Screw Master 3D Live Review: 15% D30 retention & 80k+ daily revenue? The real hybrid casual KPIs

Sometimes we're swapping between RV and interstitial impressions. So even though you as a player requested to watch an ad for an RV, we might be showing you an interstitial ad. Uh which might feel to you like a faster RV experience in a way. The 30 30 seconds timer that you mentioned before um because the first impression of interstitial sometimes generates better performance. So we have some like sophisticated models behind the scenes that are trying to optimize towards the higher payer either if it will be an interstitial ad or an RV ad but for the placement that is RV in the game and that helps us squeeze the best TCPMs out of every impression. Let's say it's rolling the dice drops of golden ice with ads making those coins rise. Jack up designs worlds chasing the sky with the two and a half gamers the midnight crew talking UA adverts and

game design too maku bringing the insight. We're rocking those vibes till the early daylight master eyes on the prize tracking data through the cyerspace skies. Felix sters like a wizard in disguise. Crafters lift us to the high. Two and a half gamers talking smack. Slow hockey sick. Got your back. Ads are beautiful. They light the way. Click it fast. Don't delay. Uh-huh. Uh-huh. Ah. Ah. Hello everybody. Welcome. We have a very special episode because Felix says this is very special but we always have a special. This is a very very special episode. We are going to be like ultra special anytime soon. But before we start, my name is Mate Lancherich. I'm Felix Proberg and I'm and we are your host. Your hosting at the wheel. Maybe next time.

Maybe next time. Uh and why is this special? because we're talking about Screw Master 3D and we have Guy and Samantha from the team. So, it's a live review. Welcome guys. Can you introduce yourself and the company and the game and everything? Samuel first. Ladies first. Yes. Sure. So, hi everyone. I'm Samantha. I've been working in gaming in the last seven years plus minus in supersonic since the start of it. Currently I'm leading the growth and creative activities in Supersonic and it's great to to be here. Thank you for having us. Awesome. And I'm Guy. I'm leading the new games unit in Supersonic. We are working with various studios all over the world from ideation all the way to to global launch. Awesome. and we are here to talk about screw master 3D and kind of uh I mean we

started this journey of talking about the hybrid casual a long time ago. So finally we we have you guys here as well so we can we can talk about all the all the fun regarding the screw master 3D because we we read uh interesting things about retention profile and whatever else that you could you were able to share and how you were actually working with Chinese developers and how you are translating we had mini games towards uh the west. So let's start with our usual approach where Yahoo opens the emulator or go over the game and then we you kind of ask questions because we have a lot of questions loads of them and then also look we are going to talk about few things numbers and uh and then you correct us if we are in the in kind of the right if we're wrong we like Yeah. Yeah. Exactly. If if we are Yeah. If you're right just blink twice.

Maybe we can talk about it. Okay. Yeah. So, let's talk over the game as as usual. The first So, this is I think the fourth or fifth kind of game from the screw series that we're doing or if Yeah. It's growing like crazy. Yeah. Yeah. the like whole physics genre which kind of fits our unique puzzlers prediction that's supposed to be especially this one because this is again if I understand it's still deterministic puzzler meaning that the levels are still the same in between the rounds similarly like color block gem is currently killing it on on the other Lego front which is not really opposed but yeah the screw category let's say is growing very very fast so the game itself is if we start here following on the screw jam thing basically which was these plates that you had to kind of unscrew and there was pretty much two-dimensional gameplay really to it

and now we are moving more towards 3D uh same with screwdom and screw master is pretty much going for that I guess that also helps with UA because these shapes are as you see here quite familiar it's not really just like uh plates here and there and uh we also have the usual tension meanan mechanic which is coming from the bus gem and all these other kind of titles where there's this kind of bench setup where you cannot go over your bench otherwise you lose and you have predefined slots that you can or that you need to populate and these kind of as as we let's say we finish this one a new one comes in and then it kind of continues there the model itself is very kind of zoomable and elastic These levels take much longer time. As we have said before, this is not your typical two to three minute puzzle. These can take sometimes like 10 to 15 minutes. I've seen worse. So that's there. And uh

yeah, it's very engaging because again, it's physics. So once we start to actually unfold this thing, it's going to fall apart and then reveal more of these blocks as you progress and eat it like an onion from outside to the inside. Which then comes the hard part where decisions really need to be made where I going to run my bench full. the tension for these two ad placements going to be very very strong because suddenly there's literally no moves for me to move forward other than either using boosters or watching the ad placements increasing these things and yep that's it pretty much I think very very light meta for now like I just saw a few events yet that's one of my questions later down the line but yeah that's it can you give us a little bit more like your view guy I guess I I can't keep playing and talking at the same time yeah it's very hard to play and talk for before before you guys before you start like I I we started with uh with first screw well uh screw 3D

whatever game review some time ago and then someone from uh from China actually reached out like say screw screw was screw okay it was and I was like this is a very big hit anyway in China already and this is also something that you've you've done so can you talk a little bit about like how did you adapt this for for global markets like what was the the actual process coming from on yeah or when yeah or when like how did you come up from like we chat and like transfer this into into vest? So I will [Music] start. We are always on our way of exploring uh new markets and we focus uh pay a lot of attention on on onet to see what works well. We we follow the top charts what's popular mechanics. As you

said this cool mechanic is very popular in the last year. We see it not only in China by the way in Vietnam as well super popular and we are tracking the retention of those games and we try to find this we call it juice I don't know how popular is this term but to look for this juicy part in the game of course supporting by the data once we see it so we are getting deeper to understand what they miss or what can we do better with this game and that what makes us in this case uh to move forward with the EO games our partner our Chinese partner in this case it was that simple the retention was amazing from from first place we saw something very hooking very engaging with this game and we thought that with

a bit of work with some tiny changes, we can scale it up really really fast and and that's what happens. Sam will elaborate about the how we scale it, but the fact that we could do it in a very very fast pace says a lot about about the the hooking part or the juice part of of this game. So, how does that process look like in terms of the like you said the speed? So let's say today you find a great game with the great retention and juice in WeChat. How long does it take you to actually transform that into western like launch? So in this case it was like super speed like in uh less than than than three months. Yeah. Supersonic. Yeah. It meets it's definitely meets our our values and values.

