“The Future of Mobile User Acquisition Is… Brilliant”, Dimitris Rostkov on Alternative App Stores and More

1) Tell us about your journey. You started in banking and payments, worked with micro-credit companies, and even managed operations for a 150-person team. What drove you to make that leap from traditional banking into the wild world of mobile performance marketing?

Thank you very much, it's a pleasure to be here. It’s an interesting experience in my life. Twelve years ago, I started my career in the banking sector and the payment industry. In my personal opinion, in our business, there are three main things: payments, licensing, and traffic. These are the three main things on which you decide which part you want to work on.

I decided to start my career in the payment industry. I had a lot of companies with licenses in many countries—European countries, Latin countries, Africa, and South Asia. I started with that, but about three years ago, I took a look at everything I had in my life and decided to start traffic. I already knew many things in the payment industry, and the next step was the traffic. I just started this business with my partner, with whom we are currently cooperating and working worldwide.

2) You've become known for mastering traffic from alternative app stores and OEM partnerships like Xiaomi, Huawei, and Techno. Most agencies focus only on Google Play and the App Store. What's the real opportunity in alternative stores that others are missing? Is this just about accessing untapped inventory, or is there something fundamentally different about how users behave on these platforms?

Regarding the alternative stores, it's not only Xiaomi and Huawei. Of course, there are a few more stores and smartphone developers like Realme, OPPO, and Techno. Xiaomi and Huawei are the biggest among them.

Regarding the statistics, for example, in 2023, 140 million devices were shipped to users from Xiaomi, and 35 million from Huawei. In 2024, around 170 million devices were shipped by Xiaomi, so around 15 to 17 percent growth. Huawei shipped around 45 million devices, so these stores are definitely growing.

Three years ago, when we started our business, nobody knew anything about alternative stores. Everyone looked at us like, "What traffic do you provide? Alternative stores? What is it?" But right now, this is a huge competitor to iOS and Android. 

They have more available players. The traffic from these alternative stores is traffic from the devices that Xiaomi and Huawei develop.

In iOS, it's the same, but there is only Apple Search Ads, and there are a few problems with that. For example, it's premium traffic with high quality, everything is fine, but it's a little bit restrictive. In my personal opinion, it's a little bit expensive. They have limited targeting and some attribution issues.

If we're talking about OEM traffic, it is cheaper traffic. It depends on the geo and the vertical, but generally, traffic in Xiaomi and Huawei is cheaper. So, these platforms are more scalable and flexible. We have faster approvals from the representative team. We have a very good relationship between our company and Xiaomi.

We know the representatives; we had a lot of personal meetings with them in China and Singapore. It's easier to communicate when you know the person, and they help you approve the moderation process faster or explain how to launch the application if the brand doesn't have one in the Xiaomi store.

In my personal opinion, there are more formats: pre-installs, banners, video, and brand formats. You can do anything. There are a lot of special formats if you have the core agency status, as we do.

Xiaomi provides more possibilities to run traffic on events, for example, the Championship League or the World Cup. They have a number of different formats that iOS cannot provide, and that is a deal breaker for us.

3) You also have exclusive placement formats for Xiaomi and Huawei. That's not something most agencies can claim. For agencies trying to build similar partnerships, what's the real value proposition? Is it just about getting better rates, or are there strategic advantages in terms of targeting and user experience?

It's not only about the costs. It's about the advertising formats. For example, when you open a Xiaomi device, and you are interested in the World Cup, you see notifications, for instance, if it's Liverpool on the left side, and on the right side, it's Manchester, and you see the current match score.

If you click on this banner, Xiaomi lets you install the application of the betting party—of course, only licensed betting. Such formats are something new in advertising, and you do not get to do this on iOS. So, you are not just building the relationship by email, as in iOS. Xiaomi creates a different type of cooperation with more possibilities and advantages for your agency.

4) So, now let's talk about some core verticals: betting, fintech, and e-commerce. These are all high LTV and also high complexity markets with strict regulations. How do you approach user acquisition differently for betting versus e-commerce apps? What's your framework for determining which traffic sources work best for each?

It's an interesting question because these are two absolutely different types of advertisement. We have very good communication with all the brands we work with in betting and e-commerce.

