fbpx

Our guests today are Liliya Gafitulina UA Manager in charge of all ad networks, and Bogdan Khashiev UA Manager in charge of social media traffic at AdQuantum. In today’s episode, we dive deep into the creative strategy that underpins some of the significantly scaled casual games owned by AdQuantum as Liliya and Bogdan pull back the curtains on what works for these games. We talk about user motivations that inform creative strategy. We dive deep into misleading ads, why they work, and how adapting that to onboarding and product experiences can unlock dramatic growth. 

Note:

We wrapped up the Mobile Growth Lab where over 60 marketers, executives, product managers and developers signed up to break the shackles of ATT’s performance and measurement losses. You can get access to the recorded versions of these sessions through our self-serve plan.

Check it out here: https://mobilegrowthlab.com/





GUESTS: Lilya Gafitulina | ADQUANTUM |

ABOUT ROCKETSHIP HQ: Website | LinkedIn  | Twitter | YouTube


KEY HIGHLIGHTS

🛫 How to start planning the creative strategy

🌞 How user motivation models can help with the strategy

📃Examples of creators using the models in ads

📲Transitioning from core ads to misleading creatives

⚖️ Misleading and meta-misleading ads

🧨The risk involved with using misleading/meta-misleading ads

📱 How does the appearance on the App Store change?

💰 The impact of misleading ads on monetization

🍀 The testing process in terms of geos

🕰 The approach on TikTok

KEY QUOTES

How to determine what the core game creatives need to be

we’ll first look at what our game can offer as a game mechanic itself, and then how we can implement this mechanic into the creative, make it a creative mechanic. If it works, we start to analyze what type of users it probably attracts.

The building blocks of a successful creative model

These are communication, mastership, management, self-expression, research, and escapism. These categories are also divided into subgroups, but this is the deeper part of it.

The transition from core ads into misleading creatives

We start to create misleading creatives after we see that all the core creatives, all these screencasts are burned out. Then we’ll start with meta-misleading ones by adding some misleading elements to the core gameplay. We see that meta misleading and misleading mechanics may work well with the audience that prefers to watch ads, instead of being in the game.

A different strategy for mid-core and hardcore games

 For mid-core, hardcore games, we tend not to use misleading at all, because here it is important to show the end product to the user. But we also may add some meta-misleading mechanics like storytelling, for example, before showing the core gameplay.

How misleading ads work with casual games

 in casual games, we have many ad watchers who prefer to watch ads, and they may be triggered by those meta misleading or misleading ads and pay off quicker. So yeah, when we use them mainly with the games with really long-term retention rates, the LTV is not that important but it is more important to pay off rather quickly.

How does the appearance on the App Store change

What we don’t do is actually not use misleading videos in the store because when the end user sees this page in the store, they should see the end product itself.

The testing process

There are two main considerations that you should keep in mind. The first one is the relevance of the test. You obviously need to get enough data, enough impressions, and respectively enough clicks to measure your CPI so that the results that you acquire are trustworthy. The second one is the cheapness of your test. For example, in Ad Quantum we ran up to 30 creatives within the same test sometimes for some projects. So you need to be wise about how you spend your cost because otherwise, the test can be way too expensive.

The day-one player cost model

We also use the so-called day-one player cost when we divide the CPI by retention rate. So then we get the cost of the player who sticks to the game on day one. That’s how we combine the retention rate and CPI together with one metric. 

 

FULL TRANSCRIPT

Shamanth

Today we have a special episode with two guests. Our guests today are Liliya Gafitulina and Bogdan Khashiev from AdQuantum. I’m excited to have them because they’re very much at the leading edge of marketing for casual games. As we all know, it’s become ever hairier after iOS 14 and SKAN and these guys have gone very deep into understanding how to manage creatives, how to do high-velocity creative testing, how to qualitatively understand creatives, and how to drive growth for games post-ATT. 

For all of these reasons, I’m excited to have Liliya and Bogdan on the show. Liliya and Bogdan, welcome to the show. 

To start off, when you’re first producing creatives for a game, where do you begin? How do you begin to understand what your creative strategy needs to be?

Liliya 

First of all, we need to play the game and understand the core mechanics, setting and graphic elements. After that, we make core gameplay creatives, like for example screencasts, in order to attract the most relevant users and after we understand how they perform within the testing sources and understand which exact game elements or setting elements attract the users the most, we start to add some misleading features. 

So for example, storytelling before the core gameplay, exaggerating mechanics, or something like that, at this point, we can set certain boundaries for what we consider meta misleads and what we consider misleads.

