Data Analytics in Mobile Apps & Games
The mobile application market is constantly evolving, providing us with the next solutions used every day. The Google and Apple platforms are full of interesting ideas from the “why I did not came up with it!”. Some users even say that the application has significantly changed their lives. Stories of couples who met thanks to the Tinder mobile application, recovered form with the Endomondo mobile application, real savings for drivers using Fuelio or the Yanosik mobile application – this is a picture of changes for users and developers of mobile applications.
All these mobile applications have brought solutions that have left a lasting impression on someone’s everyday life. However, most of the history of developers is not a sudden flash of genius and magical materialization of the perfect product. This is specific work and changes based on in-depth user analysis. Their needs, behavior, sympathy and antipathy. It is continuous development. That is why many mobile applications used by us today differ so much in their appearance and functionality from the moment of the premiere. Developers should never stop listening to their clients and ensure that their product is further developed.
Below we have prepared a brief description of the most important indicators of mobile application analytics that will help us to provide the best possible product. Remember that without the ability to analyze data in games and mobile applications, paid mobile application marketing becomes pointless. These were divided into main categories listed below:
- User Performance Indicators
- Number of Installations / Deinstallations
- Number of Users
- DAU – Daily active users
- MAU – Monthly Active Users
- Geo Indicators
- Engagement Indicators
- Session length
- App Usage time
- Session intervals
- Retention rate
- In-app user activity
- Business Indicators
- ARPU – Avg. Revenue Per User
- ARPDAU – Average Revenue per Daily Active User
- LTV – Lifetime Value
- CAC – Customer Acquisition cost / cost per new customer
- Connection errors
- Loading time
- App crush
User Performance Indicators
Number of Installations / Deinstallations
Tracking the number of downloads of a mobile application is one of the most “visual” ways to show its growth. This is obviously a huge simplification, because popularity does not have to be equal to financial success, but it can not be concealed that in this case, more means better. There is always a risk that our mobile application will be abandoned by the user shortly after its installation, therefore the use of such an indicator requires a wider context. The data related to the place from which the user went to the store are also important. In such a list, the number of downloads from a given channel can tell us if our activities convert into a specific result. If we see increased traffic on social media, it will be a good idea to increase our marketing efforts and further work in this field. It is also logical that in the case where we check the number of installations of our application, the number of deinstallations should be equally taken account of, to effectively reduce uninstallation value.
Number of Users
The number of times a mobile application has been downloaded and installed, perfectly reflects its status in relation to the number of application users, as not always one number equals the other. The number of users can be measured according to several criteria, therefore below we will present several indicators for active users. It is important to carefully define what active user means for us. Depending on the type of application, mobile game in which we intend to measure their quantity, we may consider an active user as a user who performed certain activity. This can be the transition of the next stage in the game, sending a message or sending an invitation to friends. It is not a good idea to consider anyone who just opened the application or a mobile game as an active user, because in this case the context and what we should expect from the user count. Active is not the same for everyone, so it is important to be sure what it means to us. Developers can treat this aspect in a Boolean-way, and the results they receive are often used to build false complacency, instead of getting to know their clients better.
DAU (Daily Active Users) is an indicator that reflects the number of unique, active users who have just started a session with our mobile application on a given day. This type of indicator is one of many that takes on greater significance along with other data, although in many cases, such a short period of time can itself give us a picture of our actions. Such a case may be the launch of a marketing campaign. If we expect users to visit our mobile application on a daily basis, the DAU is a great starting point for further observation. Among marketers there are voices that this is not a measurable indicator. If we well define the user’s activity, DAU allows us to better evaluate our monetization model of the mobile application.
An example of defining an active user? According to Facebook, an active user is one who: logged in to the application or on the page that made the action using the available functions, eg clicked “I like it”. Therefore, an active user in this case will not be considered someone who used FB plugins on other sites, even if he “gave a likes” or share some content non straight through Facebook. Good practise is that in the way that we measure DAU allows the same user to be separated when he uses two different devices on the same day. It can be a private and business phone or tablet – if this user is treated as “new”, it will disrupt our statistics. You can use user account/mail/phone verification to easily do it right. It will also help us to verify users in other sections of time when they exchange their device for new ones.
