Customer Churn Analytics

Churn Analysis is the process of using data to understand which customers are leaving, why are they leaving, which are likely to leave, what can you do to reduce churn and what are the options to reduce churn?

Customer churn is a measure of the customers lost by a subscription business. Customer churn is also called customer attrition.

Customer Lifetime Value (CLV or CLTV) - is the average revenue you can generate from customers over the entire lifetime of their account. In simple terms, it is the money you would make from a customer before churning.

Customer acquisition cost (CAC) - is the money spent on converting a prospect into a customer or obtaining a new customer entirely.

Customer retention - Customer Retention is the ability of an organization to retain its existing customers. A high retention rate shows how valuable your product or service is to your customers. It is the scale on which your customers’ value for your products and services is measured. Measuring this key metric should be one of the bedrocks for your business to succeed.

Monthly recurring revenue (MRR) churn -  is the measure of revenue lost from customers who have cancelled or downgraded their subscriptions in a given month. In other words, it's the monthly erosion of your recurring revenue.

Notes:

• • Moving down the first row shows you how much new revenue you could acquire month on month
  • Going across columns shows you how much that cohort expanded or contracted
  • In the cohort above, you can see an adverse impact on revenue growth across customers in April. But what’s more interesting is that customers acquired in the recent months seem to have churned more than the older ones – indicating a high early-stage churn.

Customer churn rate is calculated as the ratio of the number of customers lost during a specific period of time(typically a month or a year) and the number of customers present at the beginning of that time frame.

While it’s inevitable, there’s such a thing as high churn. A figure of 5-7% annual customer churn is acceptable by most standards - anything more requires deeper investigation and redressal. A monthly customer churn rate of 5% translates to approximately 46% annual churn. That is, a business with a 5% monthly customer churn rate will end the year with half the customers it had at the beginning of the year. In other words, it will have to add 50% more customers during the year just to break even with the customer base it had at the beginning of the year.

Analysis

Customer churn analysis is the process of using your churn data to understand:

• Which customers are leaving?
• Why are they leaving?
• Which customers are likely to churn shortly?
• What can you do to reduce churn?

Churn analysis is important for the below reasons:

1. Understand product weaknesses and strengths - Churn analysis often reveals patterns that indicate common motivators for customers to leave you, such as price sensitivity or poor product adoption.
2. Uncover opportunities for better communication - Churn analysis reveals trends in customer behavior at every touchpoint.
3. Ability to predict, and thus reduce future churn - Churn analysis involves analyzing historical customer data to make churn prediction possible. 
4. Acts as a secret weapon during a crisis  - giving discounted offers can help retain customers

Types of analysis

1. Churn analysis by revenue
2. Churn analysis by industry
3. Churn analysis by georgraphy

Types of churn

1. Early of late stage churn - You can start by analyzing churn by activation dates. It tells you how soon (or not) the customer churned after activating the product. This can be analyzed using MRR retention cohort analysis.
2. Voluntary Active Churn - These are customers that proactively cancel your product or service. This type of churn can occur due to various reasons, such as poor onboarding, poor customer service, or switching to a competitor.
3. Involuntary passive churn - This type of churn is a leak in your revenue stream. Occurs when the customer’s payment is not completed for reasons such as when an expired card is used.
4. "Good Churn" - Not all churn is bad! Sometimes churn tends to weed out customers that were a bad fit for your product, service, or business model. Another example of ‘good’ churn is when customers leave after their short-term need for your product is satisfied, like an event or a short-term project. It’s also called “happy” churn. These customers also tend to reactivate their subscriptions later, so one way to track “happy churn” is to track reactivation MRR.
5. Downgrade churn - As the name suggests, this churn occurs when customers downgrade to a lower-tier plan, resulting in downgrade MRR. It could happen due to something like price sensitivity. You can reduce this churn by experimenting with your product packaging and pricing. To reduce downgrade churn, think of ways to pack more value and features into your customer’s current plan.

Import the libraries

In order to understand the need for and the description of each of the classification libraries, please visit the link "ML Libraries Defined"

Import the dataset

There are several ways of importing data into the notebook from the various data sources. To access the many ways of importing records into a dataset from the data sources, visit the page "Importing data into dataframe".

Sources

1. Luca Petriconi
2. Customer churn definition
3. Customer lifetime value definition
4. Customer acquisition cost
5. Customer retention
6. Monthly recurring revenue churn