36Kr Exclusive Interview | Tu Zhengting, Senior Vice President of Appier: The costs of overseas marketing and user acquisition are rising steadily, and game companies are paying more and more attention to ROI
In August 2025, among the top 100 global mobile game publishers in terms of revenue released by Sensor Tower, there were 32 Chinese companies. The total revenue of these 32 Chinese companies reached $2.04 billion, accounting for 35.1% of the total revenue of the world's top 100 mobile game publishers. That is to say, Chinese companies have already occupied one - third of the seats in the world's leading game market.
At present, when the domestic game market is trapped in homogeneous competition and the quagmire of buying traffic, more and more Chinese game companies are making a name for themselves overseas. In the past decade, the overseas sales revenue of Chinese self - developed games has rapidly increased from $3.1 billion to $18.6 billion, and the proportion of overseas revenue in the total revenue has also expanded to over 30%.
Going global is an opportunity for established game giants as well as innovative game studios. However, for Chinese game companies lacking overseas experience, expanding into overseas markets is almost starting from scratch. How to communicate with players who speak different languages, have different cultural backgrounds, and use different social media? Under these difficulties, opportunities for multinational advertising companies have emerged.
Appier is an AI - native SaaS company targeting the global market. It mainly provides AI - driven advertising and marketing automation solutions to help Chinese game companies expand into overseas markets. Currently, it has 17 offices in the Asia - Pacific, Europe, and the Americas. Appier can help Chinese game companies through AI prediction and the technical support of automatically generating advertising materials, enabling game companies to have a better chance of targeting high - value players, accelerating the acquisition of new users and the recall of existing users.
Photo provided by Appier
After helping many Chinese game companies with their overseas marketing, Tu Zhengting, the senior vice - president of Appier's global business, has gained a lot of on - the - ground understanding of the difficulties and opportunities of going global. Tu Zhengting was interviewed by 36Kr. The following is the transcript of the interview:
Tu Zhengting, senior vice - president of Appier's global business
36Kr: What services can Appier provide for Chinese game companies hoping to go global?
Tu Zhengting: The services provided by Appier are divided into two major categories: advertising and marketing automation services.
The advertising services include the acquisition of new users, the recall of old users, and keyword advertising. In addition to advertising placement, we also provide material production. We mainly produce two types of materials: advertising videos and playable materials. Advertising videos are suitable for medium - to - heavy games, while playable materials are suitable for casual and puzzle games.
The marketing automation services we provide are mainly personalized promotions through channels such as EMAIL, SMS, and messenger.
Appier has provided refined game marketing services for many Chinese game companies, including how to attract paying players, consolidate ROAS performance, how to re - reach lost users through remarketing, or how to combine the emerging automatic generation function of playable ads to shorten the production time from weeks to one or two hours. AI can also predict the creative visuals preferred by overseas users, quickly generate materials that meet the needs of local users for testing, and iterate to a better - received version to quickly respond to market changes.
Photo provided by Appier
36Kr: Which countries and regions does Appier's current business cover?
Tu Zhengting: We cover the whole world and have offices in 17 regions. Our main coverage areas are the United States, Japan, and South Korea; the second - strongest regions are Europe, France, Germany, Hong Kong, Macao, and Taiwan; the third are Vietnam, Thailand, and Southeast Asia.
36Kr: We originally thought that the Confucian cultural circle seemed to be the "starter village" for Chinese game companies going global, but in recent years, we have also seen many cases where companies go directly to Europe and the United States as their first stop. How do you interpret the complexity of the "overseas destinations" for Chinese games?
Tu Zhengting: When Chinese games go global, it ultimately depends on the market. If they can challenge the market with the largest share, everyone will definitely do so. Currently, the United States ranks first in global game app store revenue, followed by China and then Japan. So, if a game company has the ability to go global to the United States, it will definitely choose the United States as the first choice.
Japan has its unique culture. Judging from the game rankings, the most popular game attributes in the United States and Japan are very different. The United States prefers medium - to - light games, while Japan prefers medium - to - heavy games. Therefore, it depends on the development direction of each game manufacturer to decide the overseas destination.
