The days of UA as a media buyer are pretty much over, and have been for quite some time. The mobile game industry is a crazy one to be a part of. Pick any random year from the past 15 years, and you’ll see a different environment, different winners, and different skills to win.
When it comes to UA, the role has evolved so much it almost doesn’t resemble what it was a few years back. Folks I have been speaking with recently, all of whom lead UA in some of the biggest game studios in the world, all claim it's harder than ever to keep up.
But it now feels like the change happening is one of the biggest there has ever been. It’s not simply a change in which channel works best, this measurement framework or another one, nor is it about a new campaign type.
Nope. This change feels way bigger. Let me unpack that.
1. The underlying technology that ad tech is built on is quickly shifting
As we covered in this article, it is an ML world, and the vast majority of networks are moving quickly into a future of high-performing, ML-driven UA. In practice, it mostly means that these networks are either already, or are going to be, extremely effective in identifying audiences with high predicted LTV for your games based mostly on signals from your creatives.
With almost no targeting whatsoever, these networks are either already or soon to be machines you supply with:
- A very large number of high-quality and diverse creatives, including significant numbers of playables.
- A target economic goal, usually a ROAS target.
- And that’s pretty much it.
This changes the shape of the systems you have to work with to do your job well.
2. AI is in the early innings of impacting most knowledge work
At the same time, AI, specifically agentic AI, is coming quickly and disrupting a huge chunk of most knowledge work.
It is growing from systems that are basically cute chat apps that everyone is used to by now, to sophisticated agentic systems that can reach real and valuable business goals.
Of course, some of this change is invisible to most of us - the evolution of AI tech and science in general is more than simply LLMs. The very fact that Applovin has become a $200B business with unprecedented profitability metrics is tied to this evolution, which they cleverly capitalized on early with their unique data set and their own machine learning and RL infrastructure.
But it’s not hard to imagine that the next few years will introduce agentic AI systems that will be incorporated into the day-to-day of most jobs.
We’re at Sett working on agentic AI that will help be your daily driver in how you find new creative winners, from ideation, performance and behavioral data analysis, high-quality generation of playables, deployment, and testing.
This near future will likely be characterized by UA teams managing powerful and capable autonomous systems to achieve their growth goals. Managing these systems is very different from the old human-heavy structure of often siloed teams: UA, Creatives, Analytics, Marketing, and more.
Getting a glimpse of what winners do right now
When thinking about what winners are doing, it’s easier to sift through the world and ask: who is winning right now, and how are they doing it?
There are a few clear winners in the mobile game industry, geographically speaking. But the clearest trend is the gradual increase in domination by game developers and publishers from China - now accounting for 50–60% of the world’s top-grossing games globally.
Learning from these winning teams, we can get a glimpse into the future of what skills are required to make yourself invaluable to your organization in the next few years. We can see this simply by analyzing how these teams are likely to operate given their output, which you can see on platforms like Sensor Tower and App Magic.
Four UA skills to master sooner rather than later
- System Thinking
First and foremost, we see teams that treat user acquisition as a complex and deep system - almost from an engineering mindset.
These winning teams are not thinking about success with creatives and UA as luck, or a moment of genius an artist had when coming up with a new creative concept.
They engineer a system that is designed to result in UA scale and profitability.
This involves data-driven ideation with taste, leveraging the truth of how networks are built these days and getting the most out of them, development of a high-throughput pipeline of creatives, systematic exploration of how to market their games, and intelligent, data-driven decision-making when it comes to testing and iterating.
When viewed as a system, humans with an engineering mindset are simply working to identify the system's weakest points, improving processes, and evaluating their changes based only on performance data.
They are not trying to get a random win; they are trying to build a consistent winning machine.
- Game Design Taste
Within this system, it’s clear that one skill is becoming super important: game and level design taste. When comparing, for example, the playables that winning teams are putting out there and the bottom of the pack, it’s pretty evident that the winners are thinking much deeper.
They think about the narrative of each playable and how fun it is. They think about the effectiveness of tutorials, the core loop, how other loops are introduced on top of it, juice, progression, the emotional journey, and how to create the desire to install.
This is a very different skill set than thinking only about copywriting, value propositions, and themes. It’s an understanding of what makes a game fun, a skill that traditionally sat with other teams within a game studio.
Teams that simply make low-quality copies of a trending ad concept miss out on this, thinking a concept didn’t work for them, while the truth was that the copy was not able to carry the fun factor of that original playable ad. You see these a lot in the industry right now with so-called playables that look like a trending ad concept but miss all the fun the original had.
By doing this, they ensure the playables have the best conditions to win. With higher IPMs, they allow the networks many more degrees of freedom to bid on much higher CPMs and be way more competitive in auctions for high LTV users.
Here's the level of thinking we believe is needed to be in a great system - deconstructing the trending game "Level Devil".
- Holistic Strategy
Additionally, winners of the future seem to be significantly more involved in how their creative ideation is related to the game, in a very deep way.
This is not simply about thinking about misleading ads as some do for quick wins (which win on some parts of the value equation but lose on others). Instead, it's about deeply slicing and dicing what their game is about from a fun and play perspective - developing concepts that show different types of play and exploring (not assuming) which types of play will have a high affinity for their games.
This is more about understanding a joyful moment in a game, which could be about evolving characters, and exploring whether a playable with a core game loop of tower defense with heroes will attract high LTV players that will convert in the full game.
- Rapid & Diverse, High-Velocity Experimentation
Last but not least, they are definitely not making any assumptions and have made peace with the fact that in modern UA networks you often cannot know why a creative worked.
To do this, they optimize for as much speed and diversification in playable creatives as possible while maintaining a high-quality standard (referring to the taste in game design). They let the data guide them forward and simply ensure they maximize the learnings (and minimize the time-to-learnings) the networks can have, as well as their team.
It’s an optimistic future for UA
We believe that UA folks have a great future ahead of them. Not as simply media buyers and campaign configuration architects, but as UA strategists that manage significantly more powerful resources - some of them human, some of them AI - and orchestrate and improve entire systems and operations that will make or break a game.
About the Author
Fishi is the Head of Marketing at Sett. His brain is a chaotic jukebox of ideas with more cultural references than any feed can handle. He collects sneakers and plays chess while you’re still counting sheep.