Unlike DAU, the MAU (Monthly Active Users) indicator takes account of a 30-day period. This means that the unique active user has performed the actions we have set out for once in a month – the statistic does not change if at that time he took the action repeatedly.
For the purpose of calculating the MAU, Twitter defines the active user in the following way. Active user must follow at least 30 users and be followed by at least 1/3 of them. If such a user logs in at least once within 30 days, he / she qualifies for MAU.
The use of DAU and MAU may be different for specific types of mobile applications. Certainly, the “cyclicity” of visiting a given mobile application will have a big impact on the reception of these indicators. DAU will be more important for mobile games with micropayments, MAU for applications for booking flight flights.
The DAU to MAU ratio is referred to as the “viscosity” of a mobile application. By comparing these two indicators, we can determine how often users return to us. The value obtained will always be between 0 and 1 and can be expressed as a percentage. Suppose that the DAU daily average for our application is 2000, and the MAU is equal to 10,000. This means that each user has logged in 20% of the time in the last few days, that is, on average every 6 days.
In the case of mobile games, the ratio of 20% would be a very good result, although mobile applications such as Facebook Messenger or WhatsApp can boast a score of over 50%.
How the mobile application market is built allows you to easily reach users from around the world. It does not matter whether we are creating a huge studio or if we are a small entity located in the garage – there are no obstacles to our application or a mobile game on all continents. This fact, however, means that we should consider additional factors when developing our mobile application.
Knowing where our users come from may help us in several ways. Let’s assume that the largest number of users comes from one region, even though the mobile application is properly available globally. Investigating the reasons for this by carefully analyzing the behavior of these users can give us a recipe for duplicating this success in other regions.
Looking at the outflow of users with the separation of a specific region, developers often discover problems that even the uninterested ones may not be aware of. Elements that are related to local culture, when they are not adapted to users, may be “to be or not to be” for our mobile application. There is a difference between translating the mobile application and its location to a given region. Let’s look at the calendar. In some parts of the world, the date formats differ significantly. If our mobile application with its functionality relied heavily on an agenda, it could be useless without proper changes for a part of the population. The record of units of measurements and weights may also be important, so entering them may be a good idea.
Often, the choice of a given game or color can discourage users to our application or mobile game. Tracking the uninstallation or outflow of users in the region may lead us to problems related to local culture, even forcing us to adopt completely new solutions. In different parts of the world, you can read from left to right, in others from right to left. For applications in which you need to scroll the screen to the page to go further, the user experience will vary.
Session length measurement refers to a single use of a given application as part of its single launch. It is customarily assumed that if the user does not perform any actions in the application for more than 30 seconds, the session should be considered interrupted. The session is counted from the moment of starting to closing the application and its length reflects how much the user is involved in the application. The average length of a session is very different within different application categories. Using, for example, Uber, we count on a quick execution of the activity for which we reached for the application. We are only interested in calling for transport, and the developer counts on each subsequent short visit, i.e. more courses. This is one of the exceptions to the rule “the more the better”, because generally for most categories, a long session is a good session.
App Usage Time
Although the phrase “customer is king” may seem a bit exaggerated, the fact is that customers should be at the center of every business. In the case of mobile applications and games, users are the driving force behind our venture. Therefore, our task should be to ensure that you spend as much time with the application as possible. The user’s time in the application is the sum of the length of all previous user sessions on a daily, weekly or monthly basis. We have repeatedly mentioned how much competition prevails on the mobile application market, which is why this indicator for many developers better presents the popularity of the mobile application than the number of downloads or the number of ratings and comments in the store.
This indicator tells us how much the user of the mobile application needs time to start the next one after the session. So far, we have counted a single session and their total time, this indicator complements our knowledge about how quickly the user returns to us. The authors of mobile applications are interested in making the session interval as small as possible, so if it comes to alarming values, it is necessary to act. Push notifications, mailing, text messaging, social media – we should use any channel within our capabilities to download the user back.
