Created with Sketch.

Best practices in App Analytics

When we start with app analytics we often fall into a trap thinking that all data is information. The data itself is not information. Today there is much talk of Big Data and thanks to some tools we are able to collect a lot of data from our users and use it for our mobile application. However, we have to know how to interpret that data.

 

To do so, I recommend that you follow these 4 Best Practices that we have defined and validated with several players in the app industry:

 

  1. Define the most appropriate metrics to analyze what you want in your application

In the last article we discussed the generic metrics that we need to start analyzing our apps. However, each application must have its own metrics. How can we discover the most suitable metrics for our app? Lets’s see a practical case: we have a photo application specialized in sporting events. This app works as follows: I go to a soccer match, I take a picture, I the edit it and I share it with my friends. Let’s call it LetsFreezeit.

How do we know what are the most appropriate metrics for LetsFreezeit? Start by asking yourself the following questions:

  • What is the purpose of my application? In this case, the objective of LetsFreezeit is to be the tool that helps you share your best sports moments.
  • How do I want users to behave within the application? LetsFreezeit wants to be a very easy to use app. What do we want is for users to take pictures and to share them. Both actions are our priorities for the app.
  • What is the conversion objective? This application is free. Its main objective is to get more users through the viral growth. If your application contains in-app purchases then probably the goal will be that your users buy these in-app purchases.

Based on these questions we define what are the most appropriate metrics for our app analytics. For LetsFreezeit, the duration of the session will not be a key metric, however, recurrence will be an important one.

 

  1. Create events

What is an event in a mobile application? It is any action that our user has done. If we know clearly the purpose of the app, the events will be very easy to define. And these events will help us know the path that our users are following. Do they skip any of these events?

In LetsFreezeit we have four events:

  • Choose the match
  • Take picture
  • Edit picture
  • Share

 

  1. Define the process of your app (funnels)

The funnels are those processes within our application that we want our users to follow to achieve a defined objective. The funnels or processes are defined by a succession of events that we have defined in the previous step.

In LetsFreezeit the funnel would be as follows:

app-analytics-letsfreezeit-egrowing

We defined the «share» event as conversion objective. This is a very simple and streamlined process. Each application has to find theirs. If we follow these funnels we’ll find any malfunction or bugs. For example: If we realize at some point that users do not pass the first event we may think that there is some error. And if you don’t share the photo through the app, we may think that the process is too complicated.

 

  1. Segment your users

Are all your customers equal? No. Surely there are users with multiple profiles and furthermore, each one uses a different device and a different network. If we can segment the users, we’ll find trends in their behavior and we will be able to identify how each group of application users behaves.

 

We can segment users based on:

  • Location
  • Device
  • Do they buy? (in the case of LetsFreezeit, it would be: Do they share?)
  • We can also segment our users depending on their use of the application.

Segmentation also helps us improve your engagement:  we can launch campaigns only for some users geo-located in the part of the worlds that we are targeting.