When you inquire in an AI chatbox about “which international kindergarten is better in Shenzhen,” the answer you receive may not be solely based on objective data but rather a carefully orchestrated “implantation.” This is all thanks to a burgeoning market known as GEO, which is believed to have a significant impact on the credibility and viability of AI.
Phoenix Technology reported on the 14th that when you open DeepSeek and type in “the best restaurant in xx place,” the AI answer you receive may have undergone some human “design.”
This is where GEO comes into play.
The concept of GEO (Generative Engine Optimization) was first proposed by researchers from the Indian Institute of Technology and Princeton University. It optimizes content to ensure that brand information becomes a “mentioned entity” in AI-generated answers. Studies have shown that by optimizing authoritative expressions, keyword layouts, and data citations, the visibility of content in AI responses can increase by up to 40%.
In China, large-scale dialogue models like DeepSeek are seen as new traffic gateways. Some SEO companies are transitioning to AI search optimization, and even practitioners from media agencies believe this to be the “closest to the AI trend.”
For small and medium enterprises, GEO offers new marketing avenues. Some startup companies have shown successful cases in their service manuals: in inquiries about “Shenzhen International Kindergartens,” the name of a bilingual kindergarten was recommended first.
Food and fast-moving consumer goods brands are among the earliest adopters. Some businesses even directly print signs saying “Recommended by DeepSeek in conversation, storefront” on e-commerce platforms or offline stores to promote their products or services. For them, being the “first recommendation by AI” seems to have become a kind of “authoritative” certification.
Phoenix Technology believes that as AI dialogue becomes a mainstream means of search, the era of “citation equals existence, non-mention equals death” for AI brands is approaching. Future marketing will no longer rely on traditional SEO or SEM, but will require adaptation to the new ecosystem of the AI era. Will there be a commercial model similar to “bidding ranking” in future AIs?
This emerging wave of GEO poses a serious question to people: as AI becomes our primary source of information, are the “answers” we see truly objective and fair, or are they influenced by commercial interests?
Regarding this, Chinese issues expert and senior commentator Wang He told Epoch Times, “In the past, Baidu’s bidding ranking caused many negative social impacts, even leading to some malicious incidents. Now, in the AI era, if we continue to optimize searches and bidding rankings under the influence of commercial interests, it will have a huge impact on the credibility and feasibility of AI.”
Wang He stated that the biggest issue with generative AI is that its authenticity and accuracy are not guaranteed. If capital forces intervene and promote it as a commercial model, its credibility and feasibility could suffer a devastating blow, affecting AI and its development far beyond people’s imaginations.
In reality, GEO faces numerous challenges. Phoenix Technology indicates that currently, most GEO companies adopt the strategy of “feeding AI-generated content to AI.” This involves using AI to generate various “network drafts” as language materials after obtaining corporate data, which are then distributed to accessible websites to be indexed by AI. This method is cost-effective, but its effectiveness has yet to be accurately evaluated.
A more profound issue lies in the opaqueness of the large model algorithms, making GEO still a “black box” optimization. Furthermore, each model’s adoption preferences vary, and there’s no “one-size-fits-all solution.”
Industry insiders warn that the practice of “using AI to pollute AI” carries risks. Large model manufacturers have anti-cheating strategies and may adopt a “zero-tolerance” defense against low-quality, bulk “gray production” models in the future. Some practitioners believe that the core competitive advantage of GEO lies in producing truly valuable, high-quality content to supplement the needs of the AI ecosystem.
Wang He remarked, “One thing that can be affirmed here is that if a person lacks good morals and values, the operation of any of their business models will bring traps that are hard to escape.”