Fix your D0. Grow LTV.
Send us your player events, and your acquisition data if you track it. You get back where players drop off in their first week, which UA campaigns bring the players who stick around, and the A/B test to run next. One game, one report, no contract.
Risk-free: money back if it’s not useful · built by Immutable
What you get
A real report you can act on.
A real excerpt from the report. Each finding shows why it happens, what we ruled out, and an A/B test to run.
17,826 players churn right after their first dungeon loss. 77.2% churn, $24.6K at risk.
Retained players use the quest system 3.2× more after their first dungeon. This segment's quest engagement is 167× lower, so they're missing the feature that keeps other players around.
They just don't like losing
Retained players lose dungeons at the same rate and don't churn. The loss itself isn't the problem, missing the quest system afterward is.
Send this segment a daily quest reminder (in-app + push); hold out 20% as a control group.
Broad-network video ads bring cheaper installs but churn 2.6× faster than playable ads.
Video-ad players skip the tutorial's story beats at 3× the rate of playable-ad players and rarely return past day 2. Playable-ad players, already familiar with the core mechanic before they install, retain almost as well as organic.
It's just cheap, low-quality traffic
Fraud and incentivised-install scores for this channel are normal, and CPI is in line with the rest of the network. These are real users who bounce because the ad's promise doesn't match the tutorial.
Move 25% of the broad-network video budget into an expanded playable-ad campaign; hold the current spend mix as a control group for 30 days.
Whale stall at D14: a power gate players can't get past.
This segment's login frequency drops 60% right at the level-14 power gate, before any drop in spend, which means they're stuck at the gate rather than out of money.
They've stopped wanting to pay
Spend didn't drop before the login drop, only after. If they'd lost interest in paying, spend would have fallen first. It didn't.
A personalised battlepass extension offer, 7 days, run as an 80/20 control split.
FTUE bounce. 0% ever start the first dungeon.
6,451 players sign in 8.49 times on average but never start the first dungeon. They keep coming back without ever reaching the part of the game that hooks people.
They've lost interest in the game
They keep coming back at 8.49 sign-ins on average. People who've lost interest don't return almost daily.
Send this segment a win-back deep link straight to the dungeon lobby with a starter guide overlay; hold out 20% as a control group.
The method
What happens to your data, step by step.
These are the four steps between your data landing and the report going out.
Clean the data
We resolve duplicate identities and inconsistent event names first, so one player isn’t counted five times.
Build cohorts
Players are grouped by cohort and checked against retention curves from similar games.
Find the divergence point
Event-sequencing shows the exact session, screen or mechanic where retained and churned players’ paths split.
Rank by dollars
Every finding is sized by players and revenue at risk, then reduced to the one test worth running first.
Acquisition data, tied in
We can ingest acquisition data from whichever MMP you run. Send it alongside your event data, and we’ll show you which campaigns and creatives bring players who stick around.
Works with, e.g.
How it works
Three steps. Three business days.
01
Tell us about your game
Name it, link your store or game page, and place your order. No contract, no call required.
02
Send your data
Export the last 30–60 days of player events, whatever your stack gives up easily. We’ll email you exactly what to send and where.
03
Get your report
Three business days from when we have your data. Where players drop off, why, and one test to run.
Money back if the report doesn’t tell you something worth acting on. This is a one-off audit. If you want it running continuously on live data instead, get in touch and we’ll talk through what that looks like.
Frequently asked questions
One-off diagnosis
Get your report. $99.
One game, one report, three business days once we have your data. No contract, no call required. Money back if it’s not useful.
Risk-free · Payment via Stripe · we’ll email you next steps