So the speed in this case especially that this market was I wouldn't say saturated but it's definitely had yes yes getting cloudy it's a great term I would just also add that it's about working with the right partners and we we could give a lot of credits to the studio as well because they're very fast speed with supersonic and they were able to adapt and implement and follow the guidelines that we were giving giving them in a very efficient and uh high deliverable way. So uh that also helps of course depends on who you're working with. Do I get it right that there's a Chinese studio? Yeah. Yeah. And and and therefore okay this is the other question on top of it. Were they the ones that like captured the trend and told you like okay we should hop on this trend or was this some external team initiative that you have like research team or like I mean like who's the one watching the we chats

so we have our team in in Apac amazing team that based in on unity beneath uh looking for new games popular games we're not trying to invent from scratch anything We just like to find the things that most users like and once we pull it out we work very fast to to scale it high. Would you say it would be possible even if you wouldn't have a team in APE like a Chinese team like you would be able to do it from west? It's much more complicated. We traveled to China several times to to Vietnam. erh the the the local team is super important. I would say that the the the cultural differences and the mindset can be a bit different.

So you need to have your team that can speak the same language. Yeah, it was so important when we went to China Joe last year and then we were talking to the local devs and my brain was melting after a few hours there. It's just like it was insane and you can't get this experience from outside. You just need to be there on the ground. Indeed. Fully agree. Okay. So, Jakob, anything uh regarding the the game design? I want I wanted to ask right here. Uh, I'm just super curious like what are can you talk a bit about the session lengths and how many sessions per day you have? It's like pretty good for this job. Ah I thought it's it's a I'm asking what the session lengths are but you cut on an out so you don't actually to say a real number but yeah so so uh uh it depends but we do see

plus minus like [Music] three a day with a good portion of game play it really depends on the stage of the users if it's in in early stage of the game or later stage But we do see that the sessions are very deep. It means that they watch a lot of rewarded video. They are fully engaged with each session. So um we are trying to follow their engagement rather than a specific numbers. Okay. Yeah. Could we just go and and open up the the s data if we are in the Yeah. What's what are we talking about? Yeah. Yeah, cuz then we we can again like you can we can blink twice because we do these reviews and we only have the external data points and we have some like a feeling of if this is right or wrong. Sadness multiplier. Said we call

it sadness multiplier. So share your feelings with us. Yeah. Yeah. Yeah. And it's like if you got to the the actual sessions and then the active users and then we said we can we can check if this is true or not or at least like what what do we think? Yeah, right. It's between three to four sessions a day and the sessions as guy said they're pretty uh deep. So it's a a good estimate. Here comes the first question I have from the admon side. So there's a lot of companies from Vietnam and also maybe Lion Studios, I guess, that monetize quite heavily with an interstitial on app open or making it rewarded. What are you guys' thoughts on that? And I'm assuming you must have tested it, right? Not in this game, I will say, but definitely it's something we tested the app open interstial format. I won't say that we saw huge

success with it yet. And um it's a bit in a sense it's a bit aggressive. Yeah. Because even though it's usually a very fast static image or it depends on the ad type, it's something that it's at the beginning of a session and it's um one of the most valuable impressions in a way because it's the first impression a user is going to see. And so far in our AB test, I won't say we have a strong opinion about it. And it's definitely not something that we are yet implementing by default in new games, especially in high retention games. Interstitial is something that we're thinking about carefully in regards to how to optimize. First, we want to understand the users, their behavior to be able to do the best setup and experience for the lowest uh churn and the maximum LTV you can get. Yeah,

cuz the main monetization point in this game is the unlock box rewarded placement. Correct. It must get at least 70 or 80% of the rewarded ads because it's so prevalent. It's a hard game. It's not easy. Yeah. Yes. That's the whole point. It's it's it's a hard game but very forgiving. Like we are not punishing the users like we can talk about other meta events that other games are doing but we we we're trying not to punish the users. So even when you fail, you can start over and you have many exits for this failure. Yeah. like the revive we offer is basically the same as the unlock box and the unlock box in game is a very uh successful placement is generating above 60% engagement rate um and I think it works really really well because you know it's first of all it's the best in-game booster we have by far and it

makes it easier to complete the level faster for the users it's uh more dynamic and fan fun to play and um and we actually has seen that this sense of mastery in these type of games is something that users like players are seeking for. So giving them this opportunity to advance faster and to win at the first try by having all these different RV offers in the level is something that is very successful especially on a school master. I I play the game and you know I've watched a lot of rewarded ads as a result, right? And I I can't help noticing that majority of the rewarded ads are on a 30- secondond length and like 5 to 10 second end card. Watch the kind of thinking experimentation around that because it's probably one of the hottest topics for publishers because networks push for, you know, 114 second, let's say, right? And then they always say it's the best thing since sliced bread and then uh your attention figures

tend to disagree, right? So um I will say that we are not that strict with um how long an ad that is showing in our supply should be let's say we do see that it's becoming more common on the user acquisition side let's say to run longer videos longer ads and they do generate like good performance high CPMs for the monetization uh side but we on Android specifically I don't know like What did you play? iOS or Android? iOS. Android. Um, I played iOS, but this is Android. iOS. Yeah. Yeah. On Android, we are more strict because of all the better ads experience policy. So, we are watching out a lot for the timers and the length of the videos in our supply, but on iOS a bit less. And if you caught exactly 30 second videos

and 10 second end cards, it's probably by chance because you would also get longer ones. Yeah. Did you try it with the shorter ones and longer ones? What kind of eCPM degradation did you see as a result? An hour like on the monetization side. Yeah. So we did like No, because you need to basically if you want to AB test that it's like a very complex AB test. It's not a setup that you're doing on our side. It's a setup of the networks, right? And we can we do always ask the networks to keep certain guidelines in terms of length of ads, especially on interstitials, a bit less on rewarded, but it's not usually always um onetoone. There's always ads that are longer than what you expect because the user doesn't know where to click or how to click it in the right time. So it doesn't matter how long you set the benchmark for the whole experience is a bit different. We do use um ad quality

