GA / Platform Data Log Analysis
Bridging Google Analytics and Platform Data to drive product improvements
Project Duration
Q1 2024 - Q2 2025
My Role
As the Lead UX Researcher, I spearheaded the strategy for integrating Google Analytics and Platform Data. I led the analysis of user behavior, including user path analyses, custom metric development, and cohort analysis, to identify key pains and opportunities. My work in designing actionable dashboards and delivering data-driven recommendations directly supported product roadmap prioritization and Weekly Business Reviews (WBRs).
Team
Project Overview
This quantitative UX research initiative unified Google Analytics (GA) session data with Platform Data events to create a comprehensive view of user behavior. Through log analysis, data integration & modeling, user path analysis, custom engagement metric development, and stickiness ratio analysis, we achieved measurable business outcomes: **~11% quarterly improvement** in median engagement scores, **~80% WAU/MAU ratio** indicating strong platform stickiness, and **10+ hours weekly** saved in manual reporting through automated dashboards. The integrated analytics framework enabled stakeholders to make data-driven decisions with unified user telemetry, directly informing UX improvements and product strategy across the complete user journey.
Methods
Tools
The Challenge: Unifying a Fragmented User View
Product and marketing teams operated with a fractured view of the user journey. Web analytics (GA) tracked in-product user behavior, while Platform Data and product databases held in-app engagement data. This separation made it nearly impossible to understand the complete user lifecycle, measure true engagement, or confidently attribute product success to specific development efforts.
Disconnected User Journey
How did users *really* move through the platfrom?
Disconnected User Journey
Mapping the end-to-end user journey from initial touchpoints (GA) to in-product feature usage and Platform Data milestones was fragmented and unclear.
Measuring True Engagement
What did 'engagement' *truly* mean for our product?
Measuring True Engagement
Relying on isolated metrics (e.g., GA page views vs. Platform Data activity logs) provided an incomplete picture of genuine user engagement and product stickiness.
Low Analytics Adoption in Ops
Insights weren't part of weekly rhythms.
Low Analytics Adoption in Ops
Dashboards weren't used in WBRs; decisions lacked data.
The Cost of Siloed Data
- •An absence of cohesive behavioral insight, resulting in strategies misaligned with live product usage.
- •Inability to identify critical friction points and drop-offs across the full user lifecycle.
- •Limited understanding of what truly drove long-term user engagement and value.
- •Wasted resources on unproven initiatives or poorly prioritized features.
Our Research Aimed To:
- •Create a unified Log Analysis framework integrating GA data points and Platform Data behavioral data.
- •Quantify user telemetry, user paths, sequence conversions, and drop-off points across systems.
- •Develop holistic engagement metrics reflecting true product interaction.
- •Enable data-driven decisions for UX, product, and other stakeholder strategies.
Bridging the Data Divide for Holistic Insights
To overcome these critical blind spots, this research initiative focused on integrating disparate data sources. The goal was to build a comprehensive, unified view of the user journey, enabling accurate measurement of engagement and the direct attribution of outcomes to specific interactions, thereby empowering smarter, data-informed decisions across the organization.
Key Research Questions
To bridge the gap between siloed GA and Platform Data data, our research was driven by critical questions aimed at creating a unified understanding of user behavior and its impact.
Understanding User Navigation & Interaction
To map and analyze how users move through and interact with integrated GA and Platform Data/Features.
- Q1:
How can we connect and understand the user journey between Google Analytics and Platform features? What KPIs/metrics are important to introduce and track?
- Q2:
Across GA + Platform data, what do sessions look like by the numbers (pages, paths, time, events, tech), and which patterns clearly indicate friction or high-value moments?
- Q3:
Which specific user roles (identified via Platform Data) show the highest engagement with particular platform features (tracked via GA)?
Measuring (& Predicting) Engagement
To define, measure, and predict user engagement and stickiness based on combined data.
- Q1:
How should we define and track engagement across user/user cohorts (Platform Data) using GA session signals (unique URL visits, session/sequence time) and in-product activity (events)?
- Q2:
What are our stickiness baselines (DAU/MAU, WAU/MAU, DAU/WAU), and where/why are the key drop-offs?
- Q3:
Based on current cross-platform usage patterns (GA & Platform Data), can we build a model to predict future user engagement, stickiness, and potential churn?
Guiding Methodological Choices for Integrated Insights
These research questions were fundamental in guiding our methodological choices. They necessitated an approach centered on data integration (Behavioral Data Modeling), pathway analysis (User Path & Sequence Analysis), nuanced engagement measurement (Custom Engagement Metric Development), and longitudinal tracking to provide a truly holistic view of user behavior across systems.
Research Methods
We used these quantitative techniques to turn raw GA and Platform Data data into actionable insights:
Log Data Integration & Time Series Analysis
Integrated GA & Platform Data for a complete user view
Built a unified data model integrating Google Analytics behavioral logs with Platform Data events, enabling comprehensive analysis of cross-platform user telemetry and interaction patterns over time.
What this means: Connected GA web logs and Platform Data user actions into one model, and looked at the data over time. This provided a full user story, normalizing GA hits and platform events into one schema so we can sequence web actions and in-product events for the same user.
