Mobile Ad Eaters on AI’s effect on profitability, performance, and production in mobile gaming.
“The current generation of AI tools still has notable technical limitations, especially when the goal is to match the level of polish, character nuance, and brand consistency our IP requires.” – Luka Komatar
“Most people still think of AI as its own thing: art is either AI-made or something else… it should be instead considered as a tool or even a plugin.” – Gašper Jager
“You’re only as good as the speed of your creative production. That’s the real story behind AI in performance marketing.” – Matt Veysberg
AI doesn’t change the goal in user acquisition: profitable scale. What it might change is speed- how fast teams can explore, iterate, and learn. Here’s the real question: can that velocity turn into measurable performance gains in your campaigns?
For the first edition of Mobile Ad Eaters Bites, we asked three industry experts for their unfiltered take on AI creatives and ROAS – what’s working today, what isn’t, and what’s actually driving results. Verbatim responses below.
AI Creatives vs Gameplay Ads: ROAS Lessons from Outfit7
We recently explored an AI-driven creative direction for the Talking Tom Remaster campaign. While the experiment was exciting, it did not result in an immediate lift in ROAS.
The current generation of AI tools still has notable technical limitations, especially when the goal is to match the level of polish, character nuance, and brand consistency our IP requires. Certain creative concepts are simply not feasible right now without sharing more brand assets than we are comfortable with. As a result, our gameplay-focused creatives continue to outperform the AI variations in pure campaign performance.
That said, the test was not a setback. It gave us a clearer understanding of where AI can meaningfully support our workflow and where it still comes up short. The speed of iteration, the ease of prototyping visual directions, and the ability to generate multiple creative angles quickly are all extremely promising. As these tools advance, particularly in the areas of controllability, consistency, and brand safety, I can see AI becoming a much larger driver of performance.
We may not be seeing ROAS gains from AI yet, but the trajectory is clear. With better technology and more refined workflows, AI-assisted creatives have strong potential to contribute in a measurable, scalable way moving forward.
Why AI Alone Does Not Improve ROAS at Ten Square Games
By itself, AI alone does not improve ROAS. Let me explain. Most people today still think of AI as its own thing, its own production style, a wholesome AI-generated video or image – as if this is how it goes. Art is either AI-made or something else. Here is the problem: it should be considered as a tool or even a plugin.
We use AI in our regular pipelines in video production and static design. We use Nano Banana the same as we use object selection in Photoshop, or we use video GenAI tools like VEO 3.1 to animate new intros in the style of our 3D animation, thus speeding up production.
We use Sora2 to generate UGC content, live commercials, and assets that we would never have had the budget or time to produce, because UA is a numbers game above all, and you cannot sacrifice so much time and resources on a single creative vision.
But with AI, you can. You can generate a live ad in a day, which would classically take weeks; you are given options to explore and test things that were previously beyond your reach.
So, if you only consider what you can create, with the arrival of AI, the ROAS is already there. And while wholesome AI made video ads in my case have not yet outperformed classic 3D animated videos, that does not mean that a combination of AI and classic production didn’t. In fact, it did in some cases, and in other cases, using AI allowed my team to execute some things faster, to have more time to do others in a classical way, which is why those videos probably performed better, because there was more time, and people working on them were more creative.
So do AI creatives improve ROAS? Perhaps, but for sure, AI workflows and AI tools do.
Creative Production Speed as a Performance Driver at Plarium
You’re only as good as the speed of your creative production. That’s the real story behind AI in performance marketing. Generative tools don’t improve ROAS because they look cool. They improve ROAS by letting you test 10 ideas in the time it used to take to produce one. The gains come from velocity, not novelty. When you remove production bottlenecks, you uncover winners faster, fight creative fatigue earlier, and match the right message to the right audience with far more precision.
But AI is not a replacement for strategy. If the underlying concept is weak, AI simply produces more variations of the same weakness. The measurable uplift we’ve seen comes from faster learning cycles and lower cost per asset, which ultimately translates into more consistent improvements across channels.
Join Mobile Ad Eaters Retreat 2026
If you want to keep this conversation going in person, Mobile Ad Eaters Retreat (April 23-24) is back in Prague – a creative-focused event for CMOs, UA & Creative Leads, Creative Producers, and Creative Strategists in mobile gaming.
In 2025, we brought together 100 of mobile gaming’s top creative minds for a new kind of event built on connection. This year we’re scaling to150 top industry voices, with speakers already confirmed from Zynga, Playtika, Outfit7, Nazara, Nitro Games, and more.
Want to hear more?
Talk to 150 creative minds in person
