Diagnosing Day 30 Churn: Why are they leaving?

---

Day 30 churn is the silent killer of free-to-play games. Users survive onboarding, engage for weeks, then quietly disappear after they've consumed initial content but before they've formed lasting habits. The team knew users were churning around D30, but not why. Generic answers like "they got bored" don't build roadmaps. My challenge was to diagnose the root causes in a way that was actionable: not just what users felt, but what behaviors and attitudes predicted abandonment and why.

The Approach: Behavioral Truth First, Attitudes Second

One of toughest questions to answer is "why did you leave?". In my experience most people rationalise post-hoc. They don't actually know or can't articulate why they stopped and so they construct a narrative that sounds reasonable that goes something like,

"Uhh, I got bored",

"It got super repetitive",

"I got busy".

I structured the research a little differently before so we weren't collecting data blindly.

1. The behavioral reality (Log analysis, n=5,000) - 2 weeks

• Identify what churners did differently before they left

• Spot the inflection points where paths diverged

2. Then ask why (Interviews, n=10 x 3) - 4 weeks

• Churners (hadn't played in 21+ days): Why did they stop?

• Active users (played 14+ days, still engaged): What keeps them coming back?

• At-risk players (flagged by data science as likely to churn): What's creating friction right now?

3. Understanding patterns (Survey, n=1500 x 3) - 3 weeks

• Once I had specific behavioral patterns, I could form hypotheses to test

• Compared all three groups to isolate factors

I focused on users who had both survived early onboarding. This isolated late-stage churn mechanisms rather than mixing in Day 1-7 drop-off issues. The underlying drivers and therefore solutions aren't always the same.

---

The Insights

Behavioral pattern from logs

There were many differences between churners and non churners in terms of engagement, ad watching, monetisation but what stood out the most was, Churners stopped engaging with live events then abandoned by Day 30. Event non-participation was the red flag, but more a symptom, rather than a diagnosis.

The "connect the dots" moment from qual and survey

Our core audience was students and young adults. Events triggered during school and work hours when most users couldn't participate. This created a compounding failure loop:

1. Miss event due to real-life schedule

2. Fall behind in progression/rewards

3. Feel left behind by active players

4. Lose motivation to catch up

5. Stop logging in.

Sad story.

It wasn't that they didn't want to play, the content cyclet was fundamentally out of sync with their daily rhythms.

The second pattern

Churners plateaued at similar progression points where the narrative and story arc pinnacled. Active users created self-generated goals (collecting, leaderboards). Churners described the game as "repetitive" and "aimless" after the story peaked (reaching the A-list in this case). Without intrinsic motivation and a compelling narrative or self-directed goals, the only driver was extrinsic rewards (events) which wasn't attracting continued participation.

This wasn't a content volume problem. It was product life fit and architecture of motivations.

---

Business Impact

Product Changes

• Event timing shifted to evenings/weekends aligned with student schedules

• Story-based smaller events introduced to extend narrative arc and create long-term goals that increased both revenue and engagement by ~7% for highly engaged users

  
  

Measurable Outcomes

• 12% improvement in late-stage retention overall among previously churned users who re-engaged

• Event participation increased significantly among target demographic post-timing adjustments

• Player progression depth improved as story events sustained motivation and gave them new monetisation sinks.

  
  

Cross-Functional Strategic Impact

• Data science team built churn prediction model flagging at-risk players based on additional behavioral patterns from this research

• Multi-pronged intervention strategy deployed: at-risk players receive tailored event notifications, re-engagement content, progression assistance

• Behavioral churn indicators now integrated into standard product dashboards

• Evangelized UX research's role in driving business outcomes across data science and BI teams

Churn butter not users