Shamanth 

Alright, so you start with the core gameplay, and then you say, “hey, can we figure out what kind of misleading ads we can add?” That certainly has been a big topic of conversation in gaming lately, and we’ll definitely get into that in more detail. But before we do that, you said ads based on core gameplay – are you using any sort of user motivation models? 

Facebook certainly popularized that with their Big Catch Playbook, but certainly within the field of game design there have been many other user motivation models, which we’ve actually talked about on the show before. But do you use these models to understand what the core gameplay creatives need to be? What do you think about that?

Bogdan 

We try to use different models because there are different approaches to how you divide users into different categories. The one thing I would like to emphasize is that we always first focus on creative mechanics as the key elements of the creative process. So

we’ll first look at what our game can offer as a game mechanic itself, and then how we can implement this mechanic into the creative, make it a creative mechanic. If it works, we start to analyze what type of users it probably attracts.

We can never be sure what kind of potential but we have experience and we have some knowledge about what type of mechanics attracts different users. 

This is actually up to our creative team. We have a special team that is deeply thinking into creative mechanics. They are free to use any models that they would like to use, as long as it provides the progression of the development of the mechanic. That the next iteration of the creative will be somehow developing the concept that we’re using at the moment. We also like to build a creative map. That’s when we first find something that works with a high CTR and adequate install rate, we then try to make some variants and some iterations. 

Then we put it on the map, where we can see what we change, and how to change the metrics. The key element of this process is for the UA team and creative team to communicate with each other, and for us, to provide them with metrics for us to build them the map of numbers, because we’re deep into the numbers of the creatives, we now know how they perform at the top of the funnel, and at the bottom of the funnel. 

It’s up to the creative team to tell us and explain to us what they have in mind and what type of users they’re trying to attract. That’s how we build this complex experience of both creative and UA Team. When we use creatives, we try to think about the users that we’re targeting.

Bogdan 

But speaking about the models, my creative team shared with me their most commonly used model, which consists of six related categories and they are divided by the need or the pain that the users are looking for, in-game and consequently in creative. 

So they are divided by these needs.

These are communication, mastership, management, self-expression, research, and escapism. These categories are also divided into subgroups, but this is the deeper part of it.

Shamanth 

For a lot of these games it certainly makes sense to use what works for you guys as games. Are there examples of how creatives might map to some of these models? 

Bogdan 

I think we shared some creatives from Gold and Goblins. At that point, when we used these creatives, it was I believe one year ago after we found the concepts and they were completely burned out. Up to now, it’s completely out of use because we spent too much on them. Later we try to think about what type of users will be attracted to this and what we probably should not use anymore because of the enormous cost that we spent on these creatives. 

This was a management-type creative as we saw it because as you know with popular creatives we need to match different warriors and towers and then match it with the warriors on the avatar so that the number of your warriors will defeat the number of theirs. This is kind of a management creative example. 

Now we use this for Lumber, a game that we are promoting at the moment. One of them you probably all saw, it’s collecting trees and then upgrading the factory. That’s also management. So, the audience is kind of similar to each other, but the creative approach is different. That is due to the core mechanic that we use from games which are producing factories, and then that will be put into the creative. 

We first tried to be mechanics oriented and then we tried to use these models to describe what we are looking at and then how we can develop it, maybe find something else that works. Actually, the towers that I mentioned, also worked for Lumber for a decent period of time, because we realized that probably the same audience but different segments of this audience can be targeted with different creatives of the same type.

Shamanth 

Thank you for those examples, it all makes sense. I know Liliya, you had said earlier about, starting with the core ads and then transitioning into misleading creatives. When does that happen? Are there trigger points in the game’s lifecycle when that happens? Is there a process by which you understand which ones you are going to test? How do you test these? Can you talk a bit more about that process?

Liliya 

We start to create misleading creatives after we see that all the core creatives, all these screencasts are burned out. Then we’ll start with meta-misleading ones by adding some misleading elements to the core gameplay. We see that meta misleading and misleading mechanics may work well with the audience that prefers to watch ads, instead of being in the game.

So for some casual products, this meta misleading and misleading stage comes rather quickly. 

But of course, we test the core mechanics templates. After we understand which meta-misleading mechanics are the most appealing to the users, we start to develop them further and find new ones. In the best-case scenario, we use misleads only at the point when we run out of any meta misleading core gameplay creative ideas because this may drop the short-term retention rate.

