Track user behavior across iOS and Android applications with native and embedded web views to understand user engagement and retention.
Influence business decisions using accurate and granular insights that would have previously taken significant technical resources to discover.
Deploy the Advanced Analytics for Hybrid Mobile DPA to start tracking events both inside the web view and native app code. You can load the data into a Snowflake warehouse, and model it into higher level entities such as screen views, sessions, or users.
Then, you can slice and dice as you need to access valuable customer insights, faster.
What to expect
This accelerator helps you build a deeper understanding of customer behavior on your mobile apps, so you can use data to influence business decisions. It uses the following tools:
- Snowplow – to track events across both the native app code and web
- Snowflake – to house the tracked events
- snowplow-mobile dbt package – to model raw events into higher level entities like screen views, sessions, or users
- Streamlit – to visualize the modeled data
Note: This guide uses Snowflake, however the snowplow-mobile dbt package also supports BigQuery, Databricks, Postgres, and Redshift.
The accelerator takes about 13 working hours to complete, and includes the following steps:
Step 1. Upload the data
Upload a sample Snowplow events dataset to your Snowflake warehouse.
Step 2. Model your data
Configure and run the snowplow-mobile data model. Modeling the data makes it easier to digest and derive business value from the Snowplow data either through AI or BI.
Step 3. Visualize the data
Use Streamlit to visualize your Snowplow data, to make it easier to identify patterns and trends in your data.
Step 4. Set up and deploy tracking
Mobile hybrid apps implement some app logic in the platform native code (e.g. in Swift or Java), and some in embedded Web views. Here, you learn how to instrument tracking for both events.
Step 5. Modeling your own pipeline
Now that you have set up tracking and enrichment on your pipeline and generated some test events, you’re ready to model and visualize your own pipeline data.