You are probably among the 8 out of 10 people who believe Customer Lifetime Value (CLV) is directly connected to your company’s growth and performance. However, if your company is among the 6 out of 10 that calculate CLV incorrectly, we’ll show you here how easy it is to avoid group-think and use accurate, simple, and practical CLV with little effort.
Even though in the wake of predictive analytics sophisticated modeling is emerging, simplistic ways that depart from a correct customer valuation are still a convention. There are numerous sources online providing instructions to calculate lifetime value. In our experience most of those sources suggest that CLV equals average revenue per customer
multiplied by retention rate
or repurchase rate
:
CLV = Average Revenue per Customer * Retention Rate
The above common method of CLV calculation is flawed in three main ways:
It assumes that business metrics don’t change
The retention rate or repurchase rate estimates the approximate duration of a customer’s relationship with the company. The average revenue assumes that the customer (or portfolio of customers) will always bring the same revenue in every future period. However, neither revenues nor churn rates are likely to stay the same for all customers in the future.
Revenue does not constitute value, profit/cash flow does
Another important flaw is taking revenue as the value creator instead of profit. Economic and financial theory suggests that the value of any asset is determined by the discounted cash flow that this asset will generate over its lifetime. The cash flow is always net of any necessary expenses. Thus, CLV should be equal to the cash flow or profits that the customer will generate.
It does not use a discount rate
The third shortcoming is not using a discount rate. The time value of money should be taken into account otherwise the ‘V’ (value) of CLV will be overestimated due to lack of discounting.
A sound calculation of CLV should incorporate the revenue that a customer is expected to generate and the costs necessary to generate this revenue. The profit, resulting from the difference of these two figures, should then be discounted at an appropriate rate.
A sound calculation of CLV should incorporate the revenue that a customer is expected to generate and the costs necessary to generate this revenue. The profit, resulting from the difference of these two figures, should then be discounted at an appropriate rate.
Paul Berger and Nada Nasr in their 1998 paper “Customer Lifetime Value: Marketing Models and Applications” clearly determined a theoretically and practically valid approach for CLV calculation, which is relatively simple and straightforward. The paper used several cases, modifying oversimplified assumptions (like constant retention rates) step by step to come up with an accurate CLV model, in line with economic theory about the financial valuation of assets. The paper determined that CLV consists of the following inputs:
Total gross contribution (GC)
coming from the customer. The total revenues from the customer less the direct variable costs necessary to provide the customer with a product or service.Retention costs (M)
necessary to encourage the customer to remain with the company.Retention rate (r)
which is essentially a probability that a given customer will stay with the company at any given period (e.g. a given month).A discount rate (d)
to account for the time value of money.The model looks as follows:
Despite the fancy mathematical expression this approach is relatively easy to apply and meaningful. Generating revenue always requires some costs relative to the amount of sales, like delivery costs or customer training costs. Often there is also necessary spending to keep customers from churning, whether this includes customer support, customer success, or additional marketing costs. Furthermore, irrespective of customer retention efforts some customers will inevitably churn. To reflect the churn in the model the calculated profit is multiplied by the probability of customer staying, i.e. the retention rate.
For example, if we were expecting to generate a 100€ profit from a middle-tier customer next month, and we forecast the churn to be around 3% in that month, we could say with confidence that we are expecting the customer to generate value of 100 * (1 - 3%) = 97€
. Although a real-world calculation would have more details it would be only marginally more complex. We have constructed a more elaborate example of a CLV calculation (spreadsheet) and the screencast below which explains it in more detail.
The same exercise performed for a forecasted period that covers the average lifetime value of a customer would provide all expected profits/value additions coming from the customer for each of the lifetime months. To complete the CLV calculation profits are discounted to present date, and aggregated to a single-figure customer lifetime value. This type of calculation is flexible and insightful. It may be used with assumptions of constant average inputs as well as applied to explicitly forecasted metrics, where GC
, M
and r
vary depending on business or market forecasts.
We hope you will use this as an opportunity to accurately determine the true lifetime value of your customers, if you aren’t already!
CLV calculation example: https://github.com/kostasfe/clv-calculation-example