to track the overall uh length and quality of the the ads running in our supply and to make sure that we minimize churn. And if we see that there are certain creatives that are really causing like a high churn then we will go and block them remove them try to be more strict with their video length let's say so ad quality I think is the place we go for the analytics behind it and based on the specific cases we see we might go and um decide to do optimizations around that but there is not like a set rule if that's what you're looking for. Yeah. way. Do you use that those creatives that are causing high churn to inform your brainstorming sessions and then use those creative angles to create new creatives? Okay, we're mixing now. We're saying if I take I know I know I know it's a different it's a different thing. Yes, I'm just it'll be okay for me. But uh

basically of course we take insights let's say awesome from uh creatives that are running in our supply. We try to understand what's like what's working best what's taking majority of the impressions what's trendy beyond just what we can get from a sensor tower of course. So the creative team let's say is connected into the platform of the mediation in the sense of understanding the ads that are running in our game so that they can get inspiration ideas trends and so on. Yeah. Yeah. For those mixing things for those listening me crystallize this. So level play is one of the biggest advantages with the creative review tool there. It shows churn. So basically you can see what creatives are causing the most churn from your game. So clever UA managers have figured this out. Exactly. And when I'm talking to the different UA managers like this, oh my god, we can do this like Yeah. Well, of course you can. You just need to run. Yeah, you just need to

use level play and add ad quality. That's it. Yeah. Do do you want to go through the numbers maybe and then I want to ask the rest of my admon questions after that? Let me let me finish the product first because I think that's the part that I we hopped around a lot, huh? Yeah. I mean that's why it's called discussion. I mean like come on. So the why I'm still stucking there in the main menu that that's my question like what would be next or what's the current plan because I know there were the theme park mechanic that we were talking before and then suddenly after I updated the game recently they're no more. So do I get it right that I have the latest build here and this is this is how it's supposed to look like. Yes. So, so we are testing the theme park is definitely a good part of the game, important part of the game, but we we need to to remember we are alive like more or less six months and we always exploring new ways to improve our product. So, we take it in and out and try new stuff and AB test and probably are part of an AB test

right now. So yes, the theme park is is there and we even um supposed to take it one step higher and uh to combine some um limited time events on this theme park like not only a theme park, it has a for example a Christmas rep or um whatever St. Patrick's rep. So we are working on it and I guess you are part of an AB test right now. Okay. Because I I know I saw it and for those like uh watching this this is basically how it looked like. There was this kind of symbol during the game play. It would have this percentage as as you fulfill it and that at the end of the level again this is something different I guess another variation rather than team park. This is coming from something like we did the dreamy dreamy rooms was the game like that. That's that's that's a similar run. Again, another inspiration, I guess, for

some meta progression. But do I get it right that basically what you're doing, you're testing different kind of visual based meta progressions and trying to find out which one sticks the better in the current state. Still didn't figure out the one. Yep, indeed. Actually, we want to make sure the players always feel a sense of of progression, especially since some level can be longer as you as you mentioned earlier. H that's why after every level, we have a new decoration. and it's added to the park and we have some kind of we try to have some kind of a smooth animation that shows the the progress. The the the key is to always give to the player a feedback to to their game play uh some kind of accomplishment. And that's why we have this on the right side and top right side we have this H progression that once you fill it you can see how it was placed in the theme park or as part of the meta. Okay. And uh okay that that

that sounds right. So my guess is that like at some point somewhere down later line you'll figure this one out and that will be the like the main visual progression. Similarly like some other games on hybrid casual genre going really back to something like uh I don't hexasord or block 3D like they all did this like unlocking picture and stuff like that. I know it spread like wildfire like two years ago. My other question is what about the event meta? Do you do you plan to kind of hop on the usual bread and butter tactic that everybody is doing now which is kind of similar to royal match where you have these very speedy trigger events popping up from left to right kind of driving the user towards the main progression like play level base saga at goal cost basically then we give you more motivations track your progress in between with like leaderboards heat time events so on so forth so not for now I have to be honest yeah uh uh since we have more like head

oriented model and like I think a solid difficulty curve if you feel it's it's harder we can we can debate about it so we are um let's say we are not ignoring it but we we wait with this approach and maybe for later stage or we wait for this game to be a bit more mature. The games require some kind of skill and luck elements. So, we're trying not to punish the users. And when I play Royal Match, which is amazing game, I always afraid to to fail to lose because of streak mechanics. Yeah, exactly. Exactly. Because of the streak mechanics and at the moment, we don't want the users to feel that way. More relaxing experience basically. Yes. Yes, although it's I can accept it's the

levels can be a bit hard and we still want to keep it more relaxing and maybe in later stage when the game will be more mature we will take this approach but at the moment we prefer to take it solid. Let's let's finish the difficulty question because this is my like really one of the biggest questions that I have here is in screw master when I was looking through the other screw game especially once you review screwdom like the genre genre leader with IP revenue in those games or let's say in screwdom you can click on an object and make it transparent to see screws that are behind it over it pretty much like make it make it somewhere that you put an example here so what I'm talking out. Yeah. So, for instance, these blocks, I don't know what's there, but if I click on the block, I see what's under it or what's what's hiding, which makes the thing much much easier. So, which is very smart, I have to say.

Yeah. Yeah. So, was this a deliberate decision that you follow that like for instance difficulty curve is better this way or did you tested it or what's your take on this? So um since it's a school master is still a new games we we are at the moment focusing on more refining the core the level design the APS of each level before we moving forward we we we are actually used it for inspiration the the school through okay as an inspiration for for some new ideas that we are working at the moment. But for now, we are focusing more on improving the level difficulty to find this sweet spot. And we believe that the see-through can can uh disrupted. Yeah, exactly. And that's why we prefer not to to develop it at the moment. And we are

working on some some great surprises that are inspired by this uh great feature. Yeah, that's great because I I I can fully understand that like it could pretty much take the game to the point where your design space to create difficulty gets very limited and I guess that's like a giant hammer on on the difficulty pretty much. So yeah. Yeah, I can understand. Thanks. Sh then move to the numbers then or Felix what what do you want to do? Yeah, numbers and then I had questions. Yeah. So what daily active users is your favorite metric. Let's look at downloads uh first and the the IP revenue right and then just to get an idea. So downloads are unstable somewhere 40k a day again as multiply replies I think so but then with us being first so pretty good traffic and then revenue wise again this is only IP we don't see ad one which then that's

