Meta Is Back in the Game. Here's What That Actually Means.
Guy Meiboom is Meta's EMEA Gaming Director. He spent years on the UA side before joining Facebook, which means he's sat on both sides of the table: burning ad budgets and managing the platform that takes them.
In this conversation, he doesn't pretend ATT wasn't a disaster for gaming developers. But he does make the case that Meta has genuinely changed, and more importantly, he lays out what the studios that are winning right now are doing differently.
Listen to the full conversation with Guy Meiboom on the Deconstructor of Fun podcast.
Here's what I took from our conversation.
The Honest Version of "We're Back"
Let's be direct about what happened after ATT. Meta's gaming organization was dismantled. The dedicated gaming team was dissolved and absorbed into country-level teams, where gaming clients competed for attention with e-commerce, fintech, and CPG. For a few years, the platform that helped birth mobile gaming as an industry was effectively treating it like any other vertical.
Guy doesn't spin this. He acknowledges the numbers told the story clearly: while other verticals at Meta recovered and grew after ATT, gaming didn't rebound at the same pace. In Meta's own earnings, the gap was visible. That's what finally prompted the structural change: a return to a dedicated global gaming org, based in London, working directly with product teams.
"I've never seen us working so closely with product. There isn't one day I don't have some sort of interaction with our product leaders around apps and gaming. This is very different from what it used to be."
The cynic in me, and I've been in this industry long enough to remember Facebook Canvas games, wants to push back. We've seen this movie before. Meta prioritizes gaming, then moves on.
But the structural evidence this time is different: a dedicated team with a direct line to product, with explicit mandate to bring feedback from developers into the platform roadmap. That feedback loop didn't exist before.
Is this "Meta is back" or "Meta is trying to get back"? Probably the latter. But the trying is real.
Creative is a Factory, not a Craft
If there's one thing Guy hammered consistently, it's this: the studios winning on Meta today treat creative production as manufacturing, not artistry.
This is harder to internalize than it sounds. Most studios built their creative teams around the assumption that great creative requires great creative talent, directors, animators, and conceptual thinkers who understand what makes an ad resonate. And that's not wrong. But it's incomplete.
"It's not about creating the right creative. It's about having the right mechanism in place. Volume. Velocity. Speed. Diversity. Really diverse ones, not just changing the background from light blue to blue. Really change them."
Guy uses a bicycle shop analogy that stuck with me: In a traditional store, you have one window. You put in the road bike during Tour de France season, the mountain bike when the trail crowd comes through. One window, one audience at a time. Meta's algorithm is thousands of windows simultaneously, but only if you give it enough creative variety to run them.
The data he referenced says top spenders are running around 29 creative variations per day. That number will only go up as AI lowers production costs. The studios that will struggle are those still treating each creative as a considered creative decision rather than a systematic experiment.
I pushed back on whether this creates any sustainable advantage, given that AI will democratize production capacity for everyone. And that the increased amount of creatives is a tax on studios while being another lever for networks.
Guy's response was honest: he's not looking for a moat here. He's optimizing for the best outcome over the next 18 months. The most important question in an AI-first world is genuinely unanswered. But the floor for minimum viable creative volume keeps rising, and studios below it will suffer regardless.
One important note from Guy: diversity and volume don't mean anything goes. Studios that use misleading ads or over-promise on gameplay might get short-term installs, but they're building on sand. His view, and I agree, is that if you're using guardrails for the right reasons, keep them. They protect your brand and your LTV.
The Attribution Tax: How Studios Are Bleeding Money Without Knowing It
This was the most practically valuable part of the conversation, and the one most studios will resist acting on.
Guy framed attribution like a math problem where you get the first line wrong. If you spend $500 a day and your attribution is off by 15%, that's manageable noise. If you're spending $500K a day with the same 15% error rate, that's a lot of talent and creative budget you're misallocating every single day, compounding over months and years.
The specific failure mode he called out: last-click attribution in a world of multi-touch discovery.
The scenario is straightforward. Someone sees your Instagram ad, doesn't click, keeps scrolling. Later, they're playing a game with an interstitial for your title. Muscle memory triggers an accidental tap. They visit the store page, don't install, and close it. Then two weeks later, they see your YouTube pre-roll, feel familiar with the game, and download it. Last-click gives credit to the in-game interstitial. The studio now thinks thatthe network is their best performer and pours budget into it. It's wrong, and it compounds.
"The credit goes to the platform I accidentally clicked on. And on the other side, there's a developer saying, 'that network gave me amazing performance’, and they invest more money there. The compounded impact over a long period for developers who invest heavily in performance marketing is just huge."
The solution he recommends isn't new, but it's underused: incrementality testing. Geo-lift experiments. Turning channels on and off in specific markets to measure true causality rather than correlation. Have a single source of truth for day-to-day execution, but audit it quarterly with a lift study. Don't question every assumption every day; you'll never move. But don't go years without testing whether your source of truth is still accurate.
Meta is significantly undervalued in multi-touch attribution compared to last-click models. Guy believes the same is true for most discovery platforms. If you're making budget allocation decisions based purely on MMP last-click data, you're likely under-investing in channels that are working.
