Summary of Key Points
Chinese technology giants (ByteDoubao and Alibaba Qianwen) have begun to explore monetization strategies through AI platforms: ByteDoubao charges a 12% commission on closed-loop transactions, while Qianwen has opened its platform for third-party intelligent agents to integrate. There are significant differences in the monetization approaches of AI in China and the United States (ByteDoubao has a closed-loop advantage, whereas OpenAI lacks the necessary transaction infrastructure). AI is becoming the “decision-making intermediary” between users and services, reshaping the logic of traffic distribution and profit allocation. Traditional platforms (such as Ctrip) are facing challenges, while companies that possess both an AI interface and a closed-loop transaction system have a competitive edge.
Detailed Analysis
1. Is the 12% Commission for ByteDoubao High?
ByteDoubao charges a 12% commission on hotel bookings (11.4% for software services and 0.6% for payment processing), which may seem higher than the 8% charged by the抖音 main platform. However, there are several factors to consider:
- More targeted traffic: While抖音 attracts users through casual video browsing, ByteDoubao generates traffic from user-generated searches (e.g., “affordable hotels in Beijing’s North Fourth Ring”). Targeted ads are generally more expensive than organic content.
- Cost-effective compared to traditional platforms: Although Ctrip and Meituan charge between 15%-20% in commissions, businesses also have to invest in additional advertising for visibility, increasing their overall costs. ByteDoubao only charges on successful transactions, making its commission rate more affordable.
- User behavior: Although ByteDoubao has created a closed-loop from inquiry to purchase, 66% of users still verify information on Ctrip before making a decision, indicating that the commission is only effective when users do not switch platforms.
2. Qianwen’s Open Model: An AI-based “Mini Program Store”?
On the same day, Alibaba’s Qianwen platform was launched, allowing third-party services like SF Express and Ziru to integrate through intelligent agents. This approach differs significantly from ByteDoubao’s closed-loop model:
- No involvement in transactions: Users can send files to SF Express’ intelligent agents directly within their own systems or manage financial investments using Qianwen’s AI recommendations, with the actual transactions handled by third parties.
- Why the open approach?: Alibaba has a strong transaction infrastructure with platforms like Taobao and Fliggy. By adopting an open model, it allows these services to leverage Qianwen’s AI capabilities without being directly involved in the sales process, similar to how WeChat Mini Programs provide a platform for businesses to operate.
- Common goal: Both ByteDoubao’s closed-loop model and Qianwen’s open approach aim to control the “decision-making intermediary” layer. If users stop searching and comparing on their own and rely solely on AI recommendations, these platforms gain more influence over user behavior.
3. The Challenges Faced by ChatGPT
OpenAI in Silicon Valley attempted monetization earlier but encountered difficulties:
- High traffic, low conversion: Although ChatGPT attracted significant traffic to Walmart and Etsy, few merchants (only about a dozen out of millions using Shopify) actually made purchases through its recommendations.
- Lack of a closed loop: OpenAI lacks its own transaction platform, payment system, or fulfillment capabilities, forcing it to act as a mere intermediary. Even offering a 4% commission didn’t attract enough users, leading to the abandonment of in-platform checkout options.
- Cultural differences: The success of AI monetization depends on having a closed loop. ByteDoubao and Alibaba have their own transaction ecosystems, giving them a competitive advantage.
4. The Impact of AI on Traditional Platforms
The rise of AI intermediaries poses a threat to traditional platforms like Ctrip:
- Changing user behavior: Users no longer rely solely on platform rankings; instead, they seek direct AI recommendations. This shifts the balance of power in profit distribution.
- Revised profit models: Businesses may need to pay both AI services and fulfillment platforms, increasing their costs without a significant reduction in overall revenue.
5. The Competition among Giants
In the AI era, having both an AI interface and a closed-loop transaction system is crucial for gaining control over pricing:
- ByteDoubao and Alibaba: They combine their AI platforms with existing sales channels to charge higher commissions.
- Meituan: It focuses on developing its own AI assistant (Xiaotuan) while maintaining control over its local services to retain its market position.
Conclusion
In the AI era, those who control both the AI interface and the transaction ecosystem will have the power to set prices and reshape the internet’s traffic and profit models. The ultimate goal for these giants is to transition users from simply verifying information to making direct purchases through AI recommendations—more accurate and seamless experiences can completely transform the industry.
(You could conclude by reflecting on the initial question: “Would you trust AI recommendations to make your purchase decisions directly?” This is precisely what the tech giants are working towards, aiming to create a smoother and more reliable purchasing experience that will revolutionize the entire industry.)