When most people are anxious about writing and making videos with ai, someone's looking at something deeper -- when users stop searching and start asking questions, can your brand be found
The law week reporter lu jing
In a creative park outside the eastern ring of beijing, the wind in march was cool. A sharing event around geo is under way, with more than 100 people sitting on the floor — twice as many as the organizers expected — and some even sitting on the steps。
The shared guest on stage, liu qian, who calls himself the "angel investor of the flow circle", is lighting the screen with a remote pen。
Liu has clearly divided current internet traffic into three forms: recommended traffic, search traffic and social traffic. He pointed out that the core characteristic of the recommended flow was that it was extremely explosive, but that it was impossible to escape the fate of the “one wave”, whether it was a phenomenon-level content that had been painted or a hot spot associated with the artist, which tended to keep a few months of heat in order to rapidly sink. And that is where a large number of businesses suffer — relying too much on the platform's recommended traffic, and eventually mostly caught in the trap of algorithms, and there are few cases in which practitioners on the ground can stabilize customers through recommended traffic。
In contrast to this, the search traffic. According to liu, the core value of the search flow lies in high conversion rates and strong certainty. Users do not search the platform without a destination, and every search is based on real and clear needs. Only when a specific claim is made by the user will the key word be entered to find a solution, and if the content matches the demand, conversion will occur naturally, which is the core logic that search flows are much better for commercial transformation than recommended flows。
"at the time of the ai, the rate of transformation will be higher." liu said。
What we're talking about on the stage is a new area that is emerging: the geo service. When most people are anxious about how to write with ai and make videos, someone has turned their eyes to something deeper — when users stop “searching” and start “questioning”, and your brand can still be found
From search to recommended to question
As a geo service provider, liu was one of the early explorers in the field. He drew a map of traffic evolution。
The first act is the search age. 100 degrees, google dominates people's internet portals, and users knock on the search box to search for information they want, the seo. Liu said, "the user comes with the demand, you meet the demand, and the deal happens."
The second is the era of recommendations. The rise of the tremors completely subverts the flow logic. Information is no longer being searched but is being sent. Users immersed in algorithm-weaved cocoon rooms, and merchants caught up in an endless competition — whoever can make something more powerful can take more eyeballs. But the flow is like fireworks, but it's short. Liu qian showed a comparison showing that the percentage of recommended traffic in a live broadcast was 99. 96 per cent, but only 24 per cent of the turnover. The remaining 0. 04 per cent of the search flow resulted in 76 per cent of the collection。
The third act is on, and liu has called it the "ai era". The entry is no longer a search box, but a dialogue window. Users no longer enter keywords, but ask a complete question: “what brand of long-term velvet is more expensive?” “how many years can buy mercedes e300?” “what travel agency does beijing travel rely on?”
Bean buns, deepseek, chishu, yuanbao these ai applications are becoming new flow hubs. User “question”, ai “a answer”. And what geo is going to do is not raise the ranking of the search results, but make it easier for a business or brand to be the subject of an ai-selected, written answer。
Small and medium-sized merchants in the new battleground of the ai era
In liu's view, those who have the capacity to produce large-scale content monopolize the ai answer for the industry. But he turned the corner: “for most people, it is enough to make a few precise long endings, each with one or two clues a day. It doesn't take a big team to do it alone."
