Comparative Metrics Dashboard: Enhancing Application Insights 📊

A comprehensive overview of a dashboard designed to compare critical metrics between landscape and portrait application versions.

1. Introduction to the Comparative Metrics Dashboard 🔍

The advent of mobile applications has necessitated the development of robust analytics tools to monitor and compare performance across different versions. Our comparative metrics dashboard serves as a pivotal tool for founders, product managers (PMs), and data analysts, allowing them to evaluate engagement, retention, stability, and user activity across both landscape and portrait app versions. By centralizing this data into a single view, the dashboard empowers stakeholders with actionable insights to drive decision-making.

2. Overview of Data Integration and Queries 🔗

Initially, the landscape application's event data was collected and stored within our data warehouse utilizing BigQuery. With the launch of the portrait application, a strategy was implemented to integrate Firebase event data—enhancing our analytics capabilities.

The integration process involved establishing a connection to Firebase events data and importing it into BigQuery, enabling detailed reporting of user interactions. The foundational step in this process was crafting SQL queries that extract and compute metrics for both app versions, which were then populated into a dedicated table for straightforward analysis.

Sample Query for Engagement and Session Report:

SELECT DISTINCT user_pseudo_id,
TRUNC(SUM(value.int_value / 60000), 3) AS time,
(SELECT value.int_value FROM UNNEST(event_params) WHERE key = 'ga_session_number') AS session_num,
DATE(TIMESTAMP_MICROS(event_timestamp), "Asia/Karachi") AS date,
'landscape' AS mobile_app
FROM `database1.table1.events_*`, UNNEST(event_params)
WHERE event_name = 'user_engagement'
AND key = 'engagement_time_msec'
AND _TABLE_SUFFIX = FORMAT_DATE('%Y%m%d', DATE_SUB(CURRENT_DATE(), INTERVAL 1 DAY))
AND user_pseudo_id IS NOT NULL
GROUP BY session_num, date, user_pseudo_id

UNION ALL

SELECT DISTINCT user_pseudo_id,
TRUNC(SUM(value.int_value / 60000), 3) AS time,
(SELECT value.int_value FROM UNNEST(event_params) WHERE key = 'ga_session_number') AS session_num,
DATE(TIMESTAMP_MICROS(event_timestamp), "Asia/Karachi") AS date,
'portrait' AS mobile_app
FROM `database2.table2.events_*`, UNNEST(event_params)
WHERE event_name = 'user_engagement'
AND key = 'engagement_time_msec'
AND _TABLE_SUFFIX = FORMAT_DATE('%Y%m%d', DATE_SUB(CURRENT_DATE(), INTERVAL 1 DAY))
AND user_pseudo_id IS NOT NULL
GROUP BY session_num, date, user_pseudo_id;

This query unifies engagement data for both application versions, allowing for a comparative analysis of user interaction metrics.

3. Metrics Included in the Dashboard 📈

A well-rounded dashboard is critical to provide clarity at a glance. Metrics included in this comparative dashboard feature:

• Engagement and Session Report: This provides insights into sessions per user, engagement time, and trends over daily and weekly intervals. This report allows PMs and founders to observe how user interaction differs between the landscape and portrait applications.

• Retention and Crashlytics Report: Monitoring user retention rates alongside stability metrics from Firebase Crashlytics offers valuable perspectives on application performance. This report is vital not only for PMs but also for lead software engineers who focus on ensuring app reliability.

• Activity and Funnel Report: This report details key activities, such as the number of videos watched and tests completed by users, empowering teams to understand user behavior comprehensively.

4. Initial Findings and User Engagement Trends 📊

During the initial phases post-launch, the landscape application demonstrated superior engagement metrics compared to the portrait version. However, as the user base for the portrait application expanded, a noteworthy shift occurred: engagement levels in the portrait app began to exceed those of the landscape version.

This evolution suggests that while initial impressions can be misleading, sustained user engagement may ultimately favor the newest version when equipped with appealing features. The insights obtained from the dashboard confirmed the hypothesis that the portrait app would achieve higher user adoption and engagement over time.

5. Conclusion and Best Practices for Comparative Analysis 🎯

Implementing a comparative dashboard not only enhances visibility but also helps justify strategic decisions regarding product development, marketing, and user retention efforts. The process of validating event data before constructing queries is essential to ensure accuracy in reporting.

In summary, the comparative metrics dashboard serves as a fundamental resource that facilitates data-driven decision-making and fosters a culture of continuous improvement across application versions. This tool exemplifies the blend of technology and data strategy necessary for modern app development in a highly competitive landscape.