The UA Team of the Future: From Knob-Turners to Control System Designers
The UA team for the next two to three years isn't optimizing campaigns. It's designing the system that campaigns run within.
The roles that matter:
UA Lead / Growth Director owns the portfolio, objectives, and network strategy. Defines what you're actually optimizing for.
Growth Analyst / Data Scientist builds yield models, tracks cohort maturation beyond the platform attribution window, rand uns incrementality.
Creative Strategist, generates hypotheses, owns the creative portfolio design, and segments audiences. Not the person making the ads, the person deciding which experiments to run and why.
Creative Ops / Producer manages the pipeline, tooling, versioning, and testing cadence. Runs the factory.
MarTech / Automation Specialist owns the APIs, rules, creative seeding, and governance. The person who makes the robots talk to each other.
Guy's lens on this: what all these roles share is critical thinking. AI takes over the repetitive execution. Humans own the judgment layer, defining objectives, setting constraints, and deciding when to intervene. The UA manager who was "good at Facebook" because they knew which buttons to press is being automated out. The one who can think structurally about what you're optimizing and why becomes more valuable.
His first two hypothetical hires for a new studio marketing team: creative, then data scientist. Not a performance manager. Not a media buyer. The function of UA is shifting toward system design and measurement before it shifts toward bid management.
The Market in 2026: More Revenue, Fewer Downloads, Higher Stakes
Guy is genuinely optimistic about the macro picture. Internet usage grows, demand for great experiences grows; therefore, room exists for great game developers. I share that view. But the details of how we get there are more complicated.
Mobile revenue is approaching all-time highs, while downloads have been in steady decline since late 2022. New game launches on iOS in 2025 are lower than the 2013 levels. These aren't contradictory data points; they're the same story. The market is concentrating. Winners are taking more share. The cost to enter and compete is rising faster than the tools are lowering the barrier to building.
Where Guy and I had a productive tension: he sees lower barriers to entry from AI as genuinely opening the market. Lean teams, creative ideas, and less capital required to build something great.
I think that's true for the building side, but the scaling side has gotten harder, not easier. You can make a great game with a small team now. Getting it to the audience that deserves it costs more than ever, requires better unit economics than ever, and competes against incumbents who own 70%+ of the market and can outbid almost anyone.
The Turkish and French studios he referenced as success stories aren't winning just because of creative ideas. They're winning because they understand the LTV-CPI equation cold, they move fast on iteration, they monetize hard through hybrid models, and they don't wait around for the market to find them. Speed and business sophistication together, not product quality alone.
Chinese developers continue to widen the gap. Guy's advice on learning from them, downloading their games, and playing them, is correct but insufficient. The competitive advantage Chinese studios have built is organizational and processual, not just product-level. Go to China. Work with them. Understand how they structure teams and decision-making. That’s my suggestion.
What Games Can Learn from Apps (That We're Currently Ignoring)
Five years ago, gaming was the teacher. Apps studied mobile games for engagement mechanics, retention design, and monetization frameworks. The ecosystem exported its knowledge in one direction.
That's over.
Apps now account for more than 50% of App Store in-app purchase revenue. They've hired the best marketing and data talent out of gaming studios. And in the process, they've developed capabilities that gaming still underestimates, specifically, onboarding and creative approach.
"Five years ago, I would have told you that all other industries should learn from gaming. Today, I think gaming also needs to learn from other industries. That's a big shift."
The onboarding gap is real. Gaming companies typically think of onboarding as an FTUE problem, the first session flow, the tutorial, the aha moment within the game itself. App companies treat onboarding as starting from the ad impression. Every step from initial discovery to habit formation is designed as a continuous arc. The ad is part of the onboarding. The store page is part of the onboarding. The first session is the continuation of a promise already made.
Gaming is very good at habit-building once it has a player. Apps are better at converting strangers into players. The gap is costing studios more than they realize.
The One Thing to Take Away
Guy's closing message was a challenge to challenge the status quo, in your org structure, your creative process, your measurement approach, and your budget allocation. He acknowledged this isn't gaming's traditional strength. We've been doing performance marketing longer than almost any other industry, which means we have more accumulated orthodoxy to unlearn.
"Audit for real. Change stuff. Be aggressive in changing stuff. If it doesn't work, you can always go back to what you did before."
Three things I'd translate from this conversation into immediate actions for a studio spending real money on UA:
1. Run a lift study. Pick one channel. Turn it off in one market for two weeks. See what moves. Do this before you make your next major budget reallocation decision.
2. Count your creative variations. If you're running fewer than 10 meaningfully different creative concepts right now, you are artificially constraining what the algorithm can learn about your audience. Not variants of the same concept, genuinely different angles.
3. Hire a data scientist before your next performance manager. The marginal value of someone who can model LTV curves and design incrementality tests has never been higher. The marginal value of someone who manages Meta campaigns has never been lower.
Meta being back is real, but conditional. They're back if you're willing to meet them with the infrastructure, creative, measurement, and team design that makes their tools work. The studios treating this as "Meta is running better ads for us again" will be disappointed. The ones who treat it as "the attribution environment is changing, and we need to adapt our measurement stack" will benefit.
