BigQuery SQL
Queried retention, onboarding, match, extraction, and session events.
Analytics case studyClosed alpha launch analysis
Closed Alpha Launch Analytics
I used SQL-led product analytics to identify why early players were dropping before their first PvP match and where the game experience needed tuning.
My role
Data analytics, BigQuery SQL, funnel diagnosis, retention analysis, insight storytelling
Launch analytics
D123%retentionPvP43%startedStep 10-12%dropM511%started
onboarding funnelSQL
Funnel signalOnboarding friction and first PvP conversion
Project setup
A launch readout focused on early drop-off.
This project analyzed an anonymized closed alpha game launch across retention, onboarding, first PvP entry, match progression, and extraction behavior.
Funnel + behavior
Mapped where players left and how skill, risk, and extraction shifted.
Slides + Granola
Turned the analysis into a concise readout with P0/P1 actions.
Metric snapshot
The early journey broke before the core loop.
Signal
23%
Day 1 retention
7 points below the 30% launch target.
Signal
43%
First PvP start
57% of players dropped before the core loop.
Signal
-12%
Extraction step drop
Largest onboarding friction point.
Signal
11%
Fifth match start
Steep decline after early match exposure.
Early journey funnel
The core loop was leaking before match one.
Install cohort
Baseline
100%
Drop before PvP
Onboarding loss
57%
First PvP start
Core loop entry
43%
Third match start
Early decline
18%
Fifth match start
Retention risk
11%
Diagnosis to decision
Fix the first PvP path before adding more content.
What mattered most
- Day 1 retention was 23%, below the 30% launch target.
- Only 43% of players reached their first PvP match.
- The extraction step was the clearest onboarding friction point.
Recommended actions
- P0: Reduce final extraction difficulty and simplify onboarding.
- P0: Reinforce extraction incentives during matches.
- P1: Add early rewards or streaks to support return visits.
Engagement signal
Players learned to fight, but not to extract.
Across the first ten matches, kills improved and deaths declined, while extraction rate trended down. That points to a risk/reward issue, not only an onboarding issue.
Skill trend
Kills rise and deaths fall across early matches.
Risk signal
Extraction rate declines as confidence grows.
Product implication
Reward extraction more clearly mid-game.
Reflection
Good analytics points to a product decision.
This project helped me practice the analytics skill that matters most in product work: turning behavioral signals into a clear decision.
Privacy note
Anonymized company data.
Company and dataset details are anonymized; the page focuses on method, interpretation, and recommendations.