钛镁AI吉晨亮：GEO的第一性原理 是构建大模型对品牌的信任机制

This article is an edited transcript of a talk by Chenliang Ji, co-founder of Taimeng AI, at the Global New Brands AI Competitiveness Conference. It shares core insights on Generative Engine Optimization (GEO) in the AI era, helping readers quickly grasp the key information and practical logic of this new direction in AI marketing.

1. The industry has fundamentally shifted: AI has become a new entry point for consumers to access information and make shopping decisions. Combined monthly active users of major global AI platforms exceed 1.2 billion. 75% of global consumers accept AI-generated shopping recommendations, and 68% of overseas consumers have already completed purchases based on AI suggestions. A full one-stop AI-powered shopping闭环 has formed, making AI marketing an entirely new industry growth opportunity.

2. The core of GEO is building a trust mechanism for large models to recognize brands, enabling brands to secure effective recommendations when large models generate responses. For cross-border brands, the practical core of GEO lies in optimizing high-authority information sources including official websites, Reddit, YouTube, Wikipedia and industry-leading media outlets. When paired with analysis of real user intent and optimization of model perception, this approach delivers clear performance improvements.

AI has become the new competitive arena for brands, shifting brand competition from keyword rankings in the search era to large model perception competition in the AI era. GEO is the core marketing strategy for brands to acquire traffic and drive growth in the AI age, and offers critical guidance for brands looking to capture new traffic entry points.

1. Consumer trends have shifted: 75% of global consumers accept AI shopping recommendations, and 68% of overseas consumers have already purchased products based on AI suggestions. AI has become a one-stop connector between consumer demand and transactions, and traffic is rapidly shifting to AI platforms. Brands need to布局 early to seize first-mover advantage.

2. GEO has a clear four-step implementation framework: identify high-intent user questions, build a brand knowledge graph to improve large model perception, use probes to continuously monitor large model iterations, and optimize布局 of high-authority information sources. For cross-border brands, official websites account for over 58% of large model source weight, and brands should also expand their presence on Reddit, YouTube, Wikipedia and other channels.

3. GEO cannot be replaced by GEM (Generative Engine Marketing) ads. To maximize traffic capture and conversion, brands need to combine GEO and GEM to cover the entire user decision journey.

AI-powered shopping recommendations are growing rapidly, representing an entirely new growth opportunity for sellers, especially cross-border sellers. By leveraging the GEO trend and布局 early, sellers can capture upcoming AI traffic shifts and secure first-mover growth advantages.

1. Opportunity overview: Combined monthly active users of major overseas AI platforms exceed 1.2 billion, and more than 60% of overseas consumers have already completed purchases via AI recommendations. A new full AI-powered shopping decision闭环 has already formed. While current direct conversion volume remains modest, early布局 comes with lower competition costs and makes it far easier to seize first-mover advantage.

2. Practical implementation takeaways: For cross-border sellers implementing GEO, the first priority is restructuring the content and architecture of your official website, followed by布局 of high-authority sources including Reddit, YouTube, Wikipedia and industry-leading media. B2B sellers should additionally build professional credibility on LinkedIn. The full implementation process is: identify core user search intent, build a brand knowledge graph, and continuously monitor and adjust for model iterations.

3. Risk warning: Do not focus solely on GEM ads and neglect GEO. Inaccurate natural perception of your brand from large models will significantly drag down ad conversion rates. GEO and GEM must work together to effectively capture and convert traffic. Case studies from leading 3C brands have validated the approach: after GEO optimization, core brand mention rates rose from 20% to over 95%, and conversions increased 56% month-over-month, proving the clear effectiveness of the strategy.

Large AI models have reshaped global consumer decision-making and brand marketing logic, creating new business opportunities for factories building their own brands, and offering new insights for factories pursuing digital transformation and branded e-commerce布局.

1. New business opportunity: AI is now an entirely new traffic entry point, and a growing share of consumers rely on AI recommendations for shopping decisions. Factories building own-brand cross-border businesses can acquire AI traffic through GEO optimization. Compared with mature traditional traffic entry points, competition for AI traffic remains underpenetrated, so early布局 enables factories to secure first-mover advantage at lower cost and unlock new growth opportunities.