the question like what's the actual revenue split the ratio the revenue split here it must be 80 or 90% ads because You watch so many rewarded ads in this game. Yeah, that was my my assumption as well based on the active user base as well. I mean, come on. It has to be. Yeah, we we can we can uh confirm that it's a very heavy IAA game. There's no doubt around that. But um it still generates some inapp purchases and we're still as guy said on the very early stages. So there may be changes that we see going forward also regarding that. Then in fact the most popular sorry IP products are are like the the skipits and the boosters and it it says a lot about the fact that the main focus on this game is like on rewarded

videos. they watch the users watch dozens a day. Okay. So if you go to actual retention profile if you have one then we can we can then compare the the notes because you released some numbers which says where I have it here day one retention is 45% but then there's day 30 which is 15%. And then like is also let's see what sensor those are US numbers or global or what's the I can I can US okay well okay yeah we have US we have one 11% day 30 which is okay so we have 15% and then I would say day 90 should be also quite strong we have four here but what I would say what six seven at least yeah definitely we like this is a I think what makes a screw master in our

eyes at least a hybrid game. Like the way we're looking at a hybrid game is not necessarily it has 6% AP revenue but it's mostly about the stickness of the users how long they can play this game for and the deepness that we can go in it and screw master has like pretty good benchmarks I think for the world of hybrid casual when it comes to retention. we do see around like between five to seven% average let's say on day 90 and um I think this engagement that we generate in this game is what leads us to keep it at a high skill for such a long period of time um and it's like a it's what we're looking for you know the revenue comes with it the moment you have like a good strategy on ad monetization and with time we can also optimize further towards have your IEP if this is what we are aiming for you know it's all a matter of like building

the right game with the right potential yeah well that was my question next like okay well can you imagine if you flip the the 80 20 to 5050 this can generate such high amount of IIPS which I mean it's easier said than done obviously and I'm the UA guy so I think it's So my question there is just like with those insane retention numbers, right? Uh you do monetize with interstitials. Interstitials absolutely destroy retention but you still have quite a lot of them. How do you balance that? So I think in supersonic like the way we're looking at the ads optimization we call it like full screen ads. It's a combination of both interstitials and RVs. But you're totally right that

interstitials usually have like a negative uh correlation to retention let's say and it's something that doesn't go hand in hand. We are trying to first, as I said, understand the user experience. You know, see how they're playing, what they like, how they engage, how often they're watching an RV at understanding when they get the hook to the game before we add interstitials into the pie. And the way that we are AB testing interstitials is a lot softer, I would say, than the usual IAA games. So here it's on clean interval when it feels like very like a very natural pause for the users in the game in between levels and it doesn't seem disruptive let's say and we know like interstitials in a way even though it is negative towards retention if balanced well is actually something that we're looking at as like a pressure point even to convert users into IEP

right and I think that's one of the beauty ities of what the hybrid market in general is showing to the industry is that if you do it in the right way, you will minimize the impact to retention, but you also push those high quality users to make their first deposit even if it won't be a high value deposit because it's very common if you think about it to pay so you don't have an ads experience in many different platforms, right? even outside gaming and interstitials is that sort of like mindset for the gaming world where you have the option to play in a more premium environment if you make a payment to remove ads and or buy skip it or other formats that you have that are like lower cost and it gives you as a studio let's say a very early signal into who are my high quality potential paying users from a very beginning stage you cuz those will convert the convert

fast. Okay, that's super interesting. So, how do you control the cooldown timers and how do you kind of segment the user base on the back of that? How do we control the you the cool down? You mean like what technology? You could I mean do you control it or does everyone have the same ad experience or what Felix is asking like how do you do your segmentation if you have any meaning that you don't repeal the payers by interstitials too soon rather kind of move it somewhere there later until they manifest and then split it into these guys going to get interstitials these guys not because those are pairs or any kind of similar system. Jacob I'll just say you speak very fast. Yeah. Oh yeah. Welcome to the Welcome to Two and a half gamers. Is this just take that into account? So much knowledge. No time for slow speaking. Welcome to Two and a half gamers. This is every time course and get an answer for you. How do we do it? So, you know, like first of all, as I said, we don't do it from the get-go. We

first of all run clean of interstitials on the beginning because of the sensitivity of it. we understand the KPIs and the players um journey throughout the game. Let's say once we feel like we understand our players, that's when we'll start thinking and brainstorming about different setups for interstitials, different triggers, intervals. We're usually AB testing. We don't have something that we go live with without having first of all an AB test. And the AB test takes into account like we have pretty sophisticated like AB testing systems in supersonic even from back in the hyper casual times. And we are looking at all of those KPIs. So we're taking into account depositors rate. We're taking into account total you know rpu we're taking into account total LTV trying to understand the best group and to do it for a long enough period of time that is uh what we validate as a good uh statistical

significance for us to go live with. So, so again it's like it's it's a matter of optimization I think and the main points that we optimize is usually facing placements triggers you know this these sort of things and try to set something that is a bit more standard usually users that will have a high engagement with RV have like a longer grace period for interstitials because the timer or the way we're looking at it is kind of connected as I said as a full screen experience. So in games like Screw Master where you have such a high engagement rate from rewarded ads, you're usually naturally not seeing a lot of interstitials even if we have this sort of interstitial setup because the the intervals they take into account any sort of a full screen ads. So, it makes the RV engaged players have a more premium experience and still lets us

monetize for those that don't. And as players, usually this is happening pretty quickly. Yeah. Like the FTDs, the first time deposit happens like at a very early stage in the game. So, we are already able to understand who are the payers. And what I said about pressure points of interstitials is for those that didn't do that on the usual early stage, we still push them to do it at later stages by having these commercial breaks. Very cool. So I guess that's a really good segment for my ad revenue estimate here. So I think that you guys are generating anywhere between 8 to 8 and a half rewarded ads on 65% of the user base along with five interstitials on 75% of the user base. And according to my estim 5% you're saying yeah on rewarded and 75 on interstitial.

