About Božo: LinkedIn | GameBiz Consulting

ABOUT ROCKETSHIP HQ: Website | LinkedIn  | Twitter | YouTube


KEY HIGHLIGHTS

📊 While bidding is becoming a more and more prominent technology via which ads are served, optimization of mediation and waterfall setups is still an important strategy in maximizing ad revenues in mobile apps. Available tactics include:

• Price point optimization of non-bidding ad networks 💰
• Expanding optimization to non-tier1 markets 🌍
• Bid floor optimization 📈
• Latency optimization ⏱️
• Price ceilings 🚫
• AB testing new ad providers 🧪📊

FULL TRANSCRIPT BELOW

Web advertising has long been dominated by bidding technology. Advertisers with the highest bid would win the impression without much effort from the publisher. But mobile? It’s a different game. Here’s why:

• Web focuses on brand demand; mobile is about performance (particularly mobile games and apps).
• Web inventory mostly consists of banners and some interstitials, while mobile inventory consists mostly of video ads (meaning rewarded and interstitial), especially in mid-core games.
• Mobile requires SDK (or Software Development Kit) integration for smooth ad delivery, while other options are not very used.
• The ad provider landscape is significantly different. Some of the ad networks that are irreplaceable on mobile don’t operate within the web ecosystem, including Applovin, UnityAds, ironSource, Mintegral, DT Exchange, Liftoff Monetize, and others.

Despite the buzz, mobile bidding took its time. But 2022 and 2023 saw a shift, with giants like Google announcing a move towards bidding-only models in 2024.

Now, Ad Monetization Managers have relied on waterfall optimization for years. It’s all about strategically placing each ad network in several different positions in the waterfall (with different floor prices, which are called instances, placements, calls, or line items) to maximize the eCPM. Sounds simple, but it’s complex. You need to decide:

• Which ad networks to use?
• Which floor prices to use?
• How many times should each network be placed in the waterfall?
• Getting the best eCPM from an aggressive setup without creating excessive latency that impacts impressions/DAUs and user experience.

Given these complexities, is mediation optimization still effective? I believe so. 

Here’s why.

The average share of revenue of non-bidding networks among our clients is 54% on iOS and 59% on Android. Now, Google is taking 12% and 25% of overall revenue on each of the platforms, respectively, which means that, even when Google moves to bidding only, we will still have 42% of revenue on iOS and 34% of revenue on Android generated by non-bidding networks. Leaving all of this revenue unattended (without waterfall optimization) does not seem like a good strategy, does it?

In the past few years, especially in 2023, one myth was commonly cited by ad monetization managers – bidding is better than waterfalls. While bidding is efficient and hands-off for ad monetization managers, it sacrifices their control over the inventory. Our tests often showed that bidding doesn’t always equal better performance. In fact, switching to bidding often led to several key outcomes:

• eCPMs going down (because no floors are set anymore).
• Impressions/DAU going up (because there is less latency).
• In some cases, even impressions/DAU would go down without any obvious reason.

Depending on how much eCPM and impressions/DAU have changed, overall ad ARPDAU might go up and down. This is highly individual and depends on the following:

• Ad network in question and how sophisticated their bidding algorithm is.
• Game.
• Platform, meaning iOS or Android.
• Ad format.

We’ve run tests that showed no improvement or even negative results when we tried to move from instances to bidding. To name a few:

• No improvement when trying to switch from DT Exchange instances to bidding
• 2.5%, 3%, and 3.5% decrease in ad ARPDAU when trying to switch from UnityAds instances to bidding
• 7.3% decrease in ad ARPDAU when trying to switch from AdMob instances to AdMob hybrid setup, meaning AdMob instances plus Google bidding.

The latter was particularly worrying because we haven’t replaced instances with bidding – we only added another way for Google to access our inventory.

In light of Google’s announcement that starting 31st October 2023 (which was later postponed to early 2024) they would fully move to a bidding model, we decided to test removing AdMob instances from the waterfalls. This resulted in a decrease in ad ARPDAU both on rewarded video and interstitial ads on Android. This experiment didn’t even take into account the removal of GAM partner instances, which would surely have even more negative results.

All of the results above leave us thinking – why are some of the networks (and mediators) insistent on moving away from a waterfall to a bidding model when we see that there is no one-size-fits-all and that, in many cases, the hybrid model (instances plus bidding) is the best option in terms of performance.

While the debate on mediation setups continues, it’s clear that a significant revenue chunk comes from non-bidding networks. This underscores the importance of waterfall optimization for eCPM control. Here are a few actionable tips you can use to maximize ad revenue:

1. Optimize Price Points: Emphasize high price floors for non-bidding networks. I’ve found success with floors like $200, $300, and up to $900 where possible, consistently boosting performance.
2. Diversify Markets: With bidders handling about half of the revenue, there’s room to optimize in previously overlooked markets, especially tier 2 and tier 3 countries.
3. Don’t neglect bidders: Implementing bid floors can enhance control and often uplift ad ARPDAU, sometimes by as much as 12%.
4. Address Latency: Aggressive waterfall optimization can lead to latency. By analyzing waterfall-level and instance-level latency and removing high-latency, low-revenue instances, we’ve seen over a 5% increase in ad ARPDAU.
5. Explore Price Ceilings: Major mediation providers are introducing tactics like price ceilings to streamline waterfalls, reduce latency, and boost impressions. 
6. Expand AB Testing: With more bandwidth, it’s time to AB test a wider range of ad networks, diversifying ads and enhancing overall performance.

To wrap up, I would say that, based on years of experience, a hybrid mediation setup is the one that ensures the best performance in terms of eCPM and ad ARPDAU. Also:

• A significant portion of revenue still comes from non-bidding networks, and that portion should be heavily optimized for the best performance.
• By optimizing the bidding part of mediation, we can increase ad revenue by using bid floors. We hope mediation providers will offer more control over bidders soon. By doing so, mediation won’t become a black box that we can’t control, resulting in poor performance and frustration among Ad Monetization Managers.

A REQUEST BEFORE YOU GO

I have a very important favor to ask, which as those of you who know me know I don’t do often. If you get any pleasure or inspiration from this episode, could you PLEASE leave a review on your favorite podcasting platform – be it iTunes, Overcast, Spotify, or wherever you get your podcast fix. This podcast is very much a labor of love – and each episode takes many many hours to put together. When you write a review, it will not only be a great deal of encouragement to us, but it will also support getting the word out about the Mobile User Acquisition Show.

Constructive criticism and suggestions for improvement are welcome, whether on podcasting platforms – or by email to shamanth@rocketshiphq.com. We read all reviews & I want to make this podcast better.

Thank you – and I look forward to seeing you with the next episode!