I ran a set of experiments on player stickiness and game flow in Merge & Splash. The video below walks through the details; the short story is that the latest update hurt retention and engagement, and the game no longer looks economically viable.
Video: https://www.youtube.com/watch?v=7CIhEAGvQwg&t
Android Game: https://play.google.com/store/apps/details?id=com.gamedrobe.dropmerge
iOS Game: https://apps.apple.com/us/app/drop-merge-journey/id6711361562
TL;DR
Before the Update
- D1 Retention: 33–34%
- D7 Retention: 11–12%
- D28 Retention: 2–3%
- Avg Session Length (D0): 19–20 min
- Sessions / Player / Day: ~2.8
- Daily Ad Impressions: 3,020
- Daily Revenue: $9.18
- CPI: $1.45
After the Update
- D1 Retention: 27–28%
- D7 Retention: 7–8%
- Avg Session Length (D0): 8–9 min
- Sessions / Player / Day: ~1.9
- Daily Ad Impressions: 610
- Daily Revenue: $1.85
- CPI: $1.32
These results clearly show the game is not profitable. Post-update revenue barely covers the cost of a single user.
What Changed in the Update
The original game was endless mode only. I converted it to a level-based structure, added lots of levels, a stronger sense of progression, bonus systems, and a cleaner first-time user experience. In hindsight, these additions negatively impacted the core loop—they diluted what made the endless mode feel simple and sticky, without creating a better long-term goal.
Compounding this, many players kept choosing endless mode even after the level update. I tried bringing levels to the forefront and acquiring a new cohort to validate the change, but the analytics worsened.
Financial Summary
- Total revenue (ads + IAP): ~$1,087
- Total spend (Meta, AdWords, other): ~$2,894
Given the trajectory, I’ve decided not to continue investing in this title. The endless mode started promising but couldn’t support enough ad inventory per session; the level mode hurt retention and didn’t lift monetization.
Try It & Share Feedback
The latest update is live on Android (levels) and the iOS build still includes the endless mode. I’d love your perspective:
- Why do you think the update underperformed?
- If you played both versions, which loop felt better—and why?
Please leave thoughts in the comments. I’ll read everything and carry the lessons into the next project.