In recent years, more and more Chinese game companies have directly challenged the US market because there have been successful cases. For example, some games ranked among the top in the US rankings, including the match - three game of Lexin Shengwen and the strategy game of XD Inc. These games have performed very well in the US rankings, indicating that Chinese game developers have the ability to directly challenge the US market and have achieved a considerable share of success.
The Japanese market is more suitable for anime - style, medium - to - heavy games and games with well - known IPs. Locally, as long as the game quality is good enough, it can get a good share.
Each country and region has very different preferences for game types. In terms of advertising materials, the aesthetics of each region are also different. For example, the United States is suitable for a large number of playable materials, while Japan is suitable for a large number of video materials. At present, strategy games are more suitable for simultaneous promotion in multiple markets, such as "Endless Winter" by XD Inc.
36Kr: According to your observation, what is the biggest change that Chinese game companies face when going global?
Tu Zhengting: In the past three years, the biggest change I have seen is the increasing emphasis on game distribution costs. Three years ago, game companies might directly invest a large amount of money in Japan, South Korea, Hong Kong, Macao, and Taiwan, investing several million US dollars in a month and then waiting for the return. But now, everyone is becoming more conservative. They will first conduct small - scale tests in regions such as Taiwan, Australia, New Zealand, and Canada to see if the user payment behavior and retention rate meet expectations.
For example, if a game company wants to promote an anime - style game and knows that the target market is Japan, it will first test the waters in Taiwan because the preferences of Taiwanese players for anime - style games are relatively similar to the behavior patterns of Japanese users; if it wants to promote a medium - to - light casual game and knows that the target market is the United States, it will first test the waters in English - speaking countries such as Australia, New Zealand, Canada, and even the Philippines.
Chinese game companies conduct small - scale advertising in regions with relatively lower traffic - buying costs to see how users' payment behavior is, whether it can meet expectations, and whether to adjust the marketing strategy, etc. In this process, the company will continuously optimize the plan based on market feedback until it confirms that it is ready to enter the Japanese and US markets.
That is to say, there have been great changes in both the pre - distribution optimization and the post - distribution media placement strategies. Now, we rarely see the way of large - scale advertising investment. Everyone will first use 1/10 of the budget for testing, look at the ROI (Return on Investment), and then select the media channels with good results and increase the investment. In the cases I have experienced, there was a company that postponed the official launch of the game for more than six months to ensure that it could recover the cost, although this is a very rare case.
36Kr: What is the changing trend of overseas marketing costs and traffic - buying costs in recent years? If Chinese game companies decide to go global, do they first need to prepare a huge investment in this regard?
Tu Zhengting: On average, the user acquisition cost increases by 3% - 5% every year because the selling prices of all media, including platforms such as Google and Meta, are rising. Especially two and a half years ago, most media increased the unit price of exposure, and the CPA (Cost per Acquisition) also increased.
Another significant change is that the original rule of the bidding - based advertising was that if you won the first place in the bid, you only needed to pay the price quoted by the second - place bidder to the platform. But two and a half years ago, the platforms generally changed the rule - the first - place bidder needs to pay the price it quoted. So, the cost is steadily increasing, and everyone is paying more and more attention to cost recovery.
There are three ways to recover costs: the first is almost completely relying on in - app advertising (IAA) for monetization, the second is almost completely relying on in - app purchases (IAP) for monetization, and the third is a combination of the two, which is hybrid monetization.
Games that rely on advertising for monetization are usually relatively lightweight, and the game development team may be small. As long as a hit game is developed, the monetization speed will be very fast, and the cost of media placement can be fully recovered in 7 days. However, the drawback is that the retention rate of such games is usually not good. After 7 days, 93% - 95% of users may be lost.
Games that rely on in - app purchases for monetization usually have high development costs and high game quality. The monetization cycle of such games will be relatively long, possibly taking 6 - 18 months. This kind of operation is also a bit more difficult.
