How to Maximize Your App's Ad Revenue: Leveraging Mediation Platforms
In-app advertising, alongside in-app purchases, is evolving into one of the largest and most efficient monetization models for app services. For engineers and companies running free services in particular, ads are often the most basic—and sometimes the only—source of revenue.
For Alarmy, the app we service here at DelightRoom, ad revenue also accounts for the lion's share of our total revenue. By testing and applying various strategies through Mopub, the mediation platform we use for Alarmy, we're generating hundreds of millions of won in monthly ad revenue. Thanks to this, we were even featured as a Best Practice case by Mopub. In this post, I want to share how we integrated and leveraged a mediation platform for Alarmy's ad monetization.
Leveraging Ad Networks Through Mediation Platforms
There are countless ad networks out there, each with its own strengths in specific ad placements or geographic regions, so you have to pick and choose based on your app's characteristics. Once integrated, you can manage these ad networks efficiently using mediation. It lets you view and manage all your statistical data in a single dashboard, which you can use to set priorities and send ad requests under various conditions (e.g., by country or ad placement). On top of that, mediation platforms offer a variety of features like adjusting ad refresh rates and capping ad impressions, allowing you to run multiple experiments to improve ad efficiency.
In Alarmy's case, over 85% of our users are based overseas. Since this user base isn't heavily concentrated in just one or two countries but spread out globally, we test and integrate a massive number of ad networks, including local ones that excel in specific regions. A mediation platform allowed us to quickly test these numerous ad networks. After integration, we continuously analyze ad performance and adjust priorities to boost our ad efficiency.
Mediation platforms are also incredibly useful for analyzing the ad performance of different ad networks against the same standard. Ad performance is typically measured by eCPM (effective cost per mille, or revenue per 1,000 valid impressions). However, if you look at this data individually per ad network, their definitions of a "valid" impression often vary (e.g., the moment the ad appears vs. X seconds after it appears vs. when at least XX% of the ad image is visible on screen). This discrepancy can lead to errors if you're using their native eCPM as the sole basis for prioritizing ad requests. By leveraging mediation, you can analyze multiple ad networks using a unified standard and easily avoid these pitfalls.
To dive even deeper into this analysis, it's a good idea to utilize the Reporting APIs provided by each ad network (which provide their ad-related data in API format). For instance, there are times when revenue data from certain ad networks won't sync with the mediation platform, leaving you completely in the dark. To solve this and conduct a more accurate performance analysis by comparing the two data sets, we at Alarmy linked the mediation platform's data with the data from each individual ad network to build and analyze a single, unified database.
Beyond Mediation
If you're looking for ways to further improve monetization beyond the standard features offered by mediation platforms, my top recommendation is to peek into what these ad platforms (including mediation platforms) are planning next.
Ad platforms are essentially at the forefront of ad-tech. They broadcast their upcoming ad strategies and new features in all sorts of ways, from newsletters to tech blogs. They have to constantly promote their unique selling points to survive the cutthroat competition against other platforms.
You can gather hints from this information to make your monetization much more efficient. For example, mediation platforms introduced a feature that dynamically changes the priority of each ad network by estimating ad unit prices based on historical data. However, some ad networks support this, while others don't. In response, our team internally developed a tool that predicts estimated ad prices based on historical data for all the ad networks we use. By using this tool to set priorities ourselves, we're driving even greater efficiency in our ad revenue.
As another example, mediation platforms have recently shifted their development focus from a waterfall model—where they request ads from ad networks sequentially—to a header bidding model, where they request ads from multiple networks simultaneously and pull the highest-priced ad. While the pace varies by mediation platform, this is the undeniable trend. Knowing this allows you to plan ahead: how will you integrate this into your app? How will you design a structure so that ad networks that don't support this model can still peacefully coexist?