Uh, so eight rewarded ads and five interstitial ads, meaning that you're generating anywhere between 40 tok right now from ads. I can tell that's too low. That's too low. I can't tell by the look at like honestly. Yeah. Like I'll say that like Yeah. First of all, yes, it's wrong. It's very low. It's a lot higher than that. Even in the Dow sense, the amount of impressions we're showing is way higher than what you're saying. But think about it that in a good game, not a great game like screw, usually you're getting a 40% engagement rate. Yeah. So in these type of games where you're heavily dependent on RV, it's part of your progression in a way. Then it's mostly speaking to almost all of the users like the engagement rate is very very very high. Interesting. So it's like a 85 to 90%. Yeah. 80 plus. There

you go. So my estimate feelings would be anywhere between 80k to 100 at least per day. Yeah. So that 90% brings it up to 75. So that's that's 100. Okay. I I have a really interesting question. It's on the high end of what you're saying sometime like I won't say the exact number but it's definitely towards that high. Nice. Here comes a really interesting question on the back multiplayer. Yes multiplayer. Even a broken clock is right twice today. But here comes a really interesting question. So if you have that high rewarded ads the impa right. One of the main things that happens is that you have uh eCPM degradation because of the recency effect. Like networks pay less for subsequent impressions than they do the first ones. What are some strategies that you guys use to kind of keep the eCPMs high when you have such an high imp? So it's a great question but um yeah so first of all I

would say that you're totally right that this wear out of the eCPMs or the degradation of the eCPMs is something that is quite painful when you have like a game that is so highly engaging with ads but it is what happens and I think that the way we are optimizing it is through different uh sort of optimizations. One thing that we have uh in the back end, let's say that it's not something that you see is that sometimes we're swapping between RV and interstitial impressions. So even though you as a player requested to watch an ad for an RV, we might be showing you an interstitial ad which might feel to you like a faster RV experience in a way. the 30 30 seconds timer that you mentioned before because the first impression of interstitial sometimes generates better performance. So we have some like a sophisticated models behind the scenes that are trying to optimize towards the

higher payer either if it will be an interstitial ad or an RV ad but for the placement that is RV in the game and that helps us uh squeeze the best TCPMs out of every impression let's say. Um and in addition to that we also have of course what the industry has yeah as a benchmark which is capping mechani mechanism. So we are optimizing around how many impressions to show users on each of the placements especially the placements like we mentioned before that are top engaging placements. Usually you want to generate the highest engagement rate that you can because that means the first impression of a unique user and that every single unique user is very valuable. But once you're already squeezing that placement into 15 plus impressions, you sometimes prefer to cap it. So you keep the experience a bit more premium in a way. You give users that also that like nudge

to go towards maybe an AP direction rather than just ads. But the important I thinking behind the capping is just to make sure that the users don't feel like they're being punished that it feels a bit more natural and not like a something negative or a penalty in a way that you're capping these sort of rewards because as guy mentioned as well we want to give the players like this um um what's the word you used for it was a very nice word I forgot now giving approach yeah this like easy experience or um give them the chance to pass the level, you know, and not have that difficulty too high. So basically, yeah. So we want to give them the ways to crack the level to feel like it's easier faster, but we also want to cap the rewards to make sure that we are keeping some sort of like premium CPMs um for any advertiser that is buying out of a screw master.

And you never hide the unlock box rewarded placement, right? That's uh that would make it too hard. Have you had any luck experimenting with the really cool feature on level play with the uh where you do ad units where you can optimize on that and you do setting up bid floors and you send basically uh different requests to different ad units. You mean like the like what legacy wise it's called like a waterfall by placement or multiple? No. So essentially it's like it's a really cool feature where you send the first couple of impressions to one ad unit and then on the back of whatever eCPM floor you saw for the user there you then send it to eCPM buckets where you have bid floors that artificially keep the like the bid high. Yeah. Sorry. So basically having multiple waterfalls you're saying? Yeah. Bid units essentially. Yeah. Yes. So we are testing that. I would say it's like

a relative like if it is the feature I'm understanding and maybe I misunderstood it still but like if it is the feature I'm understanding it's a relatively new feature in uh level play and we're definitely testing it to try. The idea behind it is pretty similar than what I said before between ECPMs of interstitials and RV. Here is just about setting the floor. Today with biders, the problem is that if you set a floor and you don't have multiple waterfalls, then this is true for everyone and then it kills your revenue a little bit. So by having the multiple waterfall setups and getting like competition to to be tougher in a way, it helps us raise the the beads we're getting, the CPMs we're getting and the overall outcome, but it's still something that uh we're testing. Yeah. Like I can't say that it's something that we're fully live with because it's relatively new. Yeah. Final question before we move on to UA. Skip it. How did you guys test that? How

did you decide on it? It's a very hot topic. It's a It's a hot topic and it's a old topic I think. No, like nobody figured out the solution. We'll make it as a hot topic for today. So, skip it. Um, they kind of became like in our eyes at least the standard. Yeah. In hyper casual and especially in high RV engaged hyper casual games or hybrid casual space in this case. In my opinion, it's just like a rec like a recognition of the value of rewarded ads and what we're offering and especially in games that have a high engagement rate. I think that skip it help us also save that how did you call it degradation or something like that CPMs because instead of having another ad you would just this time skip it right so users who love the game but they don't want to spend too

much time watching the ads and you know like they would just buy these packages of skipits it's usually part of like most offers we have in the store and the mindset behind it. It's almost like a subscription model. You buy a set number of skipits. You have like a x amount of play time potentially that you can do without having to watch ads just skipping it. It's a better experience. It gives the users a taste of what it is to be premium. You know, we all like the free trial periods, let's say. So, it's something around that in terms of like I think of how it works and how it speaks to players. But, you know, screw master to be honest with you, we're just like going with skipits from the get- go. So it's not like we did crazy sophisticated testing behind it because for these type of uh offer we do see it's a pretty standard thing for us today and uh it works and now that that I would say the

economy becomes a bit more complex once you add skipits. So we are trying to fully analyze it and understand how we can make it better in the sense of like more like well fit and integrated to the overall economy of the game and uh how to evolve potentially from it maybe gems and not skips maybe a different currency that goes more like that speaks broader uh to the game but uh it's still something that we are working on let's say then while you are talking Jacob watched another rewarded video in order to we appreciate your money. Yeah. So, so do I get it right that you consider it pretty much as a part of a normal feature set for heavy ad driven games. So, that's why you didn't like hesitate or like IB test because I know we talked with this about multiple hybrid hyper casual publishers and they were kind of on the fence like this game it worked for us this game doesn't work for us yada yada yada. So, I guess it's

like works for adriven games period. So for the games that are very high engagement, a ton of impressions a day, like impressions per D a day, I think it has a huge value. Usually we don't see like we we tested it enough. Yeah. In the past, I mean, and we had games that had these sort of KPIs before. So we usually see it's a positive um result, a positive lift to the LTV. And uh and it can also help generate purchases, right? Usually so it's it's like a win-win and we just went live with it. Now there is a thinking behind the economy of Skip Bits. Yeah. Like it's not like we go live with any sort of setup. Yeah. It's not something that we are besides for the economy. I feel like it's something that it became pretty standard for us. Let's go to UA because Yakob is stuck.

