How Metamoki used Kohort to catch a bad attribution signal and grow revenue 30% in a month
Metamoki, a US-based idle game studio with 50%+ organic installs, used Ktrl to identify an incrementality blind spot that was inflating campaign performance, fix the targeting, and scale the right campaigns with confidence.
Jukka Hilvonen joined Metamoki as Director of Growth in November 2025. Three months in, he was trying to understand which acquisition channels were truly driving incremental growth and which were simply benefiting from Weed Inc's large organic install base.
The challenge wasn't campaign optimization, but rather the measurement of their performance.
Metamoki runs Weed Inc, an idle game with a large organic install base accounting for over 50% of all installs. When revenue from those organic installs gets distributed proportionally across paid campaigns, low-CPI campaigns that reach organic-heavy audiences look better than they are. At Metamoki's organic rate, full blended ROAS was inflating performance on exactly those campaigns.
Before Ktrl, Jukka was estimating incrementality manually. He paused campaigns in individual countries, measured changes in organic installs, then built network-specific assumptions from those results. The process was time-consuming, difficult to scale, and constantly required new testing as campaigns evolved. Ktrl replaced this manual process with a consistent incremental measurement framework. Instead of spending time running geo holdouts and maintaining attribution assumptions, Jukka could focus on creative strategy, bidding decisions, channel diversification, and budget allocation.
With incrementality insights available continuously rather than through periodic manual testing, campaign decisions became faster and more confident.
With Ktrl, some of the campaigns on emerging growth networks now re-evaluated without the organic credit started showing genuine signals. Seeing them still trafficking at a new, more demanding, ROAS target, he was able to scale them up confidently.
Meanwhile, spend on Metamoki's dominant network showed at the campaign level that scaling was harder than the network average suggested. Budget moved between campaigns and networks, with spend increasing ~15%. With it, revenue grew 8% in January and 30% in February.
Altogether it confirmed the biggest benefit wasn't a single campaign decision. It was the ability to spend less time validating measurement assumptions and more time improving acquisition performance. As a result, the growth team could focus on optimization and scaling opportunities rather than constantly rebuilding incrementality models by hand.
In Jukka's words
Early on, Ktrl caught something I'd missed. Our full blended targets were letting organic inflate the numbers, making some campaigns look stronger than they actually were. They recommended we switch to incremental blended at a lower target, and when I actioned it, that was part of what drove a 30% revenue jump in February. Now Ktrl is the screen we pull up in every weekly UA sync. My finance lead and our exec team have access too, so I spend less time building decks and more time making decisions.
Why incrementality mattered
With 50%+ of installs arriving organically, any blended ROAS calculation that doesn't separate organic from paid is measuring something that doesn't exist. Campaigns get credit for revenue they didn't generate. Budget follows the wrong signal.
Switching from full blended to incremental blended removed that credit. It made the actual paid performance of each campaign visible, some for the first time.
Why Ktrl is now the weekly sync screen
Metamoki's team was stretched thin across the day-to-day challenges of running data-driven paid UA, with Jukka running UA while Jeff, the studio's data scientist, is split across product and marketing. Building and maintaining campaign-level LTV predictions in-house was no longer a realistic option to scale Metamoki's paid UA efforts.
Ktrl gives Jukka campaign-level predictions daily, with confidence scores. High confidence, act. Low confidence, wait. The exec and finance team have direct access, so Jukka spends his time on decisions, not on building slides to explain them.
Before, we had no scalable way to understand incrementality, so I was forced to run manual geo holdout tests and build assumptions by hand. Ktrl automated that process and allowed me to spend my time on scaling and optimization rather than measurement.
Key Results
- 30% revenue growth in February after switching an emerging growth network from full blended to incremental blended targets and scaling it in full confidence.
- ~15% increase in paid spend with a 30% revenue return, driven by budget moving from dominant, fully scaled ad network to new emerging growth channels based on campaign-level signal rather than network averages.
- Daily campaign-level predictions replaced a model that was always 3 weeks behind. Jukka went from triangulating historical cohorts to acting on the same day with confidence scores gating each decision.
- Incrementality built into every campaign decision for the first time. With 50%+ organic installs, full blended ROAS had been crediting organic revenue to paid campaigns. Ktrl's incremental blended view removed that noise entirely.
- Exec and finance team pulled out of manual reporting. Ktrl is now the screen open in every weekly UA sync, with the finance lead and exec team accessing it directly instead of waiting on slide decks.
About Metamoki
Metamoki Inc. is a mobile gaming studio specialising in idle, incremental, and tycoon-style simulation games. The studio develops and publishes its titles in-house, with a catalogue that includes Idle Fishing Story, Wiz Khalifa's Weed Farm, and Merge Universe. Its flagship title, Weed Inc: Idle Tycoon (5M+ downloads, 4.7 stars), is a strategic idle game where players build a global weed empire by collecting and upgrading strains, hiring managers, and expanding across locations.