When creating a mobile application, a developer imagines a model and the way users will use the application. One problem that developers face is that users do not always use an app the way it was envisaged by the developer.
How do users interact with the app? What do they do in the app? Do they do what the developer wants them to do? Mobile analytics help to answer these questions. Analytics allow the developer to understand what happens with the app in real life and provide an opportunity to adjust and improve the app after seeing how users actually use it. To put it simply, analytics is the study of user behavior.
With this article, we will compare some of the most popular mobile analytics systems. The process of adding analytics to an app involves consideration of many details, and our aim is to provide you with useful tips on implementing analytics. This information should help you find a mobile analytics system that fits your needs and should help you to properly implement it in your app.
Analytics In Real Life
Let’s use as an example a small iOS application that we developed. It’s called What I Eat , and it’s intended to track the user’s eating habits.
Users can track their meals, check the daily meal log, and switch between days in the calendar to review previous logs. The application has an advertisement banner, but users can pay to disable it.
When designing What I Eat, our primary focus was to let the user easily add new meal records and easily review their daily history of meals. We also wanted to monetize the application with an in-app purchase to remove the advertisement. To understand whether we have managed to do this, we track the following events in the app:
when the user starts the app the first time (application installation),
when the user opens the daily meals list (main application screen),
when the user adds a new meal record,
when the user makes an in-app purchase to remove ads.
A view of the meal list, the “add meal” screen, and the settings, with the ability to remove ads. (View large version) Later in this article, we will show how we use analytics to determine whether users have started using the app, and what percentage of users start tracking meals after installing the app.
Comparing Analytic Services
Today, plenty of analytics services are on the market, ranging from well-known systems such as Google Analytics to niche tools. Analyzing and comparing all of them would take forever; so, for this article, we will go with just those that we have found the most convenient. That is, we chose ones whose dashboard interface and data-mining toolbox are relatively easy to understand and easy to work with for those who do not have much experience with analytics, like our clients. As mobile-oriented analytics systems, they are also convenient from a development perspective because the analytics code can be easily implemented and tuned in a mobile app. Here are the systems:
Flurry by Yahoo
Answers by Crashlytics
To analyze how What I Eat performs, we use two main tools that almost every analytics system provides: events and funnels. Events describe what users do in the app, while funnels allows for a qualitative analysis of this data. Let’s examine how each of the systems implement these for What I Eat.
Mixpanel allows you to track custom events. The developer can add custom parameters to the events and use these parameters to segment conversion funnels.
We built a funnel that includes two events: “Install” (which indicates the initial launch of the app after installation) and “Add Meal” (which tracks each time the user adds a meal). These show us what percentage of users not only downloaded the app but also started using it. The conversion is estimated at 65%, which means that out of 100 people who installed the app, as many as 65 started tracking meals.
Conversion from app installation to “Add new meal” in What I Eat (View large version) Sometimes a developer needs events to appear in the analytics dashboard in real time or with minimal delay after they have happened in the application. For example, a developer may have launched a social media marketing campaign and needs to track how it affects their application in real time. Mixpanel shows events almost in real time. Newly created funnels are calculated and visualized almost instantly.
Right after the developer adds Amplitude’s software development kit (SDK) to their project, and without any further set-up of events or funnels, the software starts tracking daily and monthly active users (DAU and MAU) data. We use that a lot in What I Eat to understand how many people use the app each day.