Nice. Yakob is stuck. Yeah. Yeah. Yakob is stuck. You can watch another unlock box. That's fine. It's fine. It's okay. It's okay. It's okay. So, well, always talk about the UA of of kind of let's call it hybrid casual games and it's changing quite a lot. And uh I I mean I'm pulling out like the May 2025 which is now when we are recording but I'm pretty sure like the the UA channel mix changed quite a lot since the launch until now. So what was the like kind of initial UA channel mix or the strategy when you were launching the game? Mhm. Um okay so um basically we have some sort of like a standard in supersonic which is usually starting with big channels that we collaborate with um so on the beginning of a launch I would say we're opening about four different channels Google Facebook Iron Source and

Unity okay and that was also the case for screw master we launch the campaigns in these channels some of these channels and some of the campaigns we use they require some data and signals before we are able to open up like VO and okay and such. So we wait a little bit with it but um these are the channels that we are starting with and then after we have a little bit of a better understanding of rowas across different channels. Okay. Then we are starting to open the the rest because we work with a a lot of channels. Yeah. I I see. Yeah. I see basically everything in in sensor tyrant. I'm pretty sure like there's there's more. Mhm. Okay. Yeah. So yeah, so basically that's how we are launching it. And um for screw master specifically, I would just say that one of the um unique things we saw even though you know like supersonic iron source is like a family but we don't usually see that what we saw in screw

which is basically iron source scaled like we opened budgets in iron source and it scaled like super aggressively like we managed to increase our spend like week over week from like a I don't know much lower numbers to like 50 60k daily spend just on iron source nice without damaging the rowas and one of the signals that we are always looking for when we're launching is capping you know how fast we are capping on our campaigns h that's always like a good sign for us if we're capping quickly and we have to open up budgets so here we just went very high because we saw that the cap was happening very often and it exploded And it was like pretty insane because we haven't seen that since like you know very back in the days hyper casual times let's say and um and that gave us the understanding that this game is going to be very big and very good for us. Yeah.

Nice. Uh so and then wait a second let me check if you are so you mentioned VO. Okay. I don't see Tik Tok here. So I guess you're not running anything on Tik Tok. Oh we are. Okay. I wouldn't say that in these we love sensor tower we work with them but I wouldn't say in these dashboards they have the most u again so I'm also looking at impressions it will always be integral yeah I see like integral all over the place and unity and a little bit of YouTube but I guess like the the current UA channel mix is different for you. Yeah. Yeah. No, we are running definitely. We're running with Tik Tok and we're running with like basically all of the different partners also incentive channels. Yeah. Nice. So then what's the the percentage of the the budget going into different channels? Do you have like one big ones or like the

the four that you mentioned are the big ones and then the rest is just kind of there there just just in case. Also there are some big ones that are also not in the four that we start with. the four that we start with is just the ones that we set as good launch partners, let's say. Okay. So, the the distribution is uh different. It depends if you're looking at iOS or Android first of all. And um and I would say that usually we do tend to see like three channels that are dominating and then the rest competing each other for the like for the rest of the budget, let's say. But usually there's about three maybe sometimes four that are very very strong in a game. And it's also the same thing for screw master. Okay. And you're I guess you're optimizing ma mainly for address rather than just anything like Well, no, we use blended campaigns as well. Why not? You know what I mean? Like it's not it's not like

damage anything. It's just going to give us the additional lift that the AP revenue has. And we do want that to be a signal that we buy from. So whatever we do have hybrid or blended options, we are using that. And where we don't have, we're using separate campaigns like VIA on Facebook and V, right? So we are optimizing for both still. We have enough signals for that for the AP campaigns. Okay. But we're trying to get all the revenue users type that of course. Of course. And I see obviously like based on the channel mix you have the videos and playables. But I guess still for this type of game I mean the playable is always like better kind of format because you can play you can get the feeling and kind of qualify users basically. Yeah. So, um, so to be honest, like you're right, like I don't have much to say. It's true. The

playable format, it's what we're seeing basically with playables is that it generates like um higher engagement in terms of the users that we buy like they stay for longer, they have better retention. Playables usually are not um scale oriented in a way. Yeah. like they're more like the IPMs, the best IPMs we're usually getting from videos, but the best performance overall, especially for games like that, is coming actually mostly from playables. And I think that um it's it was for us at least I won't speak for everyone but also from our partners in Google they said that this is very unique that to see that playable standalone is something that was taking most of our spend also on Google which is not something no okay so you are actually the second company they were able to say they are can get scale

on playables on Google. Seriously, I've never seen that. I thought we were the first. No, but um our partners on on Google also said that this is not something that they are usually seeing. But but yeah, it happened like we had a very good playable um that we created based on a video and it just took scale over all channels basically not just SDKs which is usually what you expect. So in that case like how do you how do you approach the the creative testing and then how often you refresh creatives at that as this scale because you mentioned 50 60k per day I'm pretty sure like you are somewhere plus or minus still maybe a little bit more when you thought you said you have 100k per day kind of I mean not saying not saying you said that but alluded that it might be in that range kind of assuming it's either you spend anywhere between 80k or even maybe if

you kind of optimize for let's say 90 day payback period or more you say you spend even more than 100k but I mean that's a different part of the question let's get back to the the original question we spend a lot and enough let's say for this game to succeed and to to reach the best ROI we can but the way we are Like we first of all we have like a very big team in SuperSonic like a very big creative department to be honest and the way we are working around hybrid games is a little bit different than the way we used to produce for hyper casual like where in hyper casual it's all about you know quantity and scale and in hybrid it's a bit more about post install metrics performance and you know if you have a scale from a video that is very good but the quality is really low then you won't keep going with that right so you need to iterate so on the very early

