and "Credit Card (automatic)" have significantly lower churn rates (308, 258, and 232, respectively). This suggests that customers using automatic payment methods are less likely to churn compared to those using manual methods like electronic checks. SUMMARY From my exploration and analysis of the Telco Customer Churn dataset, I've performed several steps that highlight important trends and insights. Here's a summary: Summary of Insights: 1.Churn Percentage: • 26.54% of customers have churned. 2.Demographic Insights: • Gender: Churn rates do not show a significant difference by gender. • Senior Citizen Status: Senior citizens have a higher churn rate compared to non-senior citizens. 3.Tenure Analysis: • Customers with short tenure (1–2 months) are more likely to churn. • Longer tenure correlates with customer retention. 4.Contract Types: • Month-to-month contracts exhibit significantly higher churn compared to 1-year or 2-year contracts. 5.Service Usage: • Customers without Internet Services like Online Security, Tech Support, and Streaming TV are more prone to churn. • Those who have services like Phone Service and DSL Internet are more likely to stay. 6.Payment Method: • Electronic check users have a higher likelihood of churning compared to other payment methods like bank transfer or credit card. RECOMMENDATIONS! •Provide special discounts and loyalty programs for senior citizens. •Offer personalized customer support to address their specific needs and make the service easier for them to use. •Create onboarding programs to guide new customers in using the services effectively. •Offer introductory discounts to encourage them to stay longer. •Give discounts or additional benefits to customers who switch from month-to-month contracts to 1-year or 2year contracts. •Make long-term contracts flexible, allowing customers to make changes without penalties. •Educate customers about the benefits of services like Online Security and Tech Support. •Bundle these services with discounts to make them more appealing to customers. THANK YOU!