Mobile ad monetization: 25 questions answered

Just a few weeks ago Singular assembled a dream team of mobile experts to talk about mobile ad monetization. We spent an hour dredging their brains for insight but utterly failed to find the bottom. (In other words, we ran out of time.)

You can check that out right here.

That dream team, by the way, includes Mary Kim from GameHive, Mark Allen from Neonplay, Noam Yasour from MoPub, Ido Naim from Voodoo, and Lisi Gardiner from Singular. Between them, there are literally billions of app installs and probably trillions of ad impressions.

But there were more questions that we couldn’t get to on the webinar. Now we’re going to try to answer them. (By we, I mean Singular’s own Lisi Gardiner, plus yours truly.)

All the questions and answers

Question: What is the min-max ROAS for hypercasual games in general?
I’m pretty sure we can all calculate the minimum, and that’s going to be a big fat zero. As for the maximum, there’s a wonderful answer which works every time but seldom satisfies: it depends.

Each game will be different, and each region differs as well. Let’s put it this way: successful hypercasual games tend to recoup ad spend in less than a week.

What ROAS number you want to put on that depends on your spend and degree of success.

Question: For hypercasual, what is a good breakeven day, D1, D3, D7?
My favorite but unloved answer from Q1 applies: it depends.

But if you consider that successful hypercasuals recoup ad spend inside a week, a good breakeven day is any time before that. Important note: ad spend is not all of your cost when you’re assessing breakeven timetables.

Question: How do you deal with auto-click and hyper-intrusive ads?
The original question was asked with a few exclamation points and question marks, so I assume you’ve run into that problem multiple times. There’s only one answer here, and it’s not my favorite from questions one and two.

You need to vet your partners. Don’t work with disreputable partners, and reassess your partners regularly. Be in your game, and maybe try to VPN into different geos to assess how your game and its ads are looking elsewhere too.

Bad ads will drive users away. Good ads will retain them.

Question: Do you use programmatic, traditional, or hybrid mediation and why?
The reality is that the industry is headed towards programmatic. Most mobile marketers and monetizers don’t like managing waterfalls because it’s a lot of manual labor.

The concern, of course — and it’s a valid one — is that you’re missing out on some inventory because not all networks have adopted the technology. You’ll have to assess if that’s a real enough concern for you to endure the waterfall headache.

Question: How do you suggest optimizing waterfall and ARPDAU for indies who cannot afford to hire expensive and rare to find ad monetization consultants?
The best (and simplest) answer is a kind of Gordian knot answer: set up your processes so you don’t have to spend time optimizing a waterfall, and you won’t need to hire those expensive ad mon consultants.

In other words: move to programmatic.

If you cannot take TLC’s advice to Don’t Go Chasing Waterfalls and simply must have a waterfall, here are a few options to optimize it:

• Check the highest eCPM per app, per platform, and per country
• Check fill rates per instance … low fill rates means you need to set up lots of instances, but too many instances can cause latency in serving an ad
• Check where you’re getting the most DAUs per app, platform, and country to maximize the impact on the overall revenues

Then, chill with some TLC.

Question: Is ARPDAU a function of DAUs or ad Impressions per day? Does ARPDAU increase for high DAUs? Is 500 DAUs a good number for testing ad monetization metrics?
Well, ARPDAU (average revenue per daily average user) is more a function of ad impressions per day than DAUs itself, although obviously total ad revenue will increase as you add DAUs. ARPDAU doesn’t necessarily increase for high DAUs, although you might become more interesting to certain kinds of advertisers as you increase in popularity.

It is important to note that there are very different ways to measure all of these things and there is no clear industry standard (maybe we should make one). And, you’re measuring average revenue per daily average user, not average revenue per impression.

Question: Has anyone started to experiment with other monetization formats such as audio ads, playables, surveys, etc?
The facetious answer is why, yes!

The slightly more respectful answer is that as iOS 14 is reducing available inventory thanks to some SAN (self-attributing network) policies and device identifiers become more scarce, many mobile marketers are turning to non-traditional user acquisition channels including the ones you mention … and influencer marketing.

Question: Is more ad views/DAU always better? Or is there an optimal number of ad views per day?
No, not all impressions are valued the same. In fact, the more impressions you show a user the less return you get per impression. Ultimately, you need to find a balance where you don’t essentially spam your user and lose them, but you are generating reasonable revenue.

Testing is your friend.

Question: What could be a correct ratio between interstitials and rewarded views by DAU before you start seeing either user fatigue or a drop in revenue for each individual ad, especially on Hyper Casual games?
I’m tempted to use my favorite answer here: it depends. But I’m going with my second favorite: test.

You simply have to test and see what will happen in your app.

Here’s a hypothesis to start with: rewarded views are going to be different than interstitials because people choose them. That may be better, but since they’re long, typically, that’s not guaranteed.

Question: Great panel! Super interesting insights so far. Someone mentioned that you should be using multiple mediation tools, can you explain why should I do that? And … is it in the same game? That sounds like a huge headache.
Flattery is always appreciated, as long as it sounds at least semi-authentic. You are right on both counts (the flattery and your suspicion).

The reason why some are using multiple mediation tools is that they want to test out different features in different mediation platforms. This is generally driven more by FOMO (fear of missing out) than any major feature differentiation. And yes, it is both a big headache and significantly costly on an operational basis.

But … it is possible, and at scale you may find some advantage one way or another.

Question: How many in-app bidding partners do you usually use in your setup and which ones?
Time to bring out favorite non-answer answer number 3: it varies. We have seen anywhere from 2 to 20.

Question: Would you include your ad revenue and impression/user level in your re-engagement strategy (for users that only monetized through ads)?
Thank you for the easiest question of the webinar!

Yes.

(Plus, you can build look-alike audiences off those users to try to find more.)

Why do you think it takes the in-app advertising networks so long to switch to bidding compared to the desktop world where this change happened within a few months? So far there are only 1-2 serious bidding partners.
We can’t take credit for this answer.

Noam Yasour answered this, saying that essentially, it’s complicated. It’s a big shift in business models from paying by install to paying by impression (even if a fixed CPM sort of already does that). Also, it’s a different technology: auctions happen on the server, but the ads are served by SDKs, and every SDK needs to be rigorously tested and vetted.

However, the process is ongoing and in 2021, most if not all of the networks will be on bidding models.

(If this answer is wrong, blame him.)

Question: What’s your appetite to use DSPs to optimize for ad revenue?
Ahh … another opportunity to use favorite answer number 3.

(Test.)

Question: How will iOS 14 affect ASA campaigns?
Not at all. Apple Search Ads doesn’t use the IDFA at all, thanks to Apple’s wealth of first-party data.

Question: Is there any industry benchmark of the opt-in rate? or how can we tell the opt-in rate is good or bad?
We’re seeing high rates right now. I’m currently collecting data and hope to have a report on this out over the next few months.

Question: Can you report different eCPMs by geo?
Did we mention that we love easy questions? In this case, the answer is easy too.

Yes.

Check out the full webinar

There is so much more insight and information in the full webinar. It’s a great idea to click on over to check it out and keep the learning going.