How to calculate CLV index? Discussion of historical methods

According to popular Pareto principle, up to 80% of revenues in the business can come from just 20% of clients. These values may be different, but the mechanism well describes the situation of many companies, which get a significant amount of their revenues from the special, the most committed group of clients. Each company should know them and their characterization. CLV index (Customer Lifetime Value) will be helpful here.

On the Internet you can find many definitions of CLV (and also its various names, such as LTV or LCV), i.e. the value of customer life. Without going into details, we can assume that CLV is the sum of income that have been (or will be) generated during the relation between your company and its clients. Of course, you are never sure how long this business interdependence will take, and how much it is affected by various factors, therefore normally we calculated CLV for the certain period of time. For example, you can specify that “monthly/quarterly/annual CLV of X client was Y PLN.”

How to calculate CLV? Historical methods

Qualification of the value of CLV for each client or group of clients can be accomplished in two ways: historical and predictive. In this article I will focus on the historical approach, which is simpler in the use. On the other hand, examples of advanced methods of calculating CLV can be: moving averages, regression analysis, Bayesian conclusion and model Pareto/NBD. Discussion of these advanced techniques will appear in the following text, in which we will focus only on the predictive methods of calculating CLV.

Let’s go back to the historical approach. There we can distinguish two common ways of calculating the CLV using the ARPU (average revenue per user) or by cohorts analysis, which we described in details in a previous article. I’ll start with the characteristics of the method of the ARPU, which involves the use of existing customer spend in determining his or her CLV. I’ll do this on a quite simple example, to explain its mechanism. Let’s suppose, that in the last six months, clients named John and Kate have shopped in the following way:


If in a half of the calendar year you will want to calculate the average monthly CLV for John and Kate, then you can do the following:

  • John: (50 PLN + 0 PLN + 0 PLN + 50 PLN + 50 PLN + 200 PLN) / 6 = 58,33 PLN;
  • Kate: (300 PLN + 0 PLN + 0 PLN + 150 PLN + 0 PLN + 100 PLN) / 6 = 91,67 PLN.

The formula is very simple. You count the sum of all revenues received from a client and divide it by the length (number of months) of your relation. Both Kate and John bought something in the last six months.

Thus obtained ARPU value you need to multiply:

  • x 12 to receive annual CLV; one-year time perspective is assumed to clients who are interested in products with high seasonality of sales, for example shoes;
  • x 24 to receive a two-year CLV; two-year and longer CLV should calculate for clients who are interested in long-term products or services, such as computers.

Thus the annual CLV for Kate will be in 1100 PLN, in the case of John it will be 699 PLN.

The ARPU method is simple to use, which is its unquestionable advantage. While using it you must remember not to do a simple conversion of a total revenue by a total number of clients, cause the average obtained in this way, says nothing about the actual CLV. You should also be aware that it is a method of calculating CLV, which does not take into account many variables, such as the differences between the new and old clients, changes in the offer, seasonality. Characteristics of your clients and their behavior changes over time.

Therefore you might be interested in the already mentioned cohorts analysis, through which you will discover not only the level of spending of individual clients – for example, those who have made their first purchase in March – but also discover patterns of consumer behavior in the coming months. The disadvantage of the ARPU is the fact that it is a simple average. Therefore, the value calculated CLV might always be changed by the high value of purchases during the first months of the relation between the company and its clients. For example, if you have a lot of new clients who make numerous purchases at the beginning, CLV calculated for them may be overestimated. Generally, if you run a new company or one that is going through a period of very high growth or change, then you should approach the calculated CLV with caution. There is a high probability that the resulting value will not coincide with the actual behavior of your clients, they might take in the coming months.

Discrepancies between the calculated CLV and actual purchases of clients in the case of the ARPU method can reach values of 50-120%. If you want to minimize the risk of a revaluation – use the cohort analysis.
The easiest way to calculate the CLV using cohorts is to create a “medium cohort “, that is where the average of purchases is stored, which the customer did in the first month of being your consumer. Then calculate the same for the second and subsequent months:


As you can see in the chart above, which is an example of the calculation of CLV using cohorts, the level of consumer spending falls in the coming months. While the average value of purchases for the first month is 100 PLN, after a year of relation it falls by 90 PLN and stabilizes at around 10 PLN per month. Chart can be used to estimate the value of the CLV. For example, if you want to calculate the 12-month CLV for a new customer, you add up all the average values for 12 months. On the other hand, if you want to estimate how much clients who have made their purchase three months earlier, will spend the next 9 months, add up the values of the last 9 points (months) on the chart.

Is it worth to use the CLV?

Indicator such as the CLV has a significant advantage over the other parameters that characterize your business. The number of clients, sales, growth rates – all this says a lot about the “as it is”, but only to a small extent can anticipate and respond to the question “how can this be?” The calculation of CLV allows to estimate the future state of the business, so you can take appropriate early intervention. Moreover, the use of this indicator will enable you to identify the group of clients who spend most of their resources (in practice cash, but also time, interest) for your products or services.

Without this knowledge you will not be able to implement appropriate marketing strategies aimed at maintaining or increasing the involvement of a customer, for example, previously described scoring. 20% group mentioned at the beginning is a group of consumers, whom you should take particular care, because most likely they are responsible for a large part of your income. Their loss can be very painful. Another key benefit of the CLV in business practice is to optimize the process of managing clients and their acquisition. You can calculate the CLV for the different sales channels (for example an online store, mobile application and profile auction site), and then decide to increase or decrease spending on any of them.