Home  >  108 Data Analysis  >  108.20.30.20 Data Analysis - Calculating LTV

# Calculating LTV

Imagine a matrix where - y-axis is customer id and - x-axis is time period. - Each cell has the data of how much was spent by that customer during that period. (1)

The traditional way of thinking about profit is to add the columns up and get the revenue for the period. The marketing way of thinking is to add up the rows and get LTV. How much does it cost me to serve that customer? How much profit do I make on that customer? Doesn’t matter whether you add it up via the columns or rows, the sum is still the same - this is just a different way of thinking.

The bottom line is, instead of adding up the periods, add up the customers. The advantage is some customers have a negative value - fire them. Some customers have a lot of value - cultivate, retain, find more of them.

## Churn is different for contractual and non-contractual businesses

Noncontractual businesses have a situation where the customer doesn’t tell them that they’re not going to buy anymore - they just leave. Calculating CLV here takes into account the 3 factors of RFM. This makes the CLV a little more tricky to calculate. See the Fader paper, “RFM and CLV: Using Iso-Value Curves for Customer Base Analysis,” published in the Journal of Marketing Research in November 2005 and available from his website, www.petefader.com. (2)

## [Needed: a technical implementation example of calculating LTV]

References: 1. https://www.coursera.org/learn/wharton-customer-analytics/home Customer Analytics, Week 5, Eric Bradlow, Video: The Future of Marketing is Business Analytics 2. https://s3.amazonaws.com/kw-wdp/epubs/20150912-WDP-Fader-CustomerCentricity_Chapter4_CourseraCustomerAnalytics001.pdf - Page 88