If we talk about the betting sector, they of course function with set compliance. They must comply with the regulations of the country in which they are licensed. When we build new creatives, we go to them and ask if everything is okay or if we must add or remove something.

So, we only run the advertisement once they approve of it. We cannot start advertising betting or regulated fintech companies without approval. Because ofcourse if we launch something, there will be questions from the regulator to the company, and then from the company to us. I have a few cases in mind where a company ran ads and the regulator declined the license. So, it's a very important thing for us.

Regarding the differences between the creatives: In betting, it's trigger-quick, emotional creatives. You put the "last minute," "don't miss out," "let's make a bet." It involves short videos, like 10 or 20 seconds.

In e-commerce, it's absolutely different because you must show the value, product appeal, what is the best-selling item, and how you can buy in two clicks. On platforms like TikTok, you use unboxing videos.

These are different types of products. We have experience in all of these, and when we meet a new advertiser, we thoroughly understand which creatives and which ways will provide them with more reachable traffic.

5) Can we dig a little deeper into regulatory navigation? Betting advertisers face restrictions that make creative execution feel like navigating a maze. Yet you have managed over 100,000 first-time deposits in betting. How do you balance compliance with performance? Do you have different creative strategies for different regulatory environments?

Of course. On each geo and with each advertiser, we are working absolutely closely. Sometimes it's not too easy to manage all of this because you know which creatives have better performance, but when you send this creative to the advertiser to get approval, they tell you, "No, we can't do this, we need only this, change the text, change the video."

It is difficult sometimes, but after our years of experience, we now know the regulatory requirements. Not for all geos, of course, because sometimes we touch new geos where we don't understand anything.

We have a legal team and a compliance team. Also, I am a certified AML (Anti-Money Laundering) officer. So, this helps us to better understand and more quickly learn how to work in a specific geo and with a specific brand.

We have different creative strategies for each vertical. For betting, fintech, and e-commerce, we have different templates. We take the brand book, see the spy services to see what our competitors did, try to build something new, go to the compliance team for approval, and then start advertising.

6) Now lets talk about cross-platform budget allocation. Your work with premium in-app networks like Unity and IronSource, social platforms like TikTok and Facebook, plus your OEM partnerships. How do you decide budget allocation across these vastly different traffic sources? Is there a typical split you recommend?

These are three different types of traffic sources. Unity, IronSource, Liftoff, and other premium in-app networks are not like an OEM. They put the advertisement into the applications, and they retrieve traffic from that. OEM retrieves traffic only through the smartphones, using formats like banners, videos, pre-installs, and pushes. The third type is social traffic, like TikTok and Facebook.

When we start understanding what type of traffic sources we must provide on the first step to our new advertiser, we have a meeting with our buying team, designers, and account team. We see how a similar brand or vertical performed with each type of source. After that, we decide where to start to show the advertiser the best kind of traffic.

Usually, we start with the most effective kind of traffic. We scale it as we have experience with other brands. If you understand how OEM, social, and premium in-app work, you know the formats and what bid you must put on the first step.

For example, in Liftoff, you can't launch the campaign and expect to see a lower CPA on the second day. You must wait three or four weeks. That is why we do not show Liftoff as a first step to the advertiser.

The main thing is the creatives. We have a platinum award this year for our work. So, this is what we are basically doing: we take the most effective source of traffic, put brilliant creatives from our side, and just start.

7) AI is revolutionizing campaign management. How are you at Evamobi using AI in your campaigns? And looking ahead, where do you see the biggest opportunities for mobile user acquisition?

Regarding AI, yes, we are using it. We analyze the traffic, compare Excel files and statistics from a few brands, and do analytics. But the main AI tool we are using is not in the traffic—it is in application development.

On our side, we develop mobile applications. These are WebView applications or big applications that we release to iOS and the Xiaomi store to provide traffic to the advertiser. In application development, we use AI as an instrument to create the design. 

85% of the design that we release to iOS or Xiaomi is created by AI. 50% of the code that our frontend and backend developers write is created by AI as well.

This helps us reduce our expenses on hiring staff. We have maybe only 12 or 13 people right now creating applications, and they release applications every week. We don't need 25 or 50 people because these people are enough to just write a prompt, correct the result AI gives, and start using it.