We recommend starting with a bid that is similar to the CPI multiplied by maybe 1.2 or 1.1.5 a little bit higher. And after the first two weeks we would start having a clear view of of the performance. Hey everyone, this is Roman from Mintegral. You asked us to record a video about Mintegral iOS bidding. And I think I found a perfect person to talk about that. Hector, please tell us about yourself. Hi Roman, I'm Hector. I'm a EMEA GM for Mintegral and I'm responsible for monetization and user acquisition. So what is the ROAS bidding? So ROAS bidding is the strategy to
understand how much revenue is generated by each ad campaign. So Mintegral is releasing this new bidding optimization formula to help our advertisers acquire more users and increase their reach while reaching their their ROAS KPIs. Got it. Let's take a step back and just look in general what sort of campaigns do you have guys have just to see a full picture and then we'll dive deep into this like ROAS bidding. I know you've prepared some slides for me today, maybe you can start sharing. Sure. So Mintegral has different
bidding options available for our advertisers in our self-service platform. Um one of them, and one of the most used, is CPI model. CPI is a bidding approach um that is manual. Publishers pay per install. Um and we can optimize based on each of the different ad sources. We can change the bid manually and increase the number of users that install the the apps from our advertisers. We also have CPM model, where advertisers can buy per per mile, per 1,000 impressions. Um here also the optimization is manual. And we have our new bidding strategy, which is the ROAS targeting strategy. So, ROAS bidding is um a a very interesting
bidding option. It's new for us, and ROAS, as said, it is the metric used to track the gross revenue generated for every dollar spent on advertising. So, in essence, it measures how effective the the advertising campaign is. Um it compares every single dollar invested and how much revenue got out of it. Uh and this is our our newest bidding mechanism. One of the main advantages of this is that is driven by machine learning. So, it is very good for us and for advertisers because on one side, ROAS bidding analyzes and predicts the users' LTV in real time. Um and this is helpful because the algorithm is able to adjust the bids according to the target ROAS objective.
Um also this is very good for for the advertisers because it obviously reduces a lot the manual labor as everything here is uh done automatically through through machine learning. And it also opens the possibility to attract uh different high-quality users that might be missed by targeting only specific channels um or sub-sources. For example, some of the of the gaming apps might be targeting competitors um but with ROAS bidding, we might find um the same user or a profitable user uh in a different uh category. For example, in a utility app, etc. At the end of the day, ROAS bidding gives us the flexibility to bid higher um for for specific users that the algorithm predicts um it's
going to be more profitable for for the advertisers. This ends up um with the result of a much higher reach for for our advertisers. So, for example, with with um some of our our our advertisers, uh we have seen not only that ROAS bidding has increased um the ROAS um return uh the return of each advertising campaign as as we can see here. For example, with Homa, uh when compared when we compare the CPI campaigns towards the ROAS campaigns, we see first of all that uh we have higher um ROAS. And uh much more we do have a much higher uh spend from from Huuuge Games in
acquiring more users. So, this is just um a testament on how uh ROAS can really increase the scale um of our advertisers' campaigns. So, those are pretty big guys. Huuuge is legendary. Uh is it also available for indie developers? Sure. So, ROAS is available for all um advertisers looking to buy traffic through Mintegral's self-service platform. Uh in general, we have seen a better uh better results with medium to higher um budgets uh advertisers. Uh but it's available for everyone. And another question. So, it's a ROAS. I know that there are like different types of ROAS. So, there be might be in-app purchases, might be ad revenue, and
there might be like hybrid of both ones. Uh so, this one works with all those or only uh with a particular one? So, our ROAS campaigns um work um with with with with all of them. So, it works with in-app advertising. Um it also works with uh in-app purchases. So, it is uh suitable for any app that has um a clear uh relation between the the ad spend and a clear uh way of measuring how much uh revenue it's it's coming from this specific ad campaign. For example, in-app advertising, in-app purchases, subscriptions, etc. So, it is it is suitable for all these kind of advertisers.
