Driven by technological innovation and cultural integration, Chinese games have achieved remarkable success in overseas markets. According to the "2025 First Half China Gaming Industry Report" released by the China Audio-video and Digital Publishing Association, Chinese self-developed games generated $9.501 billion in overseas market revenue from January to June this year, representing an 11.07% year-over-year increase.

"Chinese game developers remain highly competitive in overseas expansion," said Li Zizhan, Senior Director of New Business for Greater China at Moloco, in a recent interview. From a trend perspective, overseas expansion strategies have undergone significant changes—moving away from global broad investment and extensive growth toward more segmented and refined approaches. Many Chinese developers are no longer focusing solely on T1 markets (mature markets like Europe and America), but are actively choosing to concentrate on T2 and T3 markets, with this strategy of deep cultivation in single markets becoming increasingly common.

Facing rapid changes in the global internet ecosystem, how does Moloco leverage AI-driven advertising technology to help Chinese developers build sustainable growth models? We spoke with Moloco executives to explore this question.

**Focusing on Machine Learning: Integrating Multi-Platform Traffic and Improving Campaign Efficiency**

"Moloco has been a machine learning company since its inception—AI is in our DNA," Li Zizhan noted. The company possesses three core AI capabilities: advanced AI infrastructure, proprietary deep neural network models, and advertiser-friendly software platforms.

In terms of AI infrastructure, Moloco pioneered the use of Google Cloud TPU chips for model training in late 2023, achieving an 8x improvement in training speed and a 2-4x reduction in costs. Li Zizhan explained that this not only enabled Moloco's cost reduction and efficiency improvement but directly translated into enhanced campaign efficiency and reduced costs for advertisers.

Du Que, Product Expert for Greater China at Moloco, further elaborated on the company's technological approach: "Our technology investment focuses on areas most relevant to advertising effectiveness. In advertising scenarios, prediction is more critical than generation. We need to accurately predict which users are the right users based on historical and market data, and determine the optimal timing and creative materials. Therefore, Moloco employs deep neural network architectures that perform more stably in prediction tasks, while continuously exploring applications of generative AI."

Additionally, Moloco launched several new internal initiatives this year, including establishing the Moloco Next team focused on innovative business development. Li Zizhan explained, "We have an internal concept called 'anywhere to anywhere,' meaning that in the future, regardless of which device users see an advertisement on, they can jump to another platform to complete conversion. For example, users might see an ad on mobile but complete the download on PC, or see an ad on PC but finally install the app on mobile."

This cross-device conversion involves many new technologies, including attribution methods, cross-device tracking, and effectiveness measurement, which the Moloco Next team continues to research and test.

Notably, in Connected TV (CTV), Moloco has achieved programmatic, machine learning model-integrated precise targeting and measurement. Du Que revealed that Moloco's CTV product is advancing rapidly and can now support CPI (Cost Per Install) optimization targets.

"This CPI can be the 'I' in mobile games, PC games, or console games, achieving cross-device, cross-scenario user behavior recognition and full-process identification and optimization from exposure to conversion. Moloco's CTV solutions now cover multiple vertical sectors including gaming and sports, with deep partnerships established with leading platforms like TVING. We will continue creating greater value for advertisers."

**Expanding Global Presence: Discovering Incremental Markets Outside "Walled Gardens"**

For Chinese developers preparing for overseas expansion, systematically tapping traffic outside "walled gardens" has become essential. Industry reports show that mobile users currently use approximately 26 apps monthly, spending an average of 13 minutes daily on each application.

Li Zizhan explained that the industry categorizes these applications into four types. The first category includes giant platforms like Alphabet and Meta—typical "walled gardens"—accounting for 32% of total user time. The second category comprises other "walled gardens"—self-attribution platforms ranking below Alphabet and Meta, such as Twitter, Netflix, and Xiaohongshu—representing approximately 18% of time share.

"At this point, usage time within 'walled gardens' accounts for roughly 50%, leaving the remaining 50% as traffic outside 'walled gardens.' Excluding apps where advertising cannot be placed, traffic truly belonging to 'outside walled gardens' where advertising is possible represents about 31%—a scale comparable to Alphabet and Meta."

Through efficient integration of this traffic segment, Moloco currently reaches 3 million apps, corresponding to 2 billion industry DAU. Li Zizhan disclosed that with AI capabilities and reach efficiency, Moloco currently manages $2 billion in annual advertising budget, ranking among the top non-"walled garden" platforms.

Moloco's services extend beyond gaming products to include many non-gaming applications, such as AI tools, AI social platforms, AI education, and e-commerce applications that have flourished in recent years. Geographically, Moloco maintains broad and balanced global traffic distribution without regional bias.

"Moloco determines traffic scheduling based on client budget needs, helping them identify markets with higher product responsiveness and activating corresponding traffic for real-time bidding in those regions. We also provide diversified, multi-layered support through our campaign teams and agencies," Li Zizhan told us.

Beyond serving developers, Moloco is exploring new partnership models through deep collaboration with OEM manufacturers like Xiaomi. "In the overseas Android ecosystem, Chinese phone manufacturers have gradually become important traffic entry points. We highly value this traffic entry point," Li Zizhan noted. The entire internet advertising industry is entering a highly fragmented stage, with traffic competition becoming more open and real-time. Mechanisms like real-time bidding give every advertising platform opportunities to acquire this traffic segment, prompting deeper collaboration with Xiaomi.

Moving forward, leveraging Moloco's AI algorithms and machine learning technology combined with Xiaomi's hardware terminal network and user ecosystem data, both parties will jointly drive comprehensive improvements in advertising targeting accuracy and monetization efficiency. Together, they aim to promote intelligent upgrades of advertising platforms, create efficient globalized marketing solutions for Chinese enterprises going overseas, and help developers build sustainable commercial growth models and value creation systems in fierce international market competition.