Data Integration Flow
User Path & Sequence Analysis
Mapping user journeys and identifying drop-offs
Quantitatively mapped common user navigation paths by analyzing sequences of page views and Platform Data events to identify high-traffic routes, critical drop-off points, and areas of friction.
What this means: Tracked user steps from landing page to conversion or other key events, locating pathways and identifying drop-offs to prioritize UX fixes and product improvements.
Funnel with 60% Drop-off Identified
Custom Engagement Metric Development
Creating tailored scores for user interaction depth
Developed and validated custom engagement scores by algorithmically weighting session/sequence duration, unique URL visits, and key event sequences to quantify user engagement.
What this means: Developed a composite engagement score combining GA session/sequence duration, unique visit frequencies, and interaction depth. This offered a better measure of true engagement from survey data, and also helped identify power users and at-risk segments.
Inputs to Engagement Score
Stickiness Ratio Analysis
DAU/WAU and WAU/MAU trends by segment
Quantified stickiness using DAU/WAU and WAU/MAU ratios, segmented by role and cohort, to understand how often users return and where habits strengthen or fade.
What this means: Stickiness ratios show how frequently users come back within weeks and months, helping us spot parts of the platform with strong habits and those at risk.
Foundation for Process: The chosen analytical methods (Data Integration, Path Analysis, Custom Engagement Metric and Stickiness Ratio Analysis) were instrumental in executing the research process. They directly enabled the crucial stages of defining data requirements, uncovering user behaviors for insight generation, and ultimately informed the evidence-based dashboard design.
Research Process
Our systematic approach to implementing analytics and uncovering actionable insights
Analytics Audit & PM Usage Challenges
Data Requirements Definition
SQL Metric Building & Data Validation
Insight Generation & Analysis
Dashboard Design & Deployment
Analytics Audit & PM Usage Challenges
Identified current analytics assets and investigated Stakeholders' utilization and adoption challenges.
Data Requirements Definition
Collaborated with Stakeholders to define critical business questions, KPIs, and the specific data needed from GA and Platform Data systems.
SQL Metric Building & Data Validation
Developed new metrics by joining GA and Platform Data using SQL, and rigorously validated the accuracy and consistency of this integrated data.
Insight Generation & Analysis
Analyzed the validated, integrated dataset to uncover actionable insights into user behavior, engagement patterns, and product stickiness.
Dashboard Design & Deployment
Designed, developed, and deployed interactive dashboards to provide Stakeholders with ongoing access to key metrics and actionable insights.
Key Findings
Our analysis of 70+ million data points revealed these critical insights about user behavior
Platform Data Sequence Transition Analysis
This pattern below suggests the dashboard serves as a primary launchpad for, and return point from, Platform Data-related tasks, underscoring its importance in the user workflow for these key activities.
Engagement Trend Analysis
Analysis of median 'Whole Session Engagement Scores' indicated a positive growth in engagement, suggesting increased user acclimatization and value derived from the platform of nearly 11%.
Early Oct '24
24.94
Late Dec '24
27.64
Median Whole Session Engagement Score
Dominant User Interactions
Analysis of over 23.2 million session events highlights core user interaction patterns. This data signifies a highly engaged user base actively navigating content and interacting with page elements, forming the foundational layer of user experience on the platform.
Streamlined Reporting & Efficiency Gains
Integrating multiple data systems into a unified dashboard saved analysts 10+ hours weekly by eliminating manual report compilation from disparate sources. This streamlined workflow significantly boosted pragmatic data analyses.
Key Opportunity
Leveraging integrated GA and Platform Data data provided deep insights into user behavior. Key achievements include an ~11% increase in median user engagement scores (Q4) and a reduction of 10+ hours per week in manual analytics preparation time.
Impact & Outcomes
How thisanalytics integration transformed raw data into actionable business intelligence and measurable results
Data Insights
Measured ~11% QoQ improvement in median engagement, aligned with a stable weekly stickiness baseline (WAU/MAU ≈ 0.80)
Identified dominant user actions (clicks, views, scrolls).
Unified dashboard visualizations revealed critical Platform Data interaction patterns.
Implementation
Blended GA session signals with in-product events, standardized at user/session/product levels, and forwarded via self-serve dashboards
Redesigned user flows based on behavioral analytics
Created automated data visualization dashboards for teams
User Impact
Improved understanding of user engagement and stickiness.
Enabled data-informed UX and product strategy decisions.
Increased team adoption of data for informed actions.
Elevated Engagement
Observed in Q4, indicating improved user acclimatization and perceived platform value.
Time Efficiency
By analysts due to unified dashboard, eliminating manual data compilation.
WBR Dashboard
Over 90% of the Product teams adopted their respective dashboard for their Weekly Business Reviews (WBRs).
Roadmap Analytics Transformation
Before Integration
10+ hours per week manually gathering analytics
Data silos between data, engineering and product teams
Very low percentage of decisions backed by user data
After Integration
Majority of analytics work automated through dashboards
Unified data platform accessible to all teams and stakeholders
Zero key decisions made without the backing of data
Visuals
User Engagement Overview
Shows overall session counts, role-based breakdowns, and page frequencies.
Use stacked bars for time-based trends; see the pie chart for client usage share.
Identifies which roles or clients drive the bulk of engagement.
