Predicting your game’s monetization future
An analysis on how top games developers calculate lifetime value
Many of us dream of building the next iconic game that will be remembered by future generations for its unique visuals and innovative gameplay. In turn, we generally spend a lot of time thinking about some of the fundamental business questions that underlie a sustainable business. For example: How much can I spend in acquiring a new player? What is the potential value of this user versus another one? How can I quantify the social effect of people sharing my game and bringing in new players? When will my players churn and what can I do to prevent it?
The breadth of data generated by gameplay and the ability to target actions at a user level has led to customer analytics becoming a core part of a games business. The complex nature of mobile game monetization and competition in the games industry means that there is pressure to find more innovative approaches to take action from your data, in order to give your business an edge. One method is to build models that help to predict the lifetime value (LTV) of your game players. This article provided an analysis on how top game developers calculate lifetime value. If you’d like to find out even more about the topic, read the full whitepaper.
Lifetime value in games
What is LTV?
As the term already implies, LTV estimates the total monetary value of a specific player during his or her entire lifetime.
Although lifetime value is a widespread metric that allows gaming developers to understand the value of their gamers, today there is no standard way of calculating it. Most developers either use their own custom method of calculating LTV, or are using a 3rd party tool. Regardless of the calculation method, there are three main pitfalls developers can fall into when working with LTV.
- The first mistake many developers make is not including all revenue sources in their LTV calculations. It’s quite common for developers to calculate LTV only for IAPs, and add at the very end a percentage on top in regards to the other sources of revenues e.g. advertising. LTV, whenever possible, should include all monetization business models, such as IAP, ad revenue, etc., as otherwise we might not include a legitimate revenue and hence miss out on opportunities to grow.
- The second pitfall some developers encounter is assessing the impact of less obvious factors such as: release of new titles by competition, the stage of the company, changes in exchange rates, etc. These can skew LTV significantly without the developer making any changes to the game at all. Overall we need to consider LTV as a dynamic metric that evolves continuously with the game thanks to: a) Internal changes — new content or functionalities, in-game economy, player behaviors. b) External changes — game trends, competition, currency fluctuations, platform changes.
- Lastly, some developers see LTV as a competitive strength. However, LTV is just is just a metric and therefore it does not represent a competitive advantage towards other developers. As well it is not an exact value hence it should be also considered as a range (e.g. LTV is in between 1 and $1.2) or within a certain accuracy (e.g. 85% confidence). Ultimately the goal is to avoid lifetime projections that are too optimistic/pessimistic and will end up driving bad behaviors.
How will LTV help my business decisions?
LTV is commonly used for player acquisition, but it can also be used for many other purposes; such as overall business profitability, optimizing Live ops, and in managing new feature releases. Among the common usages of LTV we find:
- Profitability: LTV is an estimation of the total cash-flows a user will provide throughout his/her entire lifetime. So if we compare the estimated income to the costs associated, we can estimate the bottom line impact to the business’ cash flow on a per user basis.
- User acquisition: As long as the cost of acquiring a new player is not above the net LTV, in theory we can still invest in user acquisition. A rule of thumb applied by many developers in player acquisition is that costs should ideally remain below 1/3 of the gross LTV but will depend in many other factors (expected growth, market landscape)
- New features and Live ops: Content updates can impact retention, engagement and monetization, all drivers of lifetime value. Top developers test the effect of the new updates prior to launch to see the overall impact in LTV.
- Building new games: Certain developers will assess the viability of an unreleased game by looking at the LTV while still in beta.
Popular ways of calculating LTV in mobile gaming
As mentioned above there is no standard way of calculating LTV. However developers usually agree that LTV is based on at least two variables: lifetime (looking at user engagement and retention) and monetization (average number of transactions, monetary value, conversion rate).
Although there is agreement that these two variables have to be somewhat part of the calculation, there is also a debate around the exact way to calculate them.
Lifetime is usually measured as retention. The debate starts here with different views of how retention is calculated:
The classic daily retention approach only looks at gamers that connect at a specific day after the install date, while range retention looks at those returning during an interval of time, e.g. week. Rolling retention looks at people returning after a certain period of time.
