User acquisition (UA) is one of the biggest challenges in gaming today. Whether indie developers or major studios, teams are forced to spend tens or even hundreds of thousands of dollars to test ad creatives -- often waiting weeks to identify what works with audiences. This process, reliant on traditional ad platforms and real-world A/B testing, has become a costly bottleneck, said Jonathan Zweig, CEO of AppOnBoard, in an interview with GamesBeat.
AppOnboard, the company behind do-it-yourself game development platform Buildbox, is solving this with Quvy.com, a breakthrough tool that uses simulated audiences to predict ad performance -- delivering insights in minutes at a fraction of the cost -- allowing gaming companies to grow their real audiences at lightning speed.
The team addressed the major pain point for indie developers: high user acquisition (UA) costs. They developed a solution using AI to simulate audiences, reducing UA costs from $400,000 to $4,000, Zweig said. Here's a case study.
"The AI, developed over six months, predicts ad performance accurately, saving time and money," Zweig said. "The discussion also touches on the broader implications of synthetic data in AI, the challenges of eSports sustainability, and the impact of major game releases like Grand Theft Auto VI."
Why Quvy matters now
UA costs continue to climb. Acquiring a single paying user in mobile gaming can exceed $50. Creative testing now favors studios with massive budgets, leaving smaller developers behind.
Quvy levels the playing field. It replaces expensive, slow A/B testing with fast, predictive simulations. Developers can now test thousands of creatives in minutes, launching only the top performers and saving both time and money.
"One of the biggest responses we got when we asked the community was about the biggest pain points for an indie developer, and overwhelmingly it was the cost of UA," Zweig said. "How do you compete with the big guys? Our CTO thought of this crazy idea, 'Why don't we simulate audiences to test ads to dramatically reduce UA costs?'"
The company tried it and it's "crazy effective," Zweig said.
Why It's a Game-Changer for Major Publishers
Top gaming companies spend large sums of money testing ads before launch. With Quvy, they no longer have to. Simulated audiences allow for:
A billion-dollar opportunity
The global digital ad market is projected to surpass $730 billion in 2025, with AI-powered marketing growing at 26.7% CAGR. Traditional A/B testing tools make up over $50 billion of that market. Quvy's focus -- simulated audiences -- is an emerging $20 billion to $30 billion category, growing at 30%+ annually. It's reshaping how developers validate creative ideas pre-launch.
Supporting the developer community
"Discoverability shouldn't be limited to those with the biggest budgets," said Zweig. "We're proud to offer Quvy at affordable pricing so anyone can market their games and apps as effectively as the largest studios and companies in the world."
Whether you're launching your first game or managing a global UA budget, Quvy offers a smarter, faster, and more affordable way to win.
"By lowering the barrier to creative testing, we're helping developers compete based on quality and innovation -- not just ad spend," Zweig said.
AppOnboard builds AI-first tools that empower creators. Its flagship product, Buildbox, enables anyone to create games and apps without coding. Quvy, its latest innovation, uses simulated audiences to predict creative performance before launch. Together, they form a unified ecosystem for building, testing, and scaling creative ideas.
How it works
It usually takes an ad network about seven days to learn what works in terms of ads for gamers. With Quvy, it takes about three minutes, Zweig said.
The company does this by testing the ads for acquiring users on synthetic people. It's a lot like how Jensen Huang, CEO of Nvidia, talked at CES about testing robots on "synthetic data." A company would be lucky to test a self-driving car for a million miles, but it can't figure out all of the emergency situations that could arise when driving a car. But with synthetic data, the company can test the car for billions of miles in a virtual environment that simulates the real world.
The same principle applies with synthetic data for user acquisition, Zweig said.
"They're synthetic people, but they have real emotions, real tastes, real judgment, real preferences," he said. "It's amazing. When we see an ad, what's going on in our brain is a lot different. And you can distill it into actual mathematics, which is then trainable for large data sets."
The company has run a bunch of ads on places like QVC and on Facebook to see if it can nail the best ads for those platforms. And it gets to the answer in a fraction of the time and cost of other methods.
"It's incredible what you can model with the right data," he said.
It's taken about six months to build Quvy, with a relatively small part of the AppOnBoard team. Meanwhile, Buildbox is still profitable, Zweig said.
