Mediation or a Single Network: Which Pays More?
How auctions, fill and setup effort differ between one ad network and a mediation stack, and which one suits your app.
6 min read · Updated
Sooner or later every mobile developer faces the same fork. You can add one ad network's SDK, such as AdMob, and be showing ads by the afternoon. Or you can put a mediation layer in front and let several networks compete for each impression.
Mediation is not free money. It adds an SDK, a second dashboard, more testing and a few fresh ways to break a release. It also tends to raise both fill and eCPM, and the gain is usually largest for apps with a global audience. Whether it pays for itself depends on your size, your regions and how much time you can spend.
Below we cover how each setup works, what changes in the auction, and who should pick which. The figures come from the same benchmark tables as the revenue calculators, so read them as estimates, not promises.
One network means one door to demand
A single-network setup means your app asks one company for ads. With AdMob, that company is Google, and behind it sits a large pool of advertisers bidding in Google's own auction. So "single network" does not mean a single advertiser. It means one door to demand, and the owner of the door decides which advertisers get through and what you are paid.
The upside is simplicity: one dashboard, one payment, one policy document to read. The downside shows up at the edges. If the network has no suitable ad for a player in Brazil at that moment, the request goes unfilled and the slot earns nothing.
Mediation puts a switchboard in front of several networks
A mediation platform is an SDK that sits between your app and the networks. The app asks mediation for an ad. Mediation asks AdMob, Unity Ads, Meta Audience Network and whoever else you have switched on, then shows the best result. AppLovin MAX, LevelPlay and AdMob's own mediation all work this way, with different strengths.
Each network brings its own advertisers, so the total pool of buyers grows. That matters most in Tier 2 and Tier 3 regions, where any single network has thinner demand. How payouts reach you varies by platform (some pay on behalf of the networks, others leave you with several payment relationships), so check the current terms before you commit.
Waterfall and bidding are two ways to run the auction
The older method is the waterfall. Networks are ranked by their expected eCPM, and the app asks them one at a time from the top. If the first passes or returns an ad below your floor, the request drops to the second, and so on. The weakness is that rankings are guesses built from past averages. A network willing to pay $20 for this particular player may sit at rank four and never be asked.
Bidding removes the guessing. Every participating network sees the request at the same moment and submits a live price, and the highest bid wins. You get fairer pricing and less manual tuning. Most mediation stacks now lean on bidding and keep a waterfall as backup for networks that do not bid, though the exact mix changes over time.
A single network runs only one auction, among its own advertisers. Mediation adds a second layer: each network's best offer competes against the others. The glossary defines waterfall, bidding and floor price.
Where the extra fill and eCPM come from
More buyers means fewer empty requests (fill) and a higher winning price (eCPM). A gain in the low double digits is a fair planning assumption for a Tier 1 heavy game, and some apps see less. Here is the arithmetic with assumed numbers, based on a game that sends 40,000 rewarded ad requests a day:
- Single network: 88% fill gives 35,200 impressions. At a $12.00 eCPM, that is 35.2 x 12 = $422.40 a day.
- Mediation: 92% fill gives 36,800 impressions. With a 15% eCPM lift to $13.80, that is 36.8 x 13.8 = $507.84 a day.
The difference is $85.44 a day, roughly $2,560 over 30 days, or about 20% more revenue. The $12.00 figure is the Tier 1 rewarded benchmark in our tables, and the 4-point fill gain and 15% lift are assumptions, not guarantees. Run the same math at a tenth of the volume and the gain shrinks to about $256 a month. That may not justify weeks of integration work. For how the underlying numbers are measured, see eCPM explained.
The price you pay in setup effort and SDK weight
Every extra network adds an SDK and an adapter to your build. Together they grow the app size, often by megabytes rather than kilobytes, so measure your own build. They also add start-up work and more version pairs to keep in sync. An outdated adapter is a classic cause of a bad release.
Privacy work scales with the network count too. Your consent flow has to cover every participating network, and on iOS the App Tracking Transparency story affects each one differently. You will also publish an app-ads.txt file listing each seller. Expect a single-network integration to take hours and a well-tested mediation stack to take days or weeks. Google's AdMob developer documentation and the Unity documentation both list the adapter and consent steps.
Side by side
The table below compares the two setups on the points that usually decide it.
| Factor | Single network | Mediation |
|---|---|---|
| Setup time | Hours | Days to weeks |
| Fill outside Tier 1 | Often patchy | Usually better |
| eCPM | One auction | Competing auctions, often higher |
| SDK weight | One SDK | Platform plus adapters |
| Reporting | One dashboard | Several, plus a combined view |
| Risk of breakage | Low | Moderate |
When one network is plenty
Stay with a single network if you are still prototyping, if daily active users are in the hundreds, or if your audience sits mostly in one strong region. The same goes for a utility app with one banner, where the dollars are small and every megabyte counts. A solo developer shipping a new title every month will usually get more from a better game than from a better ad stack.
AdMob alone is a respectable default: in our tables its Tier 1 rewarded eCPM sits at $12.00 and interstitial at $8.00, with blended fill near 88%. Use the AdMob revenue calculator to see what that means for your player count before deciding you need more.
Verdict by app type
Hypercasual and casual games, which depend on ad volume and often have global players, are the clearest case for mediation. Mid-core and strategy games have high-value audiences, so a lift on a higher eCPM is worth more, and bidding is worth the trouble. Run the numbers in the AppLovin MAX calculator and the Unity Ads calculator to compare.
Utility and education apps with modest traffic can start on a single network and revisit at around 5,000 daily users. Finance apps carry a high category multiplier (1.6x in our tables) but a narrower set of qualifying advertisers, so test before assuming a big lift. Apps aimed at children face policy limits on which networks they may use at all, so confirm eligibility first.
A low-risk way to test mediation
Do not switch everything at once. Run a single network for two to four weeks and record eCPM, fill and ARPDAU by country as your baseline. Then add mediation with two or three bidding networks and send a share of new users to it as an A/B test.
Judge the result on revenue per daily user, not on eCPM alone, and watch crash rate, start-up time and retention. If the numbers hold after two weeks, roll it out. Once mediation is stable, pair it with the other levers for earning more from your current audience, and check which formats to emphasise in rewarded vs interstitial ads. The mobile eCPM benchmarks help you judge whether your results look normal.