For midcore, hardcore games, we tend not to use misleading at all, because here it is important to show the end product to the user. But we also may add some meta-misleading mechanics like storytelling, for example, before showing the core gameplay.

Shamanth 

Can you define for folks who do not know what’s meta misleading and misleading and give examples for both if you can?

Liliya 

Misleading creatives, this consists of some totally misleading mechanics and graphics maybe and maybe even characters.

So I actually can give the same example of the towers for example, there were no towers in the game. And the setting was totally different with these creatives, different from the game one. So this type can seem to be misleading. For meta misleading, I will show Lumber, for example, where the mechanic itself is exaggerated. So in the game itself, you don’t have the big tractor that collects all the wood. But still, the mechanic remains the same and the characters remain the same and the setting remains the same. So only certain elements are misleading for this trade. All in all, it is related to the game.

Shamanth 

So I understand meta misleading is, you take the mechanic of the game, but it’s exaggerated. Misleading is, it’s not even present in the game.

Liliya

Yeah, it might not be represented in the game, or it might be in a different setting, different visual graphic elements or something.

Shamanth 

Go ahead, Bogdan.

Bogdan 

We had a debate about the towers and how they should be defined as a meta-misleading or misleading creative because the mechanic of one part of these towers is to merge some units into a bigger amount. That is actually what is present in the game involving goblins. It’s a key element of the game. 

But there was a debate with our colleagues. I think it’s still ongoing. We still sometimes argue about that. We think it’s more misleading because it’s just a little part of the creative, the merging, but the matching numbers. This primitive math that we use, and the antagonism of these two towers is we believe that this is the key element of the creation of the tower type creative. 

That is something that is present in-game; it’s not something that we use as a core mechanic of the game. That’s why we define this as more toward mystery. But you know, it’s like a percentage. It’s surely more misleading, but still, in some ways it’s meta-misleading.

Shamanth 

Liliya, I know you very briefly said the misleading ads don’t usually work for strategy and mid-core games, because you want the game to be very faithful to what’s in the ad. So with casual games, why is that not a problem? In other words, if the user sees a misleading or even a meta misleading ad because they go to the game, they don’t see that feature in the game, is there a risk for casual games that your retention might be poor, or your long term LTV might be poor, just because they don’t see the ad contents within the game?

Liliya 

There is a risk, but still these users sometimes pay off much quicker than the long-term LTV drops. So for example,

in casual games, we have many ad watchers who prefer to watch ads, and they may be triggered by those meta misleading or misleading ads and pay off quicker. So yeah, when we use them mainly with the games with really long-term retention rates, the LTV is not that important but it is more important to pay off rather quickly.

Shamanth 

I think that makes a lot of sense because you just need them to stick around for a very short payback period. Obviously, there are games that do change the App Store screenshots and of course, the game itself, based on what worked in the fake ads. What does that process look like for you guys?

Liliya 

Sometimes we use this practice because it shows good results in terms of retention rate and conversion rate from page view into installs. We sometimes use only 40% of ASO elements to reflect the core gameplay. We have our own experiment here. We can add some random static assets too, for example, some we can input with our characters, or end our settings with a screenshot of top performing misleading or meta misleading creatives but with the real interface. We don’t actually use very misleading assets with the ASO but still, we have room for experiments here. We also test our test icons, when we change the character, or maybe the most favorite asset that appeals more to the users in order to check which one will appeal the most to the user.

What we don’t do is actually not use misleading videos in the store because when the end user sees this page in the store, they should see the end product itself.

The essence of it is, it’s possible to use meta misleads.

Shamanth 

Does that mirroring of misleads happen within the game as well, within the onboarding, within the early levels, and later levels? Is that something you guys planned? What do you guys think about that?

Liliya  

Basically, from a long-term perspective, we believe that first of all, we can buy those meta misleading mechanics and check if some new features will be appealing to the user. So for example, we can add a screenshot of a nonexistent level with the city like coming soon and check if it is available to the user and add it into the game later. 

Shamanth 

Are there examples that come to mind?

Bogdan

The very first example that comes to mind is Golden Goblins, as we previously mentioned. At some point, they were added to the app store as screenshots. I believe that the developers and the production team were even considering adding some sort of level of the same type but that never happened. Probably because they eventually burnout. Then they thought that that was no good for the game itself. Later when we found a different approach, we switched the icon screenshots to the screenshots from the creatives and we still use this approach.

So once again, the exaggerated mechanics of what you should do in the game. That’s probably the case.