As for the third "hybrid monetization", for example, when in - app purchases account for 30% and advertising accounts for 70%, it usually means that the game is in the transition stage from IAA to mainly relying on IAP for monetization; when in - app purchases account for 70% and advertising accounts for 30%, it means that the game has established a stable in - app purchase revenue. But in order to also earn revenue from "free players", incentive advertising within the game is usually provided - players can get in - game rewards such as gold coins after watching videos or playable ads.
According to our current observation, although there are still many large - scale manufacturers insisting on making high - quality games, in the past two years, there have been more and more small studios making lightweight mini - games. Because the traffic - buying cost has generally increased and the cost - recovery cycle is too long, business is averse to risks. Small studios will tend to first monetize through in - app advertising games and then gradually transform into in - app purchase - based games after accumulating funds.
36Kr: Will Appier help overseas game companies with localization work?
Tu Zhengting: We will help our clients with localization - related work, but most of the game companies that hope we provide localization services are from Japan and South Korea.
Game companies from Japan and South Korea rely more on Appier's insights in game globalization; while Chinese mainland game companies basically set up offices directly in important overseas markets and recruit local people for localization. Because Chinese game companies basically establish their own teams directly in important overseas markets. They hope to obtain first - hand information rather than second - hand information.
Except for the largest game giants, most Japanese and South Korean game companies, if they want to go global to another regional market, will use local agents to help them with localization. However, more Chinese game companies going global will directly hire local people to operate in the target market or include them in the product team to directly obtain first - hand market insights.
What Chinese mainland game companies rely more on Appier for is to provide refined game marketing services, including how to attract paying players, consolidate ROAS performance, how to re - reach lost users through remarketing, or how to combine the emerging automatic generation function of playable ads to shorten the production time from weeks to one or two hours, etc.
36Kr: Compared with traditional advertising and marketing methods, what improvements does Appier's AI advertising technology solution bring? In which national markets has this solution been put into use?
Tu Zhengting: Currently, our AI advertising technology solution has been put into use globally.
The AI tools mainly help our clients solve three problems: Who should the advertising content reach? In what scenario should it reach the target? At what time should it reach the target?
Taking game company clients as an example, first, we need to find high - quality users who are willing to pay in the long term. Second, we need to push the advertisement to the target in a suitable scenario, rather than forcibly showing the advertisement to them when they are working or concentrating on something, because users are likely to directly ignore the advertisement, wasting the client's advertising cost. Finally, we need to use AI to find out the time when users are most likely to be impressed by the advertisement. For example, AI may suggest that we "expose the advertisement 2 times in the morning, 3 times in the evening, and 10 times at night", then we will concentrate on showing the advertisement at these appropriate time points to achieve the best advertising effect.
Through AI, we can understand what each user likes, how much they like it, where they can be exposed to the advertisement, and at what time they are more likely to accept the content recommended by the advertisement. After all these problems are solved, we finally decide what materials to use for advertising. Only in this way can we achieve scientific and standardized advertising placement.
Taking our cooperation with game publishers as an example, we have user models for heavy - duty games and medium - to - light games respectively. Based on AI's understanding of heavy - duty and medium - to - light game users, we find suitable users at the right time in appropriate in - app advertising positions.
Of course, the general model may not be 100% suitable for a specific game. In the process of practice, AI will learn about the special features of the game users' behavior and adjust the model based on new data to finally fully match the game's advertising requirements and optimize the results. For example, we assisted the well - known shooting game "Knives Out" under NetEase Games. Through AI, we identified high - value users at risk of churn, optimized the bidding strategy by combining contextual signals and high - quality inventory, expanded the remarketing effectiveness, and achieved a double - increase in ROAS and login rate.
36Kr: The last question is, do you think the macro - proposition of "de - globalization" has an impact on the micro - operations of Chinese games going global?
Tu Zhengting: I think virtual products are less likely to be directly and significantly affected like physical products. For game users, they are more concerned about whether the game quality is high enough and less concerned about which country the game is made by. The overseas success of "Black Myth: Wukong" has confirmed this.