stages of testing a new game we're trying everything basically levels color gameplay moments trying to stay true to what's actually the game experience that players will get in those early stages understanding the motivations that users are looking for which I said in in in screw Groom master it's like a sense of a mastery relaxation progress this is what's working for us afterwards once we understand what we want what elements from that we want in our creatives that's when we are uh iterating like we of course do a lot of market research we're looking at at quality sensor tower trying to understand the trends within this growing genre let's say yeah understanding what works what's picking up and trying to add those sort of like hooks into our uh creatives to sustain like a good CPI keep scale

going and you know basically we are producing I would say quite a lot of video and playable iterations and some percentage of the time also trying new completely out of the box concepts but what we see works best is mostly like going with the trends iterating ating on what's working a video that is working really well just doing like voice over on top of it doing UGC style uh together with the video or the playable and um trying to take the best videos and make a playable out of it or the other way around. So we are like trying to maximize on what we see is working after we are at scale and a little bit also explore new things until we find some new concepts that might be good and uh work as well but it's harder to to generate success from new concepts I would say. So so how many new creatives do you produce a week? A lot. A lot. Okay. Yeah, I mean I would

I would assume so I don't want to say a number to be honest 20 to 30 which is like the bare minimum and what you just I mean I'm playing while you you were talking I'm playing like I think sixth playable which is completely different concept so that's quite a lot which is amazing like different levels you mean right yeah yeah completely different yeah different playable like this is well this is a video but still I'm pretty sure like this is eventually a playable as well because yeah well this is again like a new playable which is which is quite uh unique concept unplayables also one of the things that we are always of course looking into it's like the longer experience right like the number of taps how long will like players need to engage with display until they're sent to the store and as I mentioned before like the IPMs usually that We are seeing

at least I don't know that this is true for all of the industry but at the moment on on screw specifically we see that playables generate lower IPM although better quality and they dominate scale right like they are still taking most of the spend. So we we understood also like that there's a lot of optimizations you can do around just the number of taps or how we call it like a endless playable experience you know to get the users to really hook that's what I'm just doing now. Yeah, it's like it's I can I can tap and just play forever. Would have finished it by now. I would just say that probably as but yeah. So trying to do this very long experience where you don't even remember you're playing an ad anymore and you really want to just keep going, you know, and that's usually also generating very good quality users that do come to the game already understanding what they are getting. Yeah. Because I think we had this discussion

with some of our friends that they said look like longer playables basically work way better for them because again like you said you qualify users and you can play until forever. But then there's the the other opinions like oh well but then players just will play the playable and then they will just churn and then they don't care about it anymore. But do you see better performance for longer versus shorter playables? So it depend like I would say in terms of post install metrics we see good performance with long playables yeah good like retention good long-term KPIs but it's not the only thing you're looking for also right so the shorter playables experience they're more I guess catchy in a way and have higher chances of uh converting and um then it's a balance between both you know like as long as you're keeping it true to the game play and you don't completely crash Ashure rorowass should be both of them should be like a best

practice you know testing it and each channel also UA channel I mean may bring different performance we do see by the way connecting it to the monetization let's say that the longer ads they don't run it necessarily on the same supply as the shorter ads. So it might be that also by having this variety of longer playables, longer videos versus very short videos, very short playables reaches different type of apps, right? Like supply sources. And then we are kind of like growing the incrementality of what we can get because they're not cannibalizing each other or competing with each other even with being the same UA channel. So yeah. Yeah. Because when I when I always talk about the the creative depth or like the number of concepts you have like you kind of you you need to diversify like you have static images, videos, playables. But then this is another level. You have shorter playables, longer playables. It's the same thing with the length of

the videos. You have 30 second videos and I saw one minute 30 second long videos. Again, so many different options just to get better scale. And Google you have the bumper ads like 5second videos. So it goes all the way. Yeah. Yeah. Okay. Nice. So you said you have all like the big team on the creative side which is amazing. A lot of the different people don't have that but you have so many different concepts like how do you then test them like on different channels maybe? Let's say how to test them on the biggest channel you have. Yeah. So um so we have like a pretty set testing process I would say that initially is on is happening on Facebook. Yeah like this is our main test platform I would say for creatives because we just upload a batch of like a few and we can compare the the the best performing ones quite fast with low

budgets. But in general for hybrid games I would say Facebook can't always be the best representation. It depends on the scale that it is on the game. So we are also testing if we truly believe in a concept and it's a high production let's say concept in our side we might test it directly also on the main campaigns. Okay lower quantity I would say and just to understand if this is something that can work or can't. The only trick trick side of it is that some channels may take a while to give a chance to new creatives. So uh so it takes time but if we did a high effort in the production of it we will still give it the chance you know. Yeah. Do you do you also force it in that case just to get impressions because well obviously like there's a high chance of all the kind of gaming studios have tons of creatives but then you have five winners that's it and then

it's pretty hard to kind of give the chance to the new new concepts and then if you have a big creative team they are going to be very not angry but sad that they're doing all the work but they're not getting enough feedback. Right. So I will say that because we are one unit, right, growth and creatives and we're pretty aligned. We are also um collaborating a lot on what do we need and for what channel do we need it. So we don't just produce and have no space to test it because that needs time sometimes and and there's nothing we can do because not every channel allows us to push impressions into a certain creative. Um, and it doesn't always make sense to make a dedicated campaign that will compete. So, it depends. But I would say that we are usually syncing with them like UA managers and creative producers to let them know like what channel needs the most focus that we haven't seen like some that is starting to get a fatigue let's say and we want

to to develop something for that specific channel. Um and and then the creative team will build something that is also more focused on where we need, right? Because also different UA channels have different sort of a creative strategy trees. Yeah. Okay. Well, I'm I'm I'm playing all the videos on the back and it's just so many of them and this is fogg was amazing. Yeah. Yeah. It's just uh it's perfect. So I mean you said like you are getting all inspiration from all the like competitors what's trending sensor type or whatever else and add quality as well. Okay, I I love it. I say that I can add a little bit. I don't know how much more h we're talking about it, but I would just add that we are even though we are a big team of creative in supersonic, we are also really like into AI and exploring all the different