Shamanth 

Liliya, I think you mentioned earlier about, even if the long-term value is not there, let’s just say you haven’t changed the ASO, you haven’t changed the product, just have an ad, then you can get a very quick payoff if their ad monetizes. Now, what happens if the monetization model of the game is IAP driven? Or a hybrid where you need to make money back for a long period of time? In those cases are misleading ads effective or is that not quite the case?

Liliya

So basically, almost all casual games are both in-app ads and purchase monetized. We still see pretty good results because meta misleading can be misleading creatives. Sometimes it happens that the meta misleading or misleading mechanic is catchy for the user. After they start playing the game, they really evolve into the process, and remain there and still make purchases and still watch ads. So they still pay off.

Bogdan  

For the UA team, it’s very important to know the IP and IA ratio. So the ratio between the ad monetization and purchase mentalization should be known. Both of the current levels of communication with a production team are very important in how new changes affect this ratio. So the one thing that we should keep in mind is the CPM decay, as you all probably know, but I should still elaborate and that the CPM decay is the thing that when each ad is watched by the same user, from the first one, gives you less revenue, costs less. 

Having this in mind with such metrics as retention rates, and probably conversion rates, but they become less important. So they become less important in comparison with the low CPI that you should aim for when you’re targeting because you need them to be quickly monetized. 

That’s probably the main difference between these app monetization models.

Shamanth 

 That makes sense. To switch gears a bit, can you describe your testing process? Some folks test completely new creators in tier 3 geos. Then they move to Western Europe, then they move to the US. So what does your testing process look like? What KPIs are you looking for? How does this differ for different channels that you guys may be using?

Bogdan 

I should probably start with social channels because we use them as the first step in testing creatives.

There are two main considerations that you should keep in mind. The first one is the relevance of the test. You obviously need to get enough data, enough impressions, and respectively enough clicks to measure your CPI so that the results that you acquire are trustworthy. 

The second one is the cheapness of your test. For example, in Ad Quantum we ran up to 30 creatives within the same test sometimes for some projects. So you need to be wise about how you spend your cost because otherwise, the test can be way too expensive.

So your test settings come from the compromise of these two factors. My personal recommendations and what we use in most of our projects is to choose some cheap geos in terms of CPM, cheap geo like Brazil, Brazil is one of my favorite ones, and use the relevant optimization. 

For purchases, Brazil is one of the best countries to test your creatives. You obviously should test them at the same time. 

I also have a recommendation that you should stick to the chosen setting as long as possible so you can retrospectively compare which creatives showed the metrics distinctly and also use such metrics as impression per mil is a very useful metric to measure. I’m sure that most of our listeners are familiar with these metrics. But still, it’s a very useful metric that is not affected by the CPM whatsoever, it just shows how effective your relative is from the point of attractiveness. 

We also use the so-called day-one player cost when we divide the CPI by retention rate. So then we get the cost of the player who sticks to the game on day one. That’s how we combine the retention rate and CPI together with one metric. 

Liliya 

On ad networks, we tend to use those creatives as the social network test, because it is much more expensive to test on social networks. Sometimes it is impossible because we might produce 20 to 30 new creatives for a title per week and it is too much work to proceed. So we take into consideration that what works well on social networks might work well on other traffic sources, and use these creatives. Of course, sometimes we also run internal tests on ad networks, but they are mostly for end cards to labels, not applicable for social networks.

Shamanth 

Out of curiosity,  is your approach very different on TikTok? If that is something you guys run extensively?

Bogdan

As far as I know, my colleagues work on TikTok, they use the same test setting as we use on different social networks. From my experience, most of the results that we test within different social networks are very much transferable from one network to another. If your creative shows some decent metrics on one social network, you can be 95% sure that it will be as effective within the other network. 

Shamanth 

That all makes sense. I think this is perhaps a good place for us to wrap up. Thank you so much, Liliya and Bogdan. But before we wrap up, can you tell folks how they can find out more about you and everything you guys do? 

Bogdan 

You can visit our website adquantum.com I believe it has all the information about how we perform within our profession, and what we can offer as an agency. 

Shamanth 

Yeah, we will also link to your LinkedIn so if folks want to connect with you, they can do so. But for now, I think there’s a good place to wrap. Thank you so much for being on the Mobile User Acquisition Show.

Liliya  

Thank you.

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 at 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!

WANT TO SCALE PROFITABLY IN A POST IDENTIFIER WORLD?

Get our free newsletter. The Mobile User Acquisition Show is a show by practitioners, for practitioners, featuring insights from the bleeding-edge of growth. Our guests are some of the smartest folks we know that are on the hardest problems in growth.