Written by Josh Chandley, COO at WildCard, architecting long-term direction and overseeing product, production, and marketing. He has 14 years of experience building, operating, and scaling game companies.

The graveyard of mobile ad tech is full of once-great names: Chartboost. Vungle. AdColony. Millennial Media. InMobi. TapJoy. MoPub. Fyber.

Even now, history is repeating itself: Unity. IronSource. Digital Turbine. Liftoff. InMobi (still hanging on). All are left playing catch-up.

But one network didn’t just outlast them. It evolved. AppLovin became the apex predator of mobile ad tech, not because it outworked the competition, but by changing the rules of the game.

And that’s the real story here: AppLovin didn’t just win. They rewrote the platform playbook.

The Shift to System Control

As 2018 kicked off, the mobile ad tech landscape was still very fragmented. IronSource and MoPub ruled mediation, the auction that decides which network wins the ad impression. Unity was growing in rewarded video. Vungle was faltering. Meta and Google still soaked up most UA (user acquisition) spend.

Most ad networks wanted to win the auction. To be the highest bidder and get to show their ad.

AppLovin wanted to control the board.

AppLovin made two moves that changed everything. First, they acquired MAX to control the mediation layer. Second, they leaned into publishing—but not how most people think.

The initial steps into publishing were made quietly, backing Matchington Mansion behind the scenes. They didn’t just fund Matchington Mansion. They orchestrated it through a stealth launch under a shell studio, Firecraft, registered to a San Mateo coffee shop. The playbook? Clone Playrix. Flood the market before anyone notices.

Matchington Mansion was a fast follow-up of Gardenscapes. It was built quickly, marketed aggressively, and scaled globally, powered by AppLovin’s own ad network. AppLovin wasn’t just supporting publishers. They were becoming one.

This wasn’t about revenue. It was about data. Publishing gave AppLovin raw insight into what creatives worked, what users retained, and what LTV profiles scaled. All first-party. All proprietary.

Then they industrialized it.

Lion Studios wasn’t just a label – It was a hypercasual factory. Fast launches, high volume of ad revenue data, and tight loops. Every game trained its models. Every install fed their system.

These moves weren’t about revenue. They were about control of the loop: show the ad, track the result, and bid smarter next time.

Publishing gave them an edge, but to fully lock in that advantage, AppLovin needed to own the infrastructure that powered the loop. This started with MAX, their newly acquired mediation platform built for real-time bidding.

Owning the Infrastructure

As MAX started to scale in 2019, it became clear they had a winner on their hands.

While other mediators duct-taped bidding into legacy waterfall platforms, MAX was built from the ground up to run auctions with real-time bidding.

***Waterfall auctions*** *asked one network at a time for a set price—often requiring 30–40 calls per impression.* ***Real-time bidding*** *changed everything: one change, all networks compete, best bid wins.*

Waterfall platforms used a fixed priority order—if the first network passed, the second got a chance, and so on.

Real-time bidding runs a live auction where all networks bid at once, and the highest bidder wins. Low latency. Easy testing. Better yield.

AppLovin marketed MAX as “cleaner pipes”, and developers loved it.

*Over the last 3 years, Real Time Bidding adoption has soared to 90% of revenue*

But the real advantage wasn’t having a mediation business. It was owning the data infrastructure. With the infrastructure in place, AppLovin made its play to consolidate the industry around MAX. In 2021, AppLovin acquired MoPub, the largest independent mediation platform, and shut it down.

At its peak, MoPub mediated over 50,000 apps. AppLovin didn’t want the tech. Or the integrations. It was less of an acquisition and more of a deletion. After MoPub’s shutdown, MAX didn’t just grow, it became two-thirds of the entire market.

By sitting between supply and demand, MAX became AppLovin’s single source of truth. The definitive system for what was shown, who saw it, and what happened afterwards.

Every impression fed their targeting models. All first-party. All in real time.

Meanwhile, Unity and others were still reliant on third-party mobile measurement partners (MMPs) to track ad performance. Each publisher might use a different MMP, so Unity received fragmented signals from a dozen partners, each with its own delays, metadata formats, and error rates. The result? Data that was delayed, inconsistent, and often incomplete.

In a system optimized by machine learning, bad data isn’t just messy. It’s dangerous.

And in 2022, that danger became real.

In Q2 2022, Unity blamed a major revenue shortfall on “a large advertiser’s bad data” corrupting their ML models. We still don’t know exactly what happened, but it exposed the fragility of depending on third-party signals. Shares tanked almost 40%.

