Google Gemini: Global Success and Struggles in the Korean Market – Analyzing the Lock-in Effect and Competitive Landscape in the AI Market

While reading an article from Money Today, I discovered a truly intriguing phenomenon. Google Gemini is experiencing tremendous growth worldwide, yet it is struggling uniquely in Korea. As of November 2025, this distinctive occurrence in the global AI market goes beyond mere regional preference differences and provides significant insights into the competitive landscape of the AI market and user behavior patterns.

First, let’s examine the astonishing contrast. According to Google CEO Sundar Pichai, Gemini’s global Monthly Active Users (MAU) reached 650 million in the third quarter of 2025. This is an increase of 200 million from the previous quarter’s 450 million, showing a quarterly growth rate of 44%. It ranks as the second-largest globally, following ChatGPT, which is estimated to have around 1 billion users. However, the story is entirely different in Korea.

The data from the Korean market is quite surprising. According to Mobile Index, Gemini’s MAU in Korea for October was 68,023, a decrease of about 10% from September’s 75,811. Even more surprising is that this marks a four-month consecutive decline. New installations peaked at 338,957 in June but plummeted by 41% to 199,131 in October. The average usage time per person is only 0.62 minutes, less than a minute.

In contrast, during the same period, ChatGPT saw an increase from 12.8 million to 13.04 million users in Korea, with average usage time rising to 109 minutes. These figures clearly show how focused Korean users are on ChatGPT. Even more interesting is that X’s GrokAI surged from 205,690 users in September to 484,288 in October, more than doubling. Even Anthropic’s Claude surpassed Gemini with an October MAU of 81,937.

Global Success Factors and the Uniqueness of the Korean Market

The weapons Google deployed for Gemini’s global success are truly impressive. The generative AI video model ‘Veo3’, launched in July 2025, is praised for significantly lowering the entry barriers for video production. In August, they released the AI model ‘Nano Banana’ for image generation and editing, which reportedly attracted 13 million new users worldwide with this feature alone. This accounts for a substantial portion of Gemini’s overall quarterly growth.

Google also made efforts in Korea. Since August 2025, they have been running a bold event offering AI Pro for free for one year to college students aged 18 and over, and they provided a free trial of the paid version ‘Gemini Advanced’ to Galaxy S25 buyers. They even integrated Gemini into the Galaxy S25 series through collaboration with Samsung Electronics. However, despite all these efforts, the user base continues to decline.

Industry experts point to the ‘lock-in effect’ as the key reason. The analysis is that Korean users are already accustomed to ChatGPT and find no reason to switch to another platform. Especially, ChatGPT secured a large number of paid users with its ‘Ghibli-style drawing’ feature at the beginning of the year, and as Korean users developed their own usage patterns, a stronger lock-in effect was created.

Unique usage cases in Korea, such as ‘Trying to Get Insulted by ChatGPT’ and ‘ChatGPT Fortune Telling’, illustrate this well. Users are not just using AI as a work tool but are also utilizing it in personal and creative ways, forming attachment and habits towards the platform. In such a scenario, no matter how good a competing product is, users are unlikely to switch easily.

Changes in the AI Market Competition and Strategic Implications

An industry insider’s comment was quite impressive: “It’s true that Gemini has the best performance, but the performance of generative AI models has become so standardized that there’s not much difference in convenience for everyday use, so there’s no compelling reason to switch.” They also added, “In the future, the success of a service will depend more on what kind of service experience it provides rather than the difference in model performance.”

This analysis seems to reflect the current maturity of the AI market well. Initially, pure performance differences were important, but now all major AI models are showing a certain level of performance. The performance gap among major models like GPT-4, Claude 3.5, and Gemini 1.5 Pro is not significant. In this situation, factors like user experience, ecosystem integration, and brand loyalty are becoming more important.

The phenomenon observed in the Korean market provides important clues for predicting the future of the global AI market. It shows that a platform that enters first can form user habits and create personalized usage patterns, establishing a strong entry barrier. This is a typical characteristic of platform businesses where network effects and switching costs are important.

From Google’s perspective, it’s certainly a disappointing situation. While they are closely trailing ChatGPT with 650 million users globally, they are falling behind in Korea, one of the key markets. Korea is a market with high AI technology acceptance and a developed mobile ecosystem, making it even more significant.

However, this situation doesn’t seem likely to last forever. Considering the pace of AI technology development, new features and services will continue to emerge, and in this process, user choices may change. Especially if Google leverages deep integration with the Android ecosystem, linkage with Google Workspace, and synergies with YouTube or Google Search, they could create opportunities for a turnaround.

The fact that Anthropic’s Claude has surpassed Gemini in Korea is also a noteworthy change. Claude’s October MAU exceeded Gemini’s 68,023, reaching 81,937, with a much higher average usage time of 67 minutes. This shows that the AI market is evolving into a complex market with various players competing, rather than a simple two-strong structure.

Ultimately, as of 2025, the AI market seems to be entering a stage where user experience and ecosystem integration capabilities are more important than technical performance. Google Gemini’s struggles in the Korean market are not just a regional phenomenon but a case that shows the structural challenges faced by latecomers in a mature AI market. It will be fascinating to see how this competition unfolds and how Korean users’ AI usage patterns change in the future.

This article was written after reading a Money Today article and adding personal opinions and analysis.

Disclaimer: This blog is not a news outlet, and the content is the author’s personal opinion. The responsibility for investment decisions lies with the investor, and no responsibility is taken for investment losses based on the content of this article.