Scaling user acquisition for a mobile game in today’s privacy-first landscape can feel like trying to solve a puzzle with half the pieces missing.
You find a great content creator, the ad looks fantastic, and you launch it. But then you run into a roadblock: fragmented attribution and weak data signals. Without precise tracking, how do you know which creative styles are actually driving high-value in-app purchases and ad revenue, and which ones are just burning your budget?
This was the exact challenge faced by Nanobit, the globally recognized mobile game developer behind massive narrative hits like Unfolded: Webtoon Stories.
For mobile game teams, success increasingly comes down to two things: giving platforms the right data to optimize towards valuable users, and building a creative pipeline that can keep up with constant testing. An install is only the beginning. The real value comes from players who stay engaged, make purchases, and generate long-term revenue. That means ad platforms need more than just install data to understand what makes a valuable player.
To tackle this, Nanobit deployed Audiencelab by Geeklab for campaign optimization and to dominate their TikTok performance. And the results speak for themselves.
The challenge: Moving past vanity metrics
Nanobit’s primary goals for the new campaign were straightforward but demanding: drive high-volume installs, scale In-App Purchases, and boost ad revenue.
Historically, many studios rely heavily on front-end delivery metrics to judge whether a video is a “winner.” But as we’ve explored before, high click-through rates don’t automatically mean high-value players. If your downstream post-back signals are delayed or muddy, the ad network’s optimization algorithm is essentially flying blind.
To fix this, the strategy had to pull two distinct levers simultaneously: rapid creative diversification and advanced signal engineering.
The solution: Creator diversity meets signal engineering
The collaborative blueprint between Nanobit, TikTok and Audiencelab relied on a continuous loop of testing, measuring, and auto-optimizing.
- Native TikTok content
Instead of betting the house on one or two polished marketing videos, Nanobit utilized TikTok One creators to build an engine of authentic, native-feeling content. By working with real creators, they were able to rapidly deploy completely different visual hooks, narrative angles, and gameplay styles. This variety kept ad fatigue low and gave the campaign a massive pool of creative assets to test.
- Tightening the feedback loop
This is where the magic happened. Authentic content only takes you halfway; the data infrastructure has to carry you across the finish line. The campaign tightly integrated Audiencelab with TikTok Events API right alongside TikTok’s Purchase Optimization campaign type. By feeding high-quality, real-time purchase signals back to the platform through Audiencelab’s dedicated infrastructure, the campaign didn’t just track who installed the game, it taught the algorithm exactly which creatives drove high-value, paying players.
Over time, TikTok’s delivery engine actively sought out users with the highest intent to purchase, matching them with the creative variations they were most likely to engage with.
The takeaway
Great creative hooks users, but robust signal engineering is what teaches the ad network who to serve those creatives to. When you combine authentic creator content with precise, real-time event tracking, your ROAS follows suit.
Find the original version of the case study published by TikTok here.
Want to stop guessing which creatives are moving the needle and start engineering your signals for actual profit? Let’s talk!