AppLovin avoided that risk by design: MAX gives them real-time, first-party IAA (in-app advertising) data without relying on third parties that could accidentally send bad data. MAX wasn’t just a better mediation. It was AppLovin’s walled garden of infrastructure. And the beginning of everyone else’s dependency. And more than that, it was a blueprint. A case study in how vertical integration creates strategic lock-in.

In a host network model like AppLovin’s, the mediator sees both bids and revenue. Clean, first-party data that fuels better targeting. Bidding networks rely on third-party MMPs and fragmented client signals, introducing latency, gaps, and risk

Supply Lock-In

Here’s the part most teams didn’t fully grasp until it was too late: The host network mediating inventory is the most efficient way to buy that inventory. Why? Let’s take a look at MAX.

First, AppLovin charges Unity and other ad networks a 5% fee for the right to bid on inventory they mediate, meaning Unity’s $20 becomes $19. AppLovin’s ad network does not pay that fee.

Second, AppLovin’s ad network has real-time access to impression-level data. That means faster feedback loops and better model performance than any external network can achieve. It also grants them immunity from the data incident Unity endured in 2022.

Finally, AppLovin has the ability to influence the auction itself. As the auctioneer, they control the structure. Whether or not that influence affects outcomes is unclear, but the performance gap is clear.

What are those results? AppLovin's share of revenue on MAX is typically 2-8x higher than its share of revenue on other mediators like LevelPlay.

*The mediating network wins by design. No fees. Better data. It’s the auctioneer of an auction it participates in. Competing networks play the game, but AppLovin sets the rules.*

The lesson is simple: if you want to buy ad impressions mediated by MAX, you need to run your campaigns on AppLovin’s network. AppLovin understood this, and they moved decisively. They announced that publishers could only run adROAS campaigns—the primary way ad monetized apps scale—if they mediated with MAX.

MAX wasn’t just a mediator anymore. Because the vast majority of mobile game impressions were mediated through MAX, it created a network-level lock-in, where scale came not from yield, but from control of the pipes. So if you want to scale on MAX inventory, you need AppLovin UA. If you want to run AppLovin UA, you have to mediate through MAX.

Most developers didn’t realize: the moment MAX mediated your game, AppLovin was influencing your growth strategy. Mediating with AppLovin makes them your primary UA source.

The mediation–UA link is brutal. Leaving MAX means giving up adROAS campaigns, your biggest growth driver. Even if another platform pays 5% more per ad, losing adROAS could cut your installs by 30–40%. The math doesn’t work.

That influence didn’t stop at IAA and hybrid.

By gating adROAS campaigns, they could leverage that scale to gain pricing power in IAP-driven UA, too. Want to buy MAX inventory without paying a 5% bidding tax? You’ll need to run AppLovin UA. Even if you’re chasing IAP whales, you’re still operating inside a system built for IAA.

There was no independent option. AppLovin had taken MoPub off the board. Their market share was at least 60% at that time.

*AppLovin’s feedback loop turns market share into system control. Every input feeds their advantage*

Mediation is About UA

In the new world, mediation wasn’t about optimizing yield. It was about staying eligible for installs. That’s the quiet shift most teams missed. They thought they were picking a mediation partner. In reality, they were picking who influenced their user acquisition strategy.

And on MAX? That influence isn’t subtle. One policy tweak, one targeting shift, and scale disappears. Especially as AppLovin is now a public company, policy tweaks on creative are becoming more and more common. For example, AppLovin in 2025 tightened video ad rules around what could be considered misleading or deceptive content. Right or wrong, many publishers saw double-digit install drops.

The impact? When AppLovin shifts the rules, every team on MAX adjusts. AppLovin is structurally critical for UA. Unity’s LevelPlay and Google’s AdMob mediation are still around, but mediation is about UA, and nobody can match AppLovin’s access to inventory.

AppLovin Eats First

Meta owns Instagram, ByteDance owns TikTok, and Google owns YouTube. AppLovin owns ad-monetized mobile gaming—not because they scaled faster, but because they closed the loop while everyone else missed it.

Today, MAX mediates the majority of mobile game impressions. AdROAS campaigns only run if you use it and every install feeds their system. Most networks build tools, while AppLovin built the infrastructure. If you’re running an ad-monetized game, here’s the uncomfortable truth: You don’t just use AppLovin.You operate inside it.

The next generation of winners won’t succeed by just optimizing their ad network; they need to own their infrastructure and the signals that drive it.