Um also I wanted to to mention um when you mentioned about the bigger spenders, um wanted to mention the recommended budget we have for for running ROAS campaigns. So, we actually recommend to start with around $500 uh per day uh minimum uh per per country. Mhm. Got it. What about bids? And uh how do I bid? Do I bid if I'm ad revenue if I'm doing ad revenue, I think it's crucial for me to bid on the day zero. Like, do I bid on the day zero here? What are the other other options? Exactly. So, so we would normally start with the the first the we have 3 days of cold start. How it works is that we do have uh first the postback data um say signals. So, first we need uh 2 weeks
of postbacks from all channels. Once we do have these 2 weeks of post back channels, we first set the goal the ROAS target. Uh and this ROAS target should be based in the lifetime value of the user. If the advertiser is running campaigns with other ad networks uh other ROAS campaigns, we can set the same or similar target. And uh see after few weeks how is the performance. Uh once this is defined, um we start with a we recommend starting with a bid um that is similar to the CPI multiplied by maybe 1.2 or 1.1.5 a little bit higher. Um and after the first 2 weeks uh we would start having um a clear view of of the performance. Once we have
that, we would adapt um according to to the results, make a recommendations. Um something very important is that we recommend the use of different ad creatives. Normally, we recommend uh advertisers to upload at least four sets of creatives and at least one playable ad. And at the same time, once our advertisers are running ROAS campaigns, we also highly recommend to run CPI campaigns at the same time. This would allow us to increase the reach. Uh ROAS would be again um targeting the users um that the algorithm can predict uh a good value. But CPI would also be able to reach different um ad sources
uh manually and we would be able to also manually optimize and um feed the algorithm with whatever we want wha- whatever users we want to target. Got it. I have a couple of questions to this slide. Um initial data sharing. So, you need to enable this somehow. I know our customers are already doing it. Uh so, definitely you can do it via MMP. Um I'm talking about like sending the uh ad revenue or EIP data to Integral. Uh are there any other options to do this? Yes, so we can get the postbacks in three different ways. We can get it as you said from the MMP. Uh but we can also get it directly from the mediation platform if this is in-app advertising ROAS. And we can also get it directly from our
SDK. Got it. So there are three ways of of giving us this data. And they don't need to use your mediation or administration any sort of No, so they don't need to to use our monetization. Yeah, yeah. Got it. Um how often should they update the creatives? Is it like every day? No, no. So we would recommend actually to update the creatives every two to three weeks adding some new creatives. We have seen that playable ads are quite important, so this is also one strong recommendation for all our advertisers. Got it. And for the CPI campaigns, are the this like CPIs and ROAS campaigns are connected in any way?
Uh they are not connected, but we would and this is why we would recommend our advertisers to run both campaigns at the same time, CPI and ROAS. With CPI, we can choose which specific apps to target. We we can think of targeting for example apps from competitors or where we believe that we will have similar kind of users and ROAS is fully automated. So the algorithm will find the users that will be valuable across different apps and different categories. So it not necessarily needs to be the same category. For example, doesn't need to be a game. It could be an utility app or any other kind of app and and this is one of the main features of ROAS that we would expand a lot the the the targeting
um uh the reach and always with the goal of of reaching the the the target goal. And once we reach the target goal, um we can increase the spending because we can obviously increase by a lot the the the people and the users we can we can reach that's are valuable for this specific advertiser. Got it. Makes sense. I think we can go to the next slide. Yeah, so it's about setting expectations. So sometimes it is very important actually to to speak directly with the account manager and set good expectations to be able to set a good ROAS goal.
Um because at the end of the day um this is what will be key to to find a good balance between reaching the the goal and having a lot more reach of users. So in the early phases of of of of ROAS, as as mentioned, we start focusing in the data collection, then optimizing it with our machine learning. Um, then the advertisers, yeah, will review constantly what's the their their ROAS target. Um, and then um, mitigating mid campaigns decreases, so it is also possible that at some point uh, the ROAS campaign might not be performing as expected, and when this happens, uh, what we do is that we would mainly
evaluate what's the app performance across all the different channels. Uh, we would check CTR, IVR, and um, as recommended, we would check also with the CPI campaigns, and we would um, use this data to optimize CPI campaigns. When you say we, I guess that is like uh, account management team, right? Correct. So, although so, Mintergral is quite a big company, we have thousands of advertisers, and we cannot assign a specific account manager to all our accounts. So, we have two methods of working depending on on the on the budgets and on the size of of our advertisers. We can give them either an assigned account manager, or we also have a general support email where all our account managers um, help answering all the questions. So,
Mintergral is accessible for all our advertisers, and we would be happy to answer the question from any advertiser regardless their size. All right, then. So, basically in both cases, when they launch something and they don't see a result or they launch something, they got some results, but then they saw a decrease, they can always reach out to your team. They can always reach out to our team and they will always get a a response. Got it. Super. All righty, let's see what else do you have for us today? Yeah, so finally is uh what's coming. Um so, right now uh D0 ROAS is fully ready. And now we are already working with D3 ROAS with uh some advertisers. This is something that it's uh being uh currently tested
uh in in-app advertising. And on the other side, we're also uh testing uh in-app advertising as well as in-app purchase ROAS solutions. So, here uh obviously, the main goal is to um in have our ROAS capabilities expanded. Uh although right now, ROAS D0 is fully available for all our partners. Um ROAS D3 is for for for selected partners as well as in-app advertising plus in-app purchase ROAS. Yeah, this is nice, especially with the times of hybrid. Like I I I think this is uh becoming a must. Exactly. So, if any advertiser would like to uh test uh our uh ROAS uh in-app advertising and in-app purchase uh solutions,
um they should just uh reach out to us. Uh what is the best uh way to reach out to you guys? So, they can always go to our website uh and reach out through our website. They can sign up for free with a new account and they can also contact us either through the website form or for they can contact me directly to my email which is [email protected]. All righty, thanks a lot Hector. I'll just add that please leave your questions in the comment section on YouTube. We'll try to answer them too. I really like this approach because like others will see what are you asking and probably they will also was wondering about the same thing.
Yeah, subscribe to our channel. Check our new tension website. Check Mintegral website about their features. And I wish you all a very good day. Thank you. Thank you.
Auto-generated captions from YouTube — names and numbers may be misheard.
Distilling the key insights…
Tenjin · 2026
Two & a Half Gamers · 2026
Comments