__STYLES__
Customer churn, also known as customer attrition, is when someone chooses to stop using your products or services. In effect, it’s when a customer ceases to be a customer. It is a real problem across many industries, and the average churn rate can be surprisingly high. For some global markets, churn rates can be as high as 30%. Customer churn is measured using customer churn rate. That’s the number of people who stopped being customers during a set period of time, such as a year, a month, or a financial quarter.
In this project, I have used customer data of a USA based telecom company, to find out reasons why it's customers churned and recommend actions for its reduction.The dataset can be found here.
This dataset contains information about whether a customer churned or not and along with it provides a lot of relevant and useful features related to churn reasons, customer demographics, customer plans, data usage and monthly charges and many more. The complete description can be found in the metadata grids below:
I have also attached a link to my Tableau Story used in this project.
These are some of the initial insights obtained till now using only a handful of attributes I have in the data. So I need to dive in deeper to explore other data columns as well and check if I can find something relevant while analyzing them.
In case of both male and female genders, senior customers have about 11% more churn rate than the average. This is definitely an interesting finding which shall be explored deeply as to why seniors have more churn rate for this telecom company. Based on the chart above, one can see that above 60 years, the churn percentage is higher. But it should also be kept in mind that the number of customers also are less in that age bins.
Now here is one more interesting point. The company offers group contracts to customers from same household. The advantage for the customer is discounted rate and it also lets the company grow their customer base.
So here I have analyzed the average monthly bills for customers who are a part of the group and those who are not and check if they have a low phone bill, which may in turn effect the churn rate. Based on the analysis, the average monthly bill for individual customers is approximately $34, whereas those who are a part of the group have a maximum average of $23, which is a $11 difference and it is indeed substantial. It is also clear that the churn rate is higher for those people who are not a part of a group.
Actionable Insight: The company can direct it’s customer service department to reach out to the customers who have not taken group plan, tell them about family connection plans and explain the possible benefits.
Based on an overall analysis, 32.11% of customers who take an unlimited data plan churn, whereas only 16.1% of those who do not take it churn, and it seems to be a bit counter-intuitive because if a customer takes unlimited plans then he/she shall be the one who has the intention to stay.
So, in order to solve this anomaly, I dig deeper and further divide the visualization based on average monthly data used by the customers. Upon doing this, it becomes clear that customers who use less than 5GB of data monthly and have taken the unlimited data plan are more likely to churn.
Actionable Insight: So these customers, who use less than 5GB and have taken unlimited data plan can be contacted and offered with other plans that have a cap on data usage and will be cheaper for them.
Initially a text table is created in which rows show if a customer is internationally active or not and column shows if customer has international plan or not. It is observed that customers who are not internationally active and do have an international data plan tend to churn more. So the table is further detailed out in terms of total number of customers, which shows that the group with abnormally high churn rate of 71.19% has less customers as compared to other groups.
Actionable Insight: Contact customers who are an on an international plan but have not called internationally and propose them to downgrade their plan.
But the group which is internationally active and does not have an international plan also has a high churn rate and a substantial number of customers also. On further diving in to this group, the other values like number of international calls, international minutes and extra international charges incurred is also put up in the table. It is apparent that people who are internationally active, but have not taken international plan seem to have spent a substantial amount of time on international calls and since they do not have an international package they incur high charges.
Actionable Insight: Contact these customers and offer them an international package, by telling them about their international calling minutes and explaining the cost reduction for them if they opt for international plans
Based on the visual below, it is clear that customers with month to month contracts are more likely to churn. And on the top of that, customers who do a direct debit payment, of those, 1141 out of 1796 of them are likely to churn, so this should be taken in to consideration by the company executives.
3411 of 6687 customers who churned had month to month contracts and majority got offered with some better products or offers from competitors. The churn rate for this category is 46.29%.
Actionable Insight: Since majority of the churn reasons belong to competitors, a thorough benchmarking study of competitor products and services is recommended as without knowing about them, this telecom company can’t actually analyze its own shortcomings on customer’s end.
1538 out of 5166 customers who are not a part of group belong to senior citizens. And around 40% of them are among the churned customers, against an overall churn rate of 33% if we take all age groups into consideration. 1031 out of the 1538 senior citizens prefer direct debit as method of payment and in these customers the churn rate is 45%.
Actionable Insight: Company shall revise its payment platform especially for direct debit payments. There could be technical issues that customers may be facing, causing them to get frustrated and thus churn away.
Customers doing payment through direct debit also tend to make a lot of customer service calls. So, this further reinforces the points stated in the previous dashboard analysis.
Therefore, based on above analysis, the final recommendations to the company are: