Design and evaluation of a product analytics framework for a wearable health tracking application
Mosina, Julia (2026)
2026
All rights reserved. This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:amk-202604095927
https://urn.fi/URN:NBN:fi:amk-202604095927
Tiivistelmä
This thesis focuses on the design and evaluation of a product analytics framework for a digital health application integrated with a wearable device. The aim of the study is to develop a structured analytical model that enables systematic analysis of user behavior, engagement, retention, and monetization.
An event-based tracking model was designed to capture meaningful user actions, and a hierarchical metric structure was defined, including the North Star Metric — Weekly Insight Active Users (WIAU). To demonstrate the functionality of the framework, a synthetic dataset simulating the behavior of 1,000 users over a 120-day period was generated. The data was processed using SQL and dbt within a Snowflake environment, enabling the implementation of a layered analytical architecture.
The practical part of the study includes funnel analysis, cohort-based retention analysis, and subscription conversion evaluation. In addition, logistic regression was applied to examine the relationship between user engagement and subscription probability. The results indicate a statistically significant association between the depth of interaction with analytical features and the likelihood of subscription.
The proposed framework demonstrates how product analytics can move beyond descriptive reporting toward explanatory and predictive decision support. The developed structure can be adapted to other digital services that rely on event-based data models.
An event-based tracking model was designed to capture meaningful user actions, and a hierarchical metric structure was defined, including the North Star Metric — Weekly Insight Active Users (WIAU). To demonstrate the functionality of the framework, a synthetic dataset simulating the behavior of 1,000 users over a 120-day period was generated. The data was processed using SQL and dbt within a Snowflake environment, enabling the implementation of a layered analytical architecture.
The practical part of the study includes funnel analysis, cohort-based retention analysis, and subscription conversion evaluation. In addition, logistic regression was applied to examine the relationship between user engagement and subscription probability. The results indicate a statistically significant association between the depth of interaction with analytical features and the likelihood of subscription.
The proposed framework demonstrates how product analytics can move beyond descriptive reporting toward explanatory and predictive decision support. The developed structure can be adapted to other digital services that rely on event-based data models.