人工智能（AI）驱动的搜索引擎和虚拟助手正通过用“引用经济”取代以点击为驱动的指标来重塑数字营销，专家表示，随着生成式AI内容的泛滥，生产原创且可验证的内容对品牌脱颖而出变得越来越重要。
Search engines and virtual assistants powered by artificial intelligence (AI) are reshaping digital marketing by replacing click-driven metrics with a “citation economy,” as experts say producing original and verifiable content has become increasingly important for brands to stand out amid the oversaturation of generative AI content.
普林斯顿大学及其合作机构的一份报告指出，在ChatGPT、Perplexity和谷歌AI概览等AI平台上的数字可见性取决于“生成式引擎优化”，即“GEO”。
Digital visibility across AI platforms like ChatGPT, Perplexity and Google AI Overviews hinges on “generative engine optimization,” or “GEO,” according to a report by Princeton University and partner institutions.
研究人员开发了一个名为“GEO-bench”的测试数据集，其中包含超过10,000个不同的查询，用于分析生成式搜索引擎的数据处理架构。
Researchers developed a test dataset called “GEO-bench,” which includes over 10,000 different queries to analyze the data processing architecture of generative search engines.
测试显示，包含可验证统计数据、可信引用和专家观点的数字内容在AI生成的答案中的可见性最高可提升40%。
The test showed that digital content with verifiable statistics, credible citations and expert opinions increased its visibility in AI-generated answers by up to 40%.
随着越来越多的用户依赖AI生成的摘要，而非点击搜索引擎查询结果，全球营销机构正在重塑其策略以适应这一变化趋势。
With more and more users relying on AI-generated summaries rather than clicking through search engine query results, marketing agencies worldwide are reshaping their strategies to adapt to the changing trend.
专家表示，这一转变将 usher in 一个新时代，生产高质量原创内容的品牌将胜过那些依赖巨额广告预算的品牌。
Experts say this transition will usher in a new era where brands that produce high-quality original content will prevail over those that boast massive ad budgets.
Indexa的联合创始人兼首席执行官Bunyamin Sarac向安纳多卢通讯社表示，随着互联网流量从简单的基于点击的系统转移，出现在AI聊天机器人答案和AI生成的搜索引擎摘要中的品牌更为成功。
Bunyamin Sarac, co-founder and CEO of Indexa, told Anadolu that brands appearing in AI chatbot answers and AI-generated search engine summaries are more successful as internet traffic shifts away from the simple click-based system.
Sarac 表示，AI 引擎利用“来源权威性”“时效性”和“可提取性”等信任信号来评估实时搜索结果。
Sarac stated that AI engines evaluate real-time search results using trust signals like “source authority,” “timeliness” and “extractability.”
“这些模型的开发者并未完全公开其来源选择机制的细节，因此我们无法说有一个确定的公式，但研究和观察显示，这些模型会关注某些信号，”他说。“AI 回答由两个来源训练而成：用于训练模型的数据集，以及在生成提示词答案那一刻的网络搜索查询。”Sarac 指出，当用户收到 AI 摘要时，点击率会下降，他引用皮尤研究中心的发现称，虽然阅读了 Google AI 摘要的用户中有 8% 点击了生成的链接之一，但未看到摘要的用户中这一比例几乎翻了一倍。
“Developers of these models do not fully share the details of their source selection mechanism, so we can’t really say that there’s a definitive formula, but research and observation showed certain signals that these models look for,” he said. “AI responses are trained by two sources: the data set used to train the model and web search inquiries at the moment that the answer is generated to a prompt.”Sarac noted that click-through rates drop when users receive an AI summary, citing Pew Research Center findings that showed that while 8% of users who read an AI summary on Google clicked on one of the resulting links, this rate was nearly double among those who did not see a summary.
他补充道：“我们预计来自信息搜索类查询的网络流量将会下降，因为点击网站的访客将是那些已经阅读过 AI 摘要或通过 AI 做过调研、且更接近做出购买决定的人——成功的衡量标准将从点击次数转变为产品或品牌在关键搜索查询中被推荐的次数。”
“We expect a decline in web traffic from information-seeking inquiries, as visitors who click on a website will be those who have already read the AI summary or done their research through AI and are closer to making a purchase – the metric for success will shift from the number of clicks to the number of times that a product or brand was recommended for critical search inquiries,” he added.
Analythick 联合创始人兼首席执行官 Erdem Dagdemir 告诉安纳多卢通讯社，传统的搜索引擎优化（SEO）侧重于技术代码和海量数据，这已不再足够。他敦促品牌采用一种更新的方法，突出人类语境、真实的客户满意度和有机的数字足迹。
Erdem Dagdemir, co-founder and CEO of Analythick, told Anadolu that conventional search engine optimization (SEO) focused on technical code and voluminous data, which is no longer sufficient. He urged brands to embrace a newer approach that highlights human context, real-world customer satisfaction and organic digital footprints.
“AI 看重的不是数据量，而是其定性准确性和深度，它通过交叉引用品牌的数字足迹来综合信息，并询问来源是否可信、客户是否满意，”他说。
“AI does not look at the volume of data but at its qualitative accuracy and depth, synthesizing information by cross-referencing a brand’s digital footprint and asking if the source is trustworthy and whether customers are satisfied,” he said.
“品牌必须规划其运营，建立有效的沟通，并通过优先考虑客户满意度来建立信任，”他补充道，并指出，如果这些数字足迹无法证实品牌的企业承诺，人工智能可能会完全将该品牌从推荐中剔除。
“Brands must plan their operations, establish effective communication, and build trust by prioritizing customer satisfaction,” he added, noting that if these digital footprints do not corroborate a brand’s corporate promises, AI might completely remove the brand from recommendations.
达格德米尔表示，大型语言模型（LLMs）会浏览在线论坛和评论网站，以揭示持续存在的消费者投诉，从而决定某个品牌或产品是否可以向用户推荐。他补充说，坚持过时方法的营销团队，例如过度堆砌表面内容，将会被提供可验证信息的竞争对手所超越。
Dagdemir stated that large language models (LLMs) go through online forums and review sites to reveal persistent consumer complaints to decide whether a brand or product can be recommended to users. He added that marketing teams clinging to outdated methods, like oversaturating superficial content, will lose ground to competitors who offer verifiable information.
“人工智能聊天机器人被设计为提供清晰快速的回应，这迫使人们和行业适应这一新节奏，而土耳其的生态系统正站在这个适应的门槛上——虽然一些品牌仍在使用过时的方法来维持其突出地位，但新时代的基本规则不再是关于可见性，而是关于可靠性。在信息泛滥和优质内容稀缺的环境中，我认为那些放弃以广告驱动的策略、转而提供真实准确信息的品牌，将在人工智能生成的搜索结果中获得相较于竞争对手的显著优势，”他补充道。
“The clear and fast responses AI chatbots are made to deliver force people and industries to adapt to this new pace, and Türkiye’s ecosystem stands at this threshold of adaptation – while some brands are still using outdated methods to cling to their prominence, the fundamental rule of the new era is no longer about visibility but about reliability, and amid the clutter of information and scarcity of quality content, I think brands that abandon ad-driven strategies in favor of offering genuine and accurate information will gain a significant edge over competitors in AI-generated search responses,” he added.