CIVITTA's experts on customer lifetime value modelling

03 August 2020 4 min. read
More news on

A customer lifetime value (CLV) is one of the most important metrics in marketing, outlining the total value a company is expected to reap from a customer relationship. But how can a marketer determine a customer’s lifetime value? And what modelling provides the best outcome? Rokas Salasevicius and Justas Jankunas from management consultancy CIVITTA shed light on the topic. 

While there are different definitions of customer lifetime value, at CIVITTA we follow the following definition: “Customer lifetime value is a concept which estimates how much money the customer is going to bring to a company in the long run.”

Calculating and tracking customer lifetime value is of huge importance for marketers, for a number of reasons. First, it enables marketers to observe the business – CLV is used as one of the KPIs that provide information about the health of the business and customer base. CLV allows understanding how strong the current customer base is and track how it is changing over time, as well as identify reasons why it is changing.

Second, it provides insights on how to optimise acquisition. The concept of CLV – especially when used together with marketing attribution modelling to drive information about customer acquisition costs – is a strong tool which allows to identify and target user groups which are expected to bring the most value, as well as understand if these customers are generating positive return on investment. 

The customer lifetime value (CLV) model

Thirdly, insights coming out of customer lifetime value analysis can help increase the value of existing customers. If sophisticated enough methods are used, such as calculating the CLV for individual customers, the business can leverage the data to early identify the risky customer segments and act accordingly.

All in all, customer lifetime value enables companies to develop sales & marketing strategies, outlining how to differentiate between keeping existing customers onboard versus attracting new clients. Using the better information around profitability management, contours for the investment room can be crafted, and important information can be plugged into the broader customer experience lifecycle, such as pricing and monetisation opportunities, and touch point factors. 

Calculating a CLV

How then can the customer lifetime value be calculated? At the core, it depends on three factors:

  • How mature is the business?
  • What is the role of CLV, and how will it be used?
  • What kind of data is available? 

In our experience working with clients across Europe, we see four different levels of CLV maturity:

  • Beginner. Usually CLV is calculated as a very simple formula by taking the average of customer spend and multiplying by the tenure expected for the average customer. The method is simple and intuitive, but appropriate only for very early stages – as it does not provide any distinction between different customer base segments.
  • Moderate approach. This approach is still based on a rather simple formula allowing to calculate and compare CLV for distinct user segments. The drawback of this model is that the absolute CLV number might not be accurate, as customers tend to have a fluctuating churn rate over time. This is extremely important for new customers as in the first months they typically have rather high churn rate and then, over time, this stabilises.
  • Advanced (cohorts based) approach. This method allows to have accurate CLV evaluation for various segments, however, it requires a rather high amount of historical data and big segments sizes. The main idea behind the cohorts approach is that a cohort should be observed for a long enough period of time, allowing to calculate long term churn.
  • Probabilistic/machine learning approach. The most mature approach is based on complex calculations allowing to calculate CLV not only for customer segments, but also for individuals. These methods are the most advanced ones at the moment, and require a low amount of historical data. However, they are rather complex to implement.

To successfully apply customer lifetime value modelling, take a four-step structured approach. Start by identifying which business problem (or problems) you want to solve using CLV, then understand the most critical threats to customer transactions, followed by evaluating the current maturity and, finally, choose the most appropriate CLV calculation method.