# Why you need to know LTV

**LTV, aka Lifetime Value, aka Customer Lifetime Value (CLV)**is an indicator of a customer’s lifetime value. It shows how much money on average one user will bring for the entire time of using the product. LTV is universal; it is calculated both in web analytics and in mobile. It is considered for most types of products, whether it is Starbucks coffee houses, mobile operators, banks, SaaS products or games. In this article we will talk about why you need to know LTV and how to use it.

It should be noted right away that the Lifetime Value, calculated across the entire user base, is such a spherical horse in a vacuum. It can be used, but an indicator calculated by individual sections gives a more accurate result. For which ones - read below.

**LTV> CPI**

This is the main formula for all traffic analysis and the main condition for the effectiveness of attraction. The user must bring in more money than was spent on his attraction.

By CPI (Cost Per Install) in this case we mean the

**average cost of attracting one user across all channels at once**. If you are familiar with the abbreviation CPA (Cost Per Acquisition), traditional for web products, use it in the further formulas in this article.

In fact, that the average CPI, that the average CPA - indicators are rather arbitrary, because as a rule we pay one partner the amount A, the other - the sum B, the third - the sum C, and the total average CPI is most likely a value that is not equal Neither A, nor B, nor C. It is

**better to consider the LTV separately by the channels of attraction, by campaigns**, and therefore we come to the next application of this indicator.

**Assessment of the quality of the traffic source**

Each source has its own user acquisition price (CPI or CPA) and its own traffic quality, and therefore its own LTV. Therefore, it will be more efficient to calculate LTV for each channel separately.

In this case, you can get the total average LTV more than the total average CPI, however, in the context of the channels of attraction, you will see inefficient channels where this condition is not fulfilled. What to do in such a situation? You can, of course, immediately turn off the traffic source that fell out of favor. However, it will be more efficient to study it in detail, “cut” it into campaigns, countries and platforms, disable those where LTV is less than CPI. Better yet, put such an analysis into regular practice and disable the ineffective SubIDs that appear at the traffic provider.

**ROI calculation**

Metrics are operated by analysts, and money is given by owners and investors. And it is important for these serious people to know whether their investments will pay off. For this, the ROI (Return On Investment) metric was invented, which takes into account both LTV and the cost of attraction.

ROI can be considered in different ways, we are now talking about the formula

**ROI = LTV / CPI * 100%**. According to the results of calculations, the ROI should be more than 100%.

I recommend also calculating ROI for certain fixed time intervals from the moment of user registration (first entry):

**ROI N days = Cumulative ARPU N days / CPI * 100%**.

Here we introduce the new Cumulative ARPU N days metric, which shows how much money an average user brought in over the first N days of using the product. Choosing different Ns, you better understand the dynamics of ROI and can calculate another important indicator. Namely ...

**Payback period**

When will the money invested in the project pay off? See the chart:

The blue line is an indicator of Cumulative ARPU, it shows how much money an average user brings in over N days of using the product. LTV is the Cumulative ARPU limit for N tending to infinity (although in practice they take fixed N values like 120, 180, 360 days).

If the business works well, and the traffic pays off, then there is a point T at which the blue line (money brought by the user) becomes higher than the green line (money spent on the user). And the day on which this important event occurred is called

**the payback period**. Now you can tell the owner exactly when the invested money will be repulsed and when the ROI will exceed 100%.

**Cost Planning**

Back to the basic formula:

**LTV> CPI**.

When calculating, it is important to know about the concept of pure LTV, that is,

**LTV minus other expenses**: store commission, publisher’s and royalties, taxes in the end.

With CPI, everything is not easy either. To start buying traffic, you must first agree (manager’s salary), sign an agreement (lawyer’s salary), integrate (programmer’s salary), and we still don’t take into account the fixed signing fee for some contractors. Therefore, from CPI we will move on to the

**effective price of attracting eCPI**(by analogy with the effective bank rate).

As a rule, in a project there are also costs for maintaining user activity - technical support, community management, servers, and others. The final formula takes the following form:

**Pure LTV> eCPI + costs per user (variables, fixed)**.

It follows from this that costs should be planned so that the condition is satisfied after deducting all fees from LTV and adding all costs to the CPI.

**The dynamics of the**

LTV

**project is**based on the value of many metrics. It is affected by retention of users (Retention), and the share of paying (Paying Share), and income from paying (ARPPU). Instead of tracking the dynamics of several metrics, you can track the dynamics of LTV - this will show you how effective the changes you are making to the project.

If LTV is growing from month to month - great, keep up the good work. If it falls (and in most LTV projects it has a downtrend along the time axis), then it is time to take measures and improve the project.

**Prediction of future revenues**

If you know how to forecast LTV, and even calculate it in the context of channels, countries, platforms, etc., then you can quite predict how much money you will receive in N months. For example, you can answer the following questions:

- what will happen in 3 months if we now reduce paid traffic by 50%;
- if we enter the market of a new country in April, how much money we will get from it by the end of the year;
- if we make a change to the project that will increase user retention by 3%, how this will affect our revenue;
- when the traffic that we purchased from partner X will pay off;
- etc.

As you can see, LTV is the most important indicator in project analytics. But there is one difficulty - to calculate it, it takes time, but time, as a rule, is not. If you consider LTV in a short time, the forecast will not be the most accurate. If you make a calculation for a long period, then the question of forecasting ceases to be relevant, the future overtakes us.

**How to count LTV?**

This is a question that cannot be answered in one sentence. For example, here is a screenshot about calculating LTV from the book “Database Marketing: analyzing and managing customers” (a good and powerful book on analytics and marketing, if you like hardcore):

There are many ways to calculate LTV, and among them there are those that allow you to get an accurate forecast in a few days. So that you can study this indicator in practice, we invite you to a

**free webinar**, which will be held on February 1 at 18:00 Moscow time. I will talk about different methods for calculating LTV, as well as a few practical questions on predicting and using this metric. Sign up !

PS And, yes, there will not be such a hardcore at the webinar as in the last screenshot :)