2. Insights for digital and e-commerce transformation: Factories launching own brands do not need to commit large advertising budgets upfront. By restructuring official website content and布局 high-authority public-domain information sources, factories can build accurate brand perception among large models, and accumulate stable organic traffic over the long term. Factories can also leverage tools from professional GEO service providers to lower entry barriers and complete optimization quickly.

3. GEO optimization requires factories to clearly articulate their brand positioning, product specifications and core capabilities, which in turn pushes factories to standardize their product systems — a process that also improves factories' overall product management capabilities.

GEO is an entirely new brand marketing service segment emerging in the age of large AI models. It is currently in the early stage of rapid development, has broad industry prospects, and already has mature implementation frameworks to draw from.

1. Industry development trends: AI has become a new entry point for consumer shopping decisions, and brand competition has shifted to large model perception competition. Brand demand for GEO services is growing rapidly. In the future, GEO and GEM will become the two core modules of AI marketing. Industry group standards are currently under development, and the segment is gradually moving toward standardization, with enormous untapped market potential.

2. Core customer pain points: Most brands lack clarity on how to help large models accurately understand their brands and products. Large models iterate rapidly, so their perception of brands often shifts. Traditional knowledge bases only solve the hallucinations problem in content creation, and cannot address the core pain point of inaccurate model perception. Brands also lack guidance on which information sources to布局 to maximize authority weight.

3. Proven mature solution: Service providers can build an intent inference system to identify high-intent user questions, optimize model perception of brands through knowledge graph construction, deploy Agent probes to continuously monitor rule changes from large model iterations, and offer differentiated source布局 recommendations for different markets. This framework has been validated through real-world practice, and can significantly improve brands' core mention rates and conversion performance.

Large AI models have become new entry points for consumer transactions, and brands have substantial unmet demand for AI platform marketing. Platforms can build an entirely new business ecosystem centered on GEO and GEM to capitalize on the new opportunities brought by AI-driven traffic transformation.

1. Core brand needs: Brands need to secure accurate exposure on large model platforms, build consumer trust, and ultimately drive conversions. Currently, most brands lack guidance on how to布局 GEO compliantly, and need platforms to provide supporting tools, clear rules and auxiliary services.

2. Actionable development directions: ChatGPT has already launched GEM ad placement paths that support direct jumps to e-commerce transactions for order completion. Other AI platforms can follow this example to launch matching marketing products, while optimizing their own source weight mechanisms to guide brands to produce high-quality official content. This approach both improves the accuracy of model responses and opens up new marketing revenue streams for platforms.

3. Risk mitigation: Large models are prone to hallucinations that generate incorrect brand information, harming consumer experience and eroding brand trust in the platform. To address this, platforms can optimize their RAG retrieval mechanisms to increase the weight of high-authority information sources, open official information submission channels for brands, and on-board compliant GEO service providers to offer supporting services for brands and complete the platform ecosystem.

The widespread adoption of large AI models is driving profound transformation in the global brand marketing industry. The emergence of GEO as an entirely new marketing field has created many new industry trends and new research directions.

1. New industry trends: Brand competition has shifted from keyword ranking competition in traditional search to perception competition in large AI models. AI has become a new traffic entry point and a full transaction闭环 carrier, and GEO has emerged in response to this shift. The core of GEO is building a trust mechanism for brands within large models. In the future, GEO and GEM will together form the complete commercial system of AI marketing. The industry is currently in an early development stage with rapid growth.

2. New research questions worth exploring: Large models are constantly iterating black boxes, with their source structures and ranking algorithms kept private. Key open research questions include: how can brands continuously optimize their perception in large models? What are the differences in source structure across different large models? What metrics should be used to measure GEO effectiveness? How can GEO and GEM work together to improve conversion efficiency? All of these are entirely new research topics.

3. On the business model front, an independent third-party GEO service business model has already emerged, offering end-to-end optimization services for brands, and the market has already validated its effectiveness. Industry group standards are currently under development, and the segment is gradually moving toward standardization, leaving substantial room for future innovation and research.