The most common question that online retailers ask us at Custora is, “How can I increase my Customer Lifetime Value.” And not far behind is, “So what exactly is CLV?”
This is part 2 of a 3-part series discussing techniques to calculate Customer Lifetime Value (CLV) in a retail setting. In the first part, we explained why ARPU-based approaches often lead to inaccurate CLV numbers. In this piece, we dive into the pros and cons of Cohort-based CLV calculation.
Cohort analysis is a simple, yet powerful, form of analysis that tracks how a group of users behaves over time. Cohort analysis provides many valuable insights – and one of those is that it can be used to predict customer lifetime value.
Predicting CLV based on cohort analysis is not only simple – it can also be fairly accurate. In our own data sets, 1 year value predictions derived from cohort analysis have an average margin of error around ±15%. While that number might seem high, it’s a huge improvement over the numbers derived with ARPU.
However, we need to take caution. There are many situations in which cohort analysis produces inaccurate CLV numbers – and the margin of error in these cases can be higher than 150%.
How to predict customer lifetime value by using Cohort Analysis
To derive CLV numbers from cohort analysis, we first need to define our cohort. One popular approach is to define the “average cohort.” To do so, we can run simple queries to determine what a customer spends, on average, in his first month of being a customer. We can repeat the process to determine what a customer spends, on average, in his second month, and so on.
We can then chart how the average group of customers behaves as time passes after the first purchase.
Pitfalls in calculating CLV – why using an “average retention rate” can lead to grossly inaccurate CLV numbers
Customer Lifetime Value (CLV), the total profit a business will make from a new customer, is a valuable metric. It puts a theoretical cap on what a firm is willing to spend to acquire a new customer.