um companies basically that are out there. There's quite a few like that are offering pretty interesting products and making it like really much easier or better for the creative team to be able to generate more with less effort you know and to like optimize around like winning creatives in a more efficient way. So yeah, so we are also like very much into that and trying to do as much as we can with the help of AI because we think that for the creatives world at least it's something that it's really there already today like that can be very much maximized. Yeah, we're also very into AI here. Yeah, we are also into AI. I saw you you're working with uh with pool day for sure because I saw like the face the the whole avatars. channel. I I work with them as well. And the question is like, do you see if you slap a UGC on top of the gameplay videos you have, do

you see better metrics or it's just like I mean it depends because what I've seen is just you slap a face or UGC on top of this kind of like let's call it hyper casual visual style, it drives higher CPIs. Mhm. Like Okay. So, I would say that uh pool day specifically, it's something that we started testing maybe like I would say like maybe two three months ago. like it's not a very long time, but it's definitely like UGC is something that we were doing before with like real production like actors that were like doing these sort of like videos for us and you know we had like collaborations around that but it took a lot of effort and the beauty let's say with pool day is that is just like you know like very very easy and you can make a lot of them and it looks real in a way like you

don't always see that there is like some delays between the person speaking and the voice over but um it's like working for supersonic at the early stages. I won't say that there is something huge that we saw from it but we do manage to generate scale with a normal performance. It's just we haven't yet seen like something that it's very sustainable yet with the creators we had. There's like a peak and then it will go down relatively fast. But I will say that in in the games that we are testing which some of them are also a hyper casual we haven't seen necessarily okay only high CPI so we did see some wins and scalable productions of a pool day that worked for us and also in screw master. Yeah just not sustainable yet. Yeah, I mean we use Yeah, I mean well you always well you always save

maybe then maybe that's the motivation like save the save the puppy you know I mean this is amazing uh I think the only I mean not the questions just comment which I had and was kind of part of that long question regarding the creative testing was like the level of spend. So I would say you're kind of optimizing for that like day 90 maybe even longer because your retention is really strong. So I would say what like day 90 day 120 maybe 180 at some point which then unlocks quite enormous spend. So I mean that's just that's just my comment and just way of thinking which then is uh an interesting uh exercise for you just how to add new channels into the UA channel mix on the monthly basis which can be quite hard. So how do you decide like what kind of channels do you do you add? Yeah. So it's all of them.

So no. So of course not like I wish it could be all of them like just to see what works best, right? But like there is cannibalization between channels, right? There's something that it's not like um easy to control and we see it a lot between incentive channels, let's say, and some channels on Android. But in the end of the day, what we're looking for when we are adding a new channel into a game is performance, right? So if it's generating a really good performance and it's scale like allowing us to scale like because if it's a good performance we'll keep opening budgets until we see you know that it starts hurting performance. So we would just keep on optimizing based on that and give the best channels according to the different games the chance to succeed. We don't have like we're not collaborating with every single channel out there in the industry but we are exploring a lot of new channels this years this year

specifically and testing incrementality in our site to understand better how each one works and the influx of new users that they are actually bringing instead of just cannibalization. Yeah. So we are testing that. We have a marketing at like a analy science team that is also helping us with all of that and hybrid games or high retention games. That's why I said this year we are focusing on new channels a lot more just because it's a different there's different types of channels that could work now for us that maybe on the hyper casual classic games didn't re weren't really relevant, right? But once you have a good retention game, there's other channels that become more relevant. So that's what we are trying to explore. and um find who are the best partners for us. Okay. Well, I think I ran out of questions. I mean, this is pretty this

is great episode. This is really good. Thank you so much for sharing all these really valuable insight. This was one of the better ones. This is really valuable. Thank you so much. Anything else you want to share with the with the audience? I think I'll say it also. I'm very open here, but that you're blurred. Yeah, Felix like that. It won't. Don't worry, including you. Yeah. I mean, as soon as we stop talking to the hedge with his estimate. Yeah. Yeah. Exactly. As soon as we start talking about UA like I'm just go I'm just go away. Pay out. Pay out. Well, no. Anything you want to share uh with the audience? I mean, obviously we we will uh add all your contacts into the said message. I'm sorry, but it's not. It's just cuz why you reading the same name for me, you know? Oh, yeah. Yeah. It's mate for for Felix as well. I'm not sure. Like he does this on purpose always. Yeah, it's on purpose. It's funny. I want to say like two small things. One, I want to encourage everyone to play the game and share their feedback. We are doing games for

users. So, we are looking forward to get as many comments and ways to improve the game because this is what we're doing. This is what we like to do. And second, I'm going back to the beginning of the conversation about WeChat. Every time Oh, here's the park. Yeah. Yeah. Just in the meantime, I wanted to tell you I'll unlock it. So, we just moved there with with the cap bar here in the bat. Yeah. So, uh every time that I'm traveling to to China and Vietnam and and exploring a platform like WeChat, it makes me be more humble. And I think that we all understand that there are many ways to to make games and not only one. And this is one more thing that I wanted to share about our philosophy like there are many ways. We just need to to to be humble and look for for new ways to to do this. That's very good. Tom, anything on on your end except like the the winking on the budget

estimates? Yeah. Yeah. If you want to share something with the audience share quite a lot. I mean yeah wouldn't expect that we get it's pretty good you know yeah as guy said we love what we're doing you know it's like really fun working in the industry there's always like new things new challenges AI is coming out so things are always like you got to keep yourself updated I think this channel that you guys have it's like a great place to like learn grow knowledge connect between people in the industry like we know here like a lot of people are like listening to you guys and we are all watching your uh deconstructing deconstruction of games and it's really valuable. So, thank you for that and yeah, we're happy to join. You know, it's like u it's a pleasure. We feel like stars now that we've been here. Definitely stars. You're definitely stars. That's for sure. Well, dropping knowledge is always the uh the

headline of of this podcast which is amazing. So, thank you very much for coming and also the listeners. Thank you very much also for listening. This was longer one but full of insights. Thank you. Join the Slack channel. if you have anything to ask Guy and Sam and see you next time. Thank you very much. Bye-bye. Cheers. Ciao. Thank you. Bye. [Music]