Introduction
Application analytics is a relatively new field that is populated by several players. These services are aimed at helping you to get more business insight into the way your apps are being used. The app usage data that you can collect and analyze will give you measurable values regarding your mobile applications.
These tools are particularly useful for developers because in so many cases, once the applications are distributed to users, it turns out that many or most users might not use the app in exactly the way that the designers and developers had anticipated. This really shouldn’t be so surprising when you think about it: Developers think in a way that is different from much of the general public.
The developer mindset is well known and documented within the industry. Programming is a methodical and tedious process that requires a certain personality type as well as a certain skill set. Because the developer knows the app so well, he/she may not do a good job of making the apps features and functions easily accessible to the average user, who doesn’t come to the app with a previous understanding of how it works. For this and other reasons, information that can be garnered from analytics systems can be extremely helpful to developers to get feedback for making the applications better.
What mobile analytics services and products do
Analytics can provide different types of information, which is useful for different purposes and different types of applications. This is important because the metrics that are important for one type of app might be less so for another. Is your main concern whether users return to use the app frequently? Are you more concerned with whether they make purchases from within the app? And so forth. You want to tailor the information to that which is most relevant for your particular app.
Analytics can measure things such as number of downloads, number of new users, user retention, in-app purchases, conversions, app crashes and more. It can assess how engaged users are with your app and how they interact with it. It can help you to see which versions of your app are being used and user distribution over different versions.
Competitors in the mobile analytics space
There are a number of popular mobile analytics tools that are vying for top spot in this market. Google Analytics for Mobile Apps is probably the best known. It supports both iOS and Android apps, with the ability to measure number of users, user characteristics, where users come from, the actions they’re taking, in-app payments and revenue, user navigation paths and behavior of different user groups. A big advantage of Google Analytics is that it’s free.
Another popular tool for analyzing mobile app usage is Mixpanel. It offers free tiers up to 200,000 data points, with monthly cost of $150 to $2000 for 500,000 to 20 million data points. A data point is the tracking of an event. Mixpanel is event-based and measures defined events and event properties. You can create complex queries based on events and demographics, and can also tie mobile to web data and tie data to specific users.
Countly is an open source mobile analytics application with a community edition that you can host yourself (free) or you can use the professional, business or enterprise edition for which pricing starts at $125 per month (500,000 sessions/events with an unlimited number of apps and users). It’s known to have a very user-friendly interface and gives you app information on users and user loyalty, sessions, countries, devices and app versions and carriers and platforms.
Localytics is another “freemium” tool (free tier up to 10,000 Monthly Active Users/1 million data points) with pricing for more MAUs/data points ranging from $200 to $600 per month. There is also an enterprise version with custom pricing. It provides apps by location, device, unique users, app version and carrier, users and sessions, heatmaps, screenflow, funnels and segments.
What Amazon Mobile Analytics brings to the table
Amazon Mobile Analytics provides AWS subscribers with application usage summary charts and lets you automatically export the analytic data to Amazon S3 and Redshift. You can also archive all of your data in S3 buckets, with Auto Export. You can even combine the analytics data with other sources of data, and you will be able to use SQL queries to directly analyze your exported data.
Mobile Analytics is part of the AWS Mobile SDK. You can use it with iPhone/iPad apps, Android apps, and it also supports Fire OS (Amazon’s own Linux-based operating system that runs on the Fire Phone and Kindle Fire tablets) and Unity, as well as with HTML5 and Javascript apps (with the Amazon Mobile Analytics SDK for Javascript).
One important benefit of AMA is its speed. Some Analytics tools may take hours or more to deliver usage reports. That’s fine in some cases, but in others where you need the information ASAP for real-time problem-solving, it’s not. AMA can deliver usage reports to you within just one hour of receiving the data from an app.
Comparing the pricing to other services, AMA follows the practice of a free tier plus paid tiers, but its free tier extends farther than most, up to 100 million free events per month. Then if you need more than that, the cost is $1 (one dollar) per million events per month. You don’t have to sign a contract and there are no charges per active user. You only have to pay for what you use, based on the number of events.
The user interface is easy to use. The Amazon Mobile Analytics Console displays for you a graphical view of your reports on app usage, and you also use it to download data (which is in comma separated value format so that you can easily work with it in any spreadsheet or database). With Auto Export, you can archive the raw data about events for your app on an ongoing basis.
Depending on your needs and how much data you work with, you have the choice of exporting to an S3 bucket or a Redshift cluster. Redshift is Amazon’s data warehousing solution that can scale up to petabytes of information. Whichever you choose, the setup process will be a little different, but either way you’ll need to configure an S3 bucket (new or existing) and configure permissions. Exporting to Redshift is a bit more complex, as you’ll need to configure settings for connecting to the Redshift cluster and EC2 instance, and there are different ways to connect to a Redshift cluster. You can find detailed step-by-step instructions for using Auto Export with either an S3 bucket or S3 plus Redshift on the AWS documentation web site here.
Building AMA into your apps is done with the AWS Mobile Analytics Software Developers Kit (SDK). You’ll need to add your app through the AMA console. Then using Mobile Analytics to record events will depend on what type of app you have.
You’ll need to create an Amazon Cognito identity pool for unauthenticated users, which you do using the Cognito console. You’ll need to enable access to unauthenticated users, create the pool and create new roles for authenticated and unauthenticated identities. Then you can download the starter code (assuming your app is either for Android or iOS) and use it to integrate Cognito with the app. You’ll need to download the SDK and integrate it into your app, as well. More information and the code for doing this are in the AWS Mobile SDK Developer Guide.
You’ll need to have the right permissions before you can use AMA. You’ll need to set permissions via the AWS Identity and Access Management service. Amazon provides an IAM policy that you can assign to roles that are associated with the Cognito identity pool for your app. IAM is part of your AWS subscription; there is no extra charge to use it (you are charged when your IAM users use other AWS services, though). For more information about IAM permissions and policies, see the IAM User Guide.
You will be able to see your data in the Mobile Analytics console. You can view or download the AMA reports via the AWS Management console. The overview tab shows you summaries of several different metrics, including daily and monthly active users, new users, sticky factor (the fraction of monthly users who use your app on a specific date), total number of sessions (uses of your app on a specific day), day 1 retention (percentage of new users that used your app two days in a row) and average revenue per paid daily active user.
There are additional tabs that show you more detailed information about each of the following: Active Users, Sessions, Revenue, Retention, and Custom Events. You can filter the displays according to application, date range and platform. You can display the data in different ways or download the data in a CSV file.
You can keep up with what’s going on with Amazon Mobile Analytics and find out about newly added features and tips for using it in the Mobile Development blog on the Amazon web site.
Summary
Amazon Mobile Analytics is one of many available tools for collecting and analyzing data pertaining to the usage of your mobile apps. It offers some benefits over many of the others, including a very generous free tier pricing structure. Perhaps most important, it is integrated with your Amazon web services. If you need more insight into users, platforms, retention and other app data, it’s easy and free to get started with AMA.