Frequently referred to and discussed indicator, equally often called one of the most important, in particular in monetization models based on a free application with the possibility of shopping inside. Measurement of retention of mobile application users is necessary to determine the value of our client. This indicator tells us about how our application was accepted by the recipients. The higher the retention rate, the greater the chance that the user will take the actions we want, giving us more opportunities to convince him.
As a large part of the application after installation is used only once, then the successes in this field are of great importance. To be able to accurately measure retention, in addition to using the right time intervals, we should also classify our users well. The division into appropriate cohorts will serve in-depth analysis, allowing more precise action. Our mobile application has a group of very involved users? We should treat this group in a special way to achieve even better results. Some of our clients give up the service? In this way, several more groups can be created, which will be divided according to the reason for the outflow. Each cohort is a different recipient, other problems and other messages. User retention in a mobile application is one of the most difficult challenges that mobile application developers must cope with.
In-app user activity
Under this slogan there can be many different activities that we should follow and count. Through the activity in a mobile application, we can understand the number of entries to a given section, the number of clicks on a specific button or just the sum of individual transactions. We create this indicator ourselves, based on what seems to be important for the further development of the mobile application, which is why we can say that this is interdisciplinary. Although the subject of the observation may be related to UX, marketing activities or options available in our mobile application, again, the user is the most important here and his actions are at the center.
Conversion is the first business indicator that tells us whether our assumptions have come true. For example, if a user visited our mobile application to make a purchase after seeing an ad on our partner’s website, we can say that the advertisement is converting. The thread itself may sound a bit intricate, although the concept itself is usually quickly adapted by the creators of mobile applications. What matters most here is the clear definition of the goal that we want to achieve. Therefore, the conversion can be watching an advertising movie, clicking a banner, registering in a mobile application, joining the newsletter, entering contact details or even visiting a specific section of the application. It is important that these activities take place after switching from a specific channel. Readers of the newsletter after its opening linked the account from social media to the account in our game? Such were our assumptions? So we have a conversion.
In the same way, we can examine the impact of specific CTAs within a mobile application. We inform our user about the possibility of buying add-ons in the game after finishing the given level, encouraging him to click a special button. If the purchase took place and we are sure about the source of the traffic, we can accurately convict the conversion. Our actions do not necessarily have to have one desired effect. It is possible for one ad to call for a direct purchase, allowing the opportunity to record for a price alert. Then we can count the conversion for both events. In the event that we achieve the same effect from two different activities, we will also have two separate indicators, this time differentiated by the place of conversion.
ARPU (Average Revenue Per User) tells us what average income we obtained on a single user of a game or mobile application. Like other indicators, this one is also calculated at a specific time, because the MAU should be taken into account here. We calculate it from the ARPU = REV / MAU formula, where REV is the sum of our revenues.
The ARPDAU (Average Revenue Per Daily Active User) is the daily average earnings from the active user. This indicator is vigilantly monitored by the developers of mobile games, because some fluctuations might not be noticed. ARPDAU allows you to precisely express the state of our monetization every day and in contrast to the ARPU, which represents the image on a monthly basis. With this indicator, it is easier to assess our activities as soon as they took place.
We can meet with a similar indicator called ARPPU, which tells us about average income from paying clients. This is an even more precise variation of revenue measurement focused only on the users we monetize, with the exception of those who did not bring us any income in the average.
It is an abbreviation of “lifetime value”, used interchangeably with CLV from “customer lifetime value” or LCV. In the literature, we can find on almost 20 different LTV deflections. A well-known economist, professor at the International Marketing Department at Kellogg School of Management, Philip Kotler defines this concept in the following way: “current value of future profit streams expected to be obtained as a result of transactions with clients in a given time horizon” .TV in a mobile application is used to measure customer value as part of our entire relationship, i.e. from acquisition to the last transaction. This is the most complex indicator from the previously mentioned. At first glance, LTV may seem an indicator similar to ARPU, but it is much more complex. ARPU only considers average revenue when LTV has more components. In the case of mobile application analytics, it is considered that if LTV is higher than the cost of a single installation by a CPI user, we can talk about a return on investment. Just as we can meet some defections for the LTV itself, we can also find several ways to calculate it. One referred to as “average LTV” is obtained by multiplying the ARPDAU by the average retention rate. The downside of this method is its small reliability at the beginning of our business, because this method requires a large user base and a certain time of presence of the mobile application on the market.
