The mediation business was launched as a solution for studios. Too many networks managed directly (especially in an ecosystem dominated by waterfall) was a hassle, so mediation appeared as a fairly seamless way to manage these integrations and how they interact with advertising inventory.
Fast forward to 2025, however, and that seemingly innocent step in favor of app developers has evolved into the heart of every monetization strategy and built an industry where the largest player (Applovin) commands a $200 billion-dollar market cap in part due to its hold on advertising inventory and data.
To a large extent, this growth was at the expense of the studios it was built to support. Despite owning the supply, studios agreed to have mediation be provided as a free service. This led to innovation being focused on the demand side. Networks, user acquisition models and so on, evolved, while studios were increasingly bound to their mediation provider in order to access their growth offering. Fee transparency disappeared as margins grew. The mediation players even moved into the gaming space and directly started competing against their clients.
To make matters worse, a shift to bidding effectively shut down entrants to the mediation space. Launching a mediation is impossible if you can’t access the main sources of demand, and all the major players restrict access to their network in the new environment. The infamous walled-garden flourished.
Along the way, the industry witnessed what I now call mediation wars, Round One. Integration bonuses in the 7 figures, aggressive marketing strategies and outlandish events framed the mediation land-grab being fought out between a somewhat slow Google and the appearances of ironSource and Applovin. Studios willingly gave up their data and the core of their business in exchange for these short-term benefits.
This culminated with a fight to see who would get to merge with Unity and create the expected winner. Ironically, ironSource merged, but Applovin won.
We’re now at the cusp of Round Two. Applovin has outgrown even the most optimistic analysts and is now poised for bigger battles. It has its sights set on Google and Meta, and it tried to buy TikTok in the US. Continuing to grow in the traditional gaming ecosystem is hard when you already dominate, so they turned towards CTV first, ecommerce second and maybe web next? If nothing else, they seem to be distancing themselves from mediation (the term MAX is now hard to find on their website) and focusing on Axon, their growth engine.
While no longer involved with games of their own, Applovin continues to exercise firm control on the industry. Policy shifts are not uncommon (even unwritten ones) and sanctions are swift when studios challenge their rules. The quality of Axon, their UA engine, has become so good at driving traffic (especially optimized to generate a return in an ad-monetized environment), that leaving their ecosystem is hardly an option for many studios. Not only will your ability to acquire users fall, but your CPMs likely will, too. I confess to admiring their ability to create this reinforcing loop.
In a market where many studios are struggling to make a profit however, this is generating frustration.
It stands to reason that mediation should be a commodity; the pipes through which supply is connected to demand. If this were so, however, an app should see similar performance and demand distribution regardless of the provider it uses. Anyone in the space knows that this is not the case, with the mediation-owned network typically drastically outperforming (i.e. larger SoW) when used within its mediation.
This differentiation among mediation providers, however, also means that it’s unlikely that a single mediation will outperform across every single user, session, placement and format for a single app. Each network has its own strengths. If mediations are all different, then studios are interested in finding the best mediation strategy to maximize their returns.
Unfortunately, this is very hard to do today. Using more than one mediation is technically possible, but puts access to user acquisition at risk. Unless you’re willing to risk losing access to your UA campaigns, using a segmented mediation approach is not a possibility.
Accessing more than one mediation through prebid/postbid setups is also possible, but after much use has been regulated out by mediation platforms through changes to their terms and conditions.
Monetization today is a game of cat and mouse. Desperate studios come up with creative ways to extract value from their inventory, while mediation companies try to figure out what studios are up to in order to change their policies and disallow this behaviour if it in any way affects their margins.
The first, most relevant example of this was the shift to bidding. Pitched as a gift to developers because of its simplicity, this wolf in sheep’s clothing was also a way to curtail the number of optimization windows that developers could use in order to find the best price for their inventory. Making a bunch of calls of a single impression was good to studios, but it was costing the demand side (who processed each such request but rarely got filled). A unified auction did wonders for their margin.
More recent examples include the ability to run pre-bid or post-bid setups (all the craze two years ago, now forbidden for most), or the shifting restrictions towards triggering multiple requests for a single ad impression.
While common for large players to get preferential treatment, the mediation space has taken this to a new level. As rules and policies change (see above), the precise application is always subject to negotiation. As mentioned above, postbid is forbidden for most (but not all).
Fill rates for ad-units weren’t monitored in the past; but now they are and the threshold imposed on studios varies widely. The lower the fill rate, the harder it is for the studio to use this feature to improve CPMs. We’ve seen everything from 90% to 0% min fill rates imposed.
As studios get increasingly restless with this setup and networks other than those belonging to mediations secretly work to find alternative paths to supply, the status quo is once again being challenged.
Google, who had so far been silent as they stripped its market position, now seems to be back. For reasons unknown, they decided to invest in the space and are once again showing up with new features, integration bonuses and more.
Unity also seems to have recovered from the runtime-fee hangover and refocused on supply. With a frustrated market asking for change, they seem to be investing in their mediation in addition to improving how their own growth engine (Vector) works. If you want to fight Applovin on mediation, you need to have an alternative to Axon.
There’s talk of mediation accepting postbid setups again. Unlimited ad-unit creation, improved ad-quality controls and maybe even the option to introduce bid-adjustment strategies to the auction (allowing studios to decide when the “highest” bid should actually win, rather than favoring those with higher revenue per second, or ad-quality metrics).
It’s not for me to say how this will play out. Like you, probably, I’m a spectator to these gladiators fighting it out, and often caught in their wake. All I can say is that I sense movement in what has recently been a somewhat stale space. The top 3 seem increasingly restless and there’s always rumors of new players stepping in.
Whatever comes from this, I hope studios can behave in a more strategic way this time. Long-term pain for short term candy didn’t play out for them last time; they should be careful to profit from this new scuffle, if it happens.