Can AI + going global give rise to global enterprises? | Insights into Going Global

#Going Global #AI #Qiantang #Globalization #Localization #Made in China

The AI wave is sweeping across the globe, and "Failure to embrace AI means being left behind" has become a consensus among enterprises. China's progress in AI technology has attracted global attention. From hardware to services, and from R & D to operations, the opportunity for Chinese enterprises lies in transforming AI into a tool for opening up incremental markets.

However, going global is by no means a simple "replication of the domestic path". To gain a foothold in the global market, it is not only about providing competitive and high - quality products. It also requires an understanding of cultural differences and the real needs of users, and playing the role of an "innovator" rather than a "copier" in the global market.

Adapting to the implementation scenarios, understanding user needs, and building a brand ecosystem are all tough battles. Recently, at the 2025 "From 'Ingenuity' to the 'World'" Going Global Conference jointly hosted by 36Kr and Qiantang Construction and Investment Group, representatives from four enterprises, namely Duowei Capital, Petkit, Yizhi Intelligence, and Tuya Smart, conducted in - depth discussions on topics such as building technological barriers and market adaptation mechanisms.

01 Beyond Product Innovation, Structural Capabilities Matter More

How can technology enterprises build global barriers? The guests generally believe that a single technological advantage is no longer sufficient to support long - term competitiveness, and the construction of structural capabilities is the key.

Guo Weike, the founder of Petkit, has a deep understanding of this. He pointed out that the pet industry has a limited scale, and commercializing top - notch AI technology faces dual challenges: On the one hand, consumers' expectations for AI have exceeded the current capabilities of enterprises. Failure to embrace AI technology will lead to being eliminated by the market. On the other hand, the dilemma of "increasing volume but not price" brought about by AI is significant. The product operation cost has increased, but the product price remains the same, and the product gross profit has dropped from 45% to 25%.

"Small industries cannot 'expand at all costs' like new energy enterprises. They must find a clear profit path between technological investment and commercialization." In response, Petkit's solution is "value transfer" - shifting from hardware sales to software services. In 2023, the revenue from pure pet software services reached 30 million, with a gross profit margin as high as 90%.

From the perspective of capital, Cao Fangning, the founding partner of Duowei Capital, divides AI enterprises going global into three categories, each facing different challenges:

AI Infra enterprises need to balance geopolitics and data security and formulate long - term regional strategies;

AI Agent enterprises need to clarify their global user positioning and make a choice between short - term interests and long - term ecosystems;

AI + smart hardware enterprises, although having supply - chain advantages, need to be vigilant against "replicating the domestic path" and must deeply adapt to overseas users' habits.

For example, pool robots, a niche product, are a "pseudo - demand" in China, but their demand has exploded in the European and American markets. "The success of pool robots lies in the fact that while the supply side cultivates users' habits, the cost and pricing reach a critical point." Cao Fangning summed up.

As a platform - type enterprise connecting technology and the market, Ren Hangbo, the market director and the person in charge of AI innovative hardware at Tuya Smart, said that more than 80% of Tuya Smart's 'Powered by Tuya' devices are targeted at overseas markets, and its core capabilities are reflected in three aspects:

Compliance First, covering regulations such as the EU's GDPR and the US's CCPA to reduce customers' compliance costs;

Interconnectivity, enabling cross - brand device interaction through a unified platform to improve the user experience;

Agile Development, with standardized tools supporting the development of AI hardware within one month to help customers respond quickly to the market.

02 Localization in Globalization: Adaptability Determines the Implementation Effect

Although the global market generally has an open attitude towards AI, whether it can be adapted to the market determines the implementation effect. Zeng Yiqian, the president of the consumer business department of Yizhi Intelligence, divides the global market into two categories and shares Yizhi Intelligence's differentiated implementation logic:

In developed markets such as Europe, America, Japan, and South Korea, the labor cost for enterprises is high, and the habit of paying for SaaS is well - established. The efficiency - improvement value of AI is prominent. "The AI outbound calling service we provided for L'Oréal and Estée Lauder can reach hundreds of millions of users, and it outperforms manual work in terms of call duration and follower - adding rate." Zeng Yiqian gave an example.

In emerging markets such as Southeast Asia, the low labor cost limits the initial advantages of AI. Zeng Yiqian emphasized: "A local team is the key to breaking the deadlock." Just as overseas enterprises need to form local teams when entering the Chinese market, Chinese AI enterprises also need to deploy local teams in the target markets to adapt to language, culture, and compliance requirements, and at the same time, reduce costs.

Ren Hangbo reminded enterprises to avoid common mistakes when going global: One is "over - self - development". A smart heater enterprise spent 3 - 4 months without solving the hardware - software adaptation problem with its self - built team. Small and medium - sized customers should make good use of platform capabilities to avoid reinventing the wheel. The other is "redundant functions". A desktop robot team delayed the product launch due to polishing 5% of niche functions and missed the market window. He believes that AI hardware should "Take small steps and move fast", first meet the core needs, and then iterate and upgrade.

03 Not Just Model Capabilities, Emphasize Both Technological R & D and Operational Efficiency Improvement

For most enterprises, the value of AI does not lie in flashy technological demonstrations, but in tangible efficiency improvement and cost optimization.

Guo Weike shared that after Petkit introduced a large - scale AI model into its customer service system, it entered data such as product problems and user concerns into the system to assist front - line customer service in responding quickly and empower the dealer team. "The core value of AI at the operational end is not to pursue 'disruption', but 'efficiency improvement'." He believes that AI is positioned as a "tool to serve front - line customer service" rather than a replacement. This model can not only reduce the pressure on customer service, improve response efficiency, but also make up for the training cost and emotional fluctuations of pure manual work.

Zeng Yiqian further pointed out that the root cause of the "ineffective cases" in the actual application of AI technology often lies in "being divorced from business scenarios" - only emphasizing model capabilities but being unable to quantify the effects. He believes that valuable AI applications need to meet three criteria:

Controllable and Monitorable, for example, AI outbound calling focuses on clear scenarios such as order notifications to avoid the "hallucinations of large - scale models".

Data Quantifiable, such as increasing the call duration from 30 seconds to 45 seconds and increasing the follower - adding rate in the private domain by 20% and other specific indicators.

Scenario - based Implementation, digital human live - streaming reduces the cost per session from 2 - 5 million yuan to 1 million yuan and covers tens of millions of users at the same time.

04 AI +, How to Create Global Enterprises

Regarding the proposition of "How can AI + give birth to global enterprises", the guests generally believe that differentiation and creating incremental value are the core.

Cao Fangning proposed different logics for different types of enterprises: For the 2B sector, enterprises should choose industries with strong payment capabilities such as industry and energy, control the performance cost, and reject "sham innovation". For the 2C sector, enterprises should deeply cultivate the local experience. She particularly emphasized: "Globalization requires a dual - team product logic. Chinese and overseas teams should adapt to different markets, and founders should also have cross - cultural management capabilities."

Guo Weike took the pet industry as an example to elaborate on the concept of "Creating Market Increment": Focus on AI technology for pet health, do not disrupt the existing rules, avoid direct competition with local enterprises, and create new demands through innovative functions such as health monitoring, rather than squeezing the existing market by relying on supply - chain advantages.

Ren Hangbo added from the platform perspective: "Differentiation is the key to customers' success." For products in the same category, enterprises should also accurately target niche groups. For example, AI toys can target children aged 6 - 8 and Gen Z women respectively to avoid homogeneous competition.

This article is from the WeChat official account "Hangzhou Qiantang Enterprise Going Global Service Base". Author: Zhejiang Enterprises Going Global. Published with the authorization of Qiantang.