1. Make an Analytics Plan
Start thinking about Android analytics before you release the first version of your app. Development goes through phases: specification, prototyping, design, implementation, testing, deployment to the Google Play store, user installation, and feedback. Make analytics something you integrate with your mobile testing program.
2. Decide Which Metrics Matter Most
Android mobile analytics tools offer perhaps hundreds of quantitative measures about app usage—counts, revenue statistics, session durations, events, crashes, etc.—plus data on operating systems, device models, and geography. Define the most important metrics that need to be reported. If ecommerce is the goal, you’ll want to know about conversion (number of visitors who buy), shopping cart size (dollars spent in a purchase), and abandonment (number of shoppers who start a purchase but don’t check out). If in-app revenue is the sole way you make money, key metrics are the number of downloads, ratings, daily/monthly active users, and purchases of the premium version of your app. Consider, too, how often the app should report each data element.
3. Don’t Forget About Usability
To get a more holistic view of how users engage with your app, define additional qualitative metrics to look at user interface elements such as design, flow, and the intuitive, creative ways people complete tasks and follow paths in your app. Your Android app analytics should tell you whether and how users complete tasks (effectiveness) and how quickly they reach their goal (efficiency). If they are not completing a task, what are the pain points, the areas where users get stuck or encounter problems that cause them to quit the app? Visual analytics such as videos of single user sessions or touch heatmaps that look at user behavior in the aggregate are helpful for evaluating usability issues.
4. Pick and Install Your Tool(s)
There are dozens of Android mobile analytics platforms to choose from. Pick at least two so that you can experiment with and implement different Android app analytics. All solutions require you to use an SDK and have some level of customization. Read the example code to see how easily you can integrate the SDK and implement the Android analytics tool into your app. Then, start tracking and collecting data. Remember, too, to tell users (in your terms and conditions) that the app is going to share information about how it is being used.
5. Know How to Use the Data
The world of Android analytics is a Goldilocks zone: Pick too many metrics and you can overload the analytics system and increase complexity. Pick too few metrics or the wrong ones and you won’t gain the insights you need. A “just right” approach for getting results that can improve your app might be to use unique qualitative data from a visual analytics tool to complement the more standard data logs and reports from another analytics tool.