Each of these calculation methods can fit certain types of games better. For example, a storytelling game that releases episodes or seasons at certain periods of time — and therefore has gamers return only once the new season has been released — will argue that classic retention is not as relevant for them and they might be more interested in range retention and season/episode completion rate. Some hyper-casual game developers look at retention per hours instead of days, as in their case the first hours (not the first days) will determine the success of their game.
One last word of caution on retention — it is usually defined as the person opening the app. It’s worth noting that there might still be significant differences between opening the app to — for example collect a daily reward, or to play another battle in an online multiplayer game. As such we might want to re-define retention as the act of opening the app to do a certain action
How far you want to go in your LTV estimate?
Before calculating the LTV metric, developers usually agree on a specific period of time for LTV calculation purposes e.g. estimate LTV for the next 90 days, 180 days, 1 yr, 2 yrs, or even 5 years.
Note that some people might find this concept contradictory as the term ‘lifetime’ itself theoretically means the total duration of the entire life of a person. However as LTV is usually an average estimation, a developer might want to remain conservative and calculate the LTV to a certain period of time. Note as well that the further you go in the future estimation, accuracy will tend to be lower.
A few factors will impact your choice of the period and developers usually look at a number of factors, including the following:
- Type/genre of game: the more specifically expected life cycle of the game genre overall. E.g. a super casual game might have a shorter life versus a hardcore game running as a service hence the LTV will be calculated at different periods.
- Business model: IAP’s vs. subscriptions vs. advertising. As an example, generally subscriptions users might have longer lifetimes and hence the chosen period might be longer
- Stage of the company: early vs. mature stage. Early stage companies will often select longer and more optimistic periods of time when calculating LTV because they are relying on the future development of a certain technology, or due simply to a lack of historical data.
- ROI horizon: well funded companies might be able to invest in user acquisition for longer periods of time, hence extending the lifetime period. e.g. 180 days is the period that will allow enough revenues to reach a breakeven point.
As with retention, there is debate on how the monetization variable is calculated. Most gaming developers look at ARPDAU (average revenue per daily active users), but some take into account ARPU (average revenue per monthly users), or ARPPU (average revenue per paying users). As we will see later on, depending on the model used to estimate LTV, we will use one or another.
Regardless of which monetisation variable you choose, it’s important to be consistent with the time period you select, and be aware of some of the limitations with averaging out the resulting number. For example, the ARPDAU of a game will probably fluctuate significantly if we estimate it based on the last quarter, month, or week.
Common LTV models in gaming
Assuming that with increasing complexity the models also increase in accuracy, we can cluster these in the following way:
- Historical averages and benchmarks — based on historical data or past games.
- Simple predictive models — forecasting some variables such as retention or spend.
- Advanced predictive models — such as the ‘buy till you die model’ (BTYD), which comes from the Pareto/NBD and BG/NBD models or machine learning models.
As we have seen, the lifetime value metric has a variety of usages and is widely adopted in the mobile gaming industry. However, we have also seen that there is no standard way of calculating it, therefore there might be multiple valid ways to do so, depending on several internal (game nature, company resources, data available), or external (type of audience, competitors) factors.
As such, whenever calculating the lifetime value of our players, we need to be ready to make trade-offs in terms of accuracy and resources needed, in order to make the most of this valuable resource.
When using the lifetime value metric for user acquisition or live ops, we need to try avoid some common pitfalls such as: overly optimistic calculations, missing out sources of revenues, or considering it a competitive strength which may lead to under- or over- estimation. We might also want to consider key aspects such as calculating the net LTV, discounting cash flows, or adequately segmenting the calculation.
I hope you have found this overview useful in giving you a better understanding of the potential of LTV in driving better business decisions. Given the complexity of the topic, and further insights available on the subject, download our whitepaper to learn more best practices on calculating LTV for games.
What do you think?
Do you have other thoughts on LTV for game developers? Join the discussion in the comments below or tweet using the hashtag #AskPlayDev and we’ll reply from @GooglePlayDev, where we regularly share news and tips on how to be successful on Google Play.