"because ai doesn't look at the number of fans, doesn't look at the number of playouts, it just depends on whether your content is logical or not. This means that whoever can produce more valuable, neutral, ai-approved content can win.” liu qin says。
Liu qian set up a group of participants. One of the law students runs a small law firm in the northeast and has never been promoted online. After learning about geo, he went back for a month to optimise, and the brand reference rate on the ai platform increased from 0 to 80 per cent. He has a new mobile phone number dedicated to telephone calls under the geo service, which is basically two calls per day. Last week, two cases were signed。
Another guangzhou student, after studying geo optimization, simultaneously optimized his brand and trademark, with the result that in ai questions and answers, first in company and second in trademarks, the competition disappeared。
Liu predetermines the end of the geo service layout: head companies build their own systems, produce thousands of articles every day, turning the official network into an industry knowledge base; medium-sized companies use ai+ efficiency processes to accurately strike long-term keywords; and small businesses and individuals can reap dividends by learning how to choose words and what to do。
Journalists participated in the sharing event and found that the physical exercise of geo service providers has become highly standardized and process-oriented: first, building text content that is more easily captured and understood by the ai mega-models, based on the identification logic of the ai algorithm; secondly, placing the mass-produced content on the website carrier, which is more trusted by the ai-mode; and finally, waiting for the ai-mode to retrieve, aggregate and generate summary outputs。
Journalists note that the current market-based outreach of geo service providers is also characterized by a high degree of commercialization: “market exclusive geo monitoring tool, real-time tracking of brand ai performance”, “systemistic methodology + hand-held hand-held, 5 days complete geo warfare system, ready to do it”, “ai tool in-depth empowerment, customization of exclusive work streams by top national experts, 10 times more effective content creation process”。
How big models respond to geo services. The reporter for the rule of law weekend interviewed the staff of one of the ai megamodels, who indicated that they could not respond。
Where are the boundaries of the geo service
Ai brought us a lot of convenience. But does the search results given by ai have to be objective
At the 3-15 p. M. Event, a large model named geo was just named "toxas" and a smart watch called "apollo-9" was invented and a software called "the power geo optimization system" was purchased. After the geo service provider generated software and mass delivery of the apollo-9 product information, two major ai models recommended the fictional bracelet on the same day in a question about the smart health handrings, and ranked ahead of the ranking, even directly following the fully fabricated term “quantum tangeration”。
Fan mingzhi, a professor at the institute of data and rule of law of the chinese university of political science and law, told journalists of the rule of law weekend that some traders, using geo optimization, had achieved the ranking of brands and trademarks as monopolistic and directly excluded from competition in ai questions and answers, which, by their nature of behaviour, constituted undue competition. Under article 13 of the anti-improper competition act, “internet-specific articles”, it is expressly prohibited to “obstruct or disrupt the proper functioning of network products or services legally provided by other operators”。
In the area of search engines, there are typical examples of search engine platforms fighting cheating in the area of search engines. In january 2023, the beijing intellectual property court rendered a final judgement in the case of beijing 100-degree telecommunications and technology ltd. (hereinafter referred to as " 100-degree " ) v. Shenzhen one-eu network technology ltd. (hereinafter referred to as " one-euns " ): a company in favour of 100 million yuan in damages and $50,000 in reasonable costs。
According to the judgement, 100 degrees are the actual operators of the 100 degrees network. One of the leading companies operates a treasure shop called seo optimization 100-degree ranking (hereinafter referred to as the shop in question), which provides products that specifically target the results of the optimization of search engines such as the 100-degree network. With the selection of keywords and links to the target website by the purchasers of the above-mentioned commodities, a good company was able to raise the ranking of the target site, which had been ranked low, to the front page of the nature search results for a certain period of time. In the view of the company, these actions interfered with the normal ranking of 100 degrees of natural search results, undermined their legitimate interests and constituted undue competition. A company of excellence was therefore requested to compensate for economic losses and reasonable expenses totalling $2 million and to eliminate the effects。
The court of first instance held that the number of hits on the target website, which is ranked higher on the site's natural search results, is not based on real search needs of real users, does not reflect the website quality of the target website and the match to the user's needs in a realistic and objective manner, and is clearly contrary to the business will of the client. Following a first instance judgement in the case, a cassation appeal was filed by a company of yu, which upheld the sentence in the second instance。
Van mingzhi stated that 100 degrees v. Shenzhen shenzhen shenzhen was clearly found to constitute an undue competition and awarded $1. 8 million for falsely clicking on the natural order of the intervention engine. The thrust of the decision is clear that the search engine has a competitive interest in the natural sequencing of its algorithm output and that third parties may not distort it by technical means. Large model content generation is highly similar to the legal nature of the search for natural sequencing, and geo volume interventions can be applied by analogy. Second, the general provisions of article ii of the anti-improper competition act (good faith, business ethics) can be applied in a bottom-up manner if the enumeration provisions cannot be accurately matched。
Fan mingji added that in five typical cases of unfair competition in the field of ai, published by the national directorate of market supervision in february of this year, there have been enforcement cases involving the use of competitive ranking + imitation techniques to interfere with the results of the ai search (the deepseek case in beijing)。
Currently, geo services have developed a standardized operating chain, with some traders opting for geo optimization and even achieving brand and trademark monopolistic rankings in ai questions and questions, and how should the larger ai model respond
