Customer Lifetime Value: How to calculate LTV

This is part 2 of our series on Lifetime Value, previous segments are available in our archives.

The hardest part of calculating your Lifetime Value metric is often determining the length of the customer lifetime. It doesn’t seem so at first, as intuitively it should be as easy as measuring the difference between when a customer is acquired and when they are lost.

Naieve model of calculating lifetime value

Unfortunately, rarely do customers let you know when you’ve lost them! It’s very common that customers will go away quietly and you will not know if they will come back or not.

It can be hard to tell the difference between customers who have gone quiet and customers who are lost. But, have no fear, we will cover two ways to do just that: windowing and last action.

Option 1: Windowing

In this approach, you choose a set period of time (window) after which you will conclude a quiet customer is lost. Each time a customer interacts, the window is reset to start at the time of the most recent interaction. The size of your window will depend on your business; a supermarket might use a window of a few weeks (people buy food every week) but a ski resort might use a window of a year (since you can only ski during the winter).

Using Windowing for calculating lifetime value

Choosing the right window is critical for this approach, because if your window is too long it will take too long to realize you have lost a customer. Likewise, if your window is too short you will consider customers lost who were really just quiet and coming back on their own.

This brings up an obvious question: what do you do if a customer returns after your window is over? You have a few options but the best is to simply count them as a new user (again).

Advanced Windowing for calculating lifetime value

That is not as bad as it sounds. If your window is long enough, say 6 months, your product may be so different that the returning customer really is approaching it fresh. It’s also possible that what brought the customer back was something entirely different than why they were a customer in the first place.

Let’s break down the pros and cons of this approach.

Pros: It is very easy to calculate and is flexible because you can choose and adjust your window.

Cons: Your choice of a window is critical to success. If your window is too long it make take too long to realize you’ve lost a customer!

Option 2: Last Touch

A more aggressive approach, which works particularly well for products that have daily interactions with customers, like video games,  is to assume you have lost the customer after every interaction (touch) they have with your product, but if they interact with your product again, then you update your assumption and extend their lifetime.

Using Last Touch for calculating lifetime value

As time goes on, and the customer interacts with your product, their customer lifetime grows.

Using Last Touch for calculating lifetime value

Obviously, this means that there are cohorts of customers who will have artificially short customer lifetimes. If a customer joined a week ago, they cannot have interacted with your product for a month! For this reason you need to be careful in considering which cohorts you use when calculating your lifetime.

Let’s break down the pros and cons of this approach.

Pros: You know immediately which customers you need to focus on bringing back.

Cons: It is difficult to calculate and overly pessimistic when measuring lost customers.

This is a complex topic so I recorded a short video (5:00) that might be easier to understand. There are many other techniques you can use, and hopefully this has gotten you thinking about the best method to use for your business.

Tomorrow we’ll dive into how to get started with calculating your LTV.

Quote of the Day: “It took me four years to paint like Raphael, but a lifetime to paint like a child.” – Pablo Picasso