Another method of calculation suggests using the user outflow indicator, and the formula for it is as follows: LTV = ARPU x (1 / outflow).
There are other ways to get this indicator. To make this one of the most useful for us, this topic should be explored further, choosing the method in accordance with the intended business goals.
CAC – Customer Acquisition cost
Customer Acquisition Cost or Cost per new Customer, in short called CAC. The use of this indicator can tell us not only how much it costs a single installation by a unique user, but also from where it comes to us.
The easiest way to calculate it is to sum up the total expenses we incur to acquire users and divide them by the number of new clients in the set time interval. Right next to LTV, it is one of the most important indicators for mobile applications, because it helps us determine whether our investment is paying back and the application earns money. Looking at the ratio of LTV to CAC, we can easily assess the state of our business. If this is 1 to 1, we actually lose money. 2 to 1 means that we are on the right path. The customer value status has a ratio to the cost of its acquisition at the level of 3 to 1, which means that our business acts like a well oiled machine. Interestingly, the ratio of 4 to 1 can be considered too high. It may mean that our application or a mobile game is growing at a lower rate than it would be when increasing the costs of acquiring new customers.
I know from experience that usually the highest LTV in a mobile application has users acquired through organic search in the App Store and Google Play.
There are many great tools to run mobile application analytics. Certainly, in the future, a text about analytics tools in mobile applications will appear on our blog.
In the analysis of mobile applications, indicators related to the application’s performance are still monitored. We present here a few of the most important ones briefly.
Network errors – most often related to problems on the side of network service providers for applications. These types of errors can lead to unstable applications, slow performance and slow down or completely disable the application.
Application latency – as not all actions taken by the user take place locally, but also on the side of the network servers, some delay may occur between the user’s action and its result. We can then experience the so-called “Lag”. In addition to issues related to the speed of Internet connection, which is characterized by a certain latency in addition to the delay, which depends on the performance of the device used, how quickly our application can receive and send data packets in distributed systems, affects the comfort of its use.
The speed of loading the mobile application – this indicator refers to the load within the session and can show how much time our application needs for given activities. It may be the time it takes to launch it, move to a given option, or in the case of games, to load a new level. The user may feel discouraged if he has to wait too long for the response from the application, therefore it is necessary to monitor and optimize such elements.
Mobile application failures – this indicator is intended to show how often an application crashes during the start of a certain number of sessions. This indicator is of great importance usually at the beginning of the application launch. Even with very intense tests before the premiere, it may turn out that users experience difficulties that we could not foresee.
Indicators related to devices and type of system – the number of different devices on the market is huge. That’s why even with the Apple system, we can deal with many constructions of models and software versions. By doing this type of statistics, we can focus on optimizing our application for configuration, which our users use most often.
For some of the indicators described above, we can find several variants and variations of how to determine them. Of course, there are many other things that can be useful in the development of our mobile application. The information provided is not exhaustive, it is only intended to provide an overview of the benefits and benefits of learning about our product and its users. Increasingly, you can hear that good ideas are worth nothing. All that counts is their implementation. Even if we are sure that the solutions we offer our clients are exactly what they expect, we should not rely on our own intuition. Therefore, in every aspect, in which instead of guessing, we can look at hard data, collect it, analyze it and use it, for the sake of both our business and its clients.
Acquiring user to a mobile application is the first stage of application marketing. Your goal is to acquire users who will be involved in the application, therefore it is necessary to measure events in the mobile application. It will also allow you to check which channels provide “engaging” users.
Remember the promotion of the mobile application without data analytics in the application and in markets will be pointless.