From SEO to GEO, from digital marketing to AI Marketing—standards first, metrics to lead. How can auto brands excel at marketing in the AI era? This article begins with the birth of a standard.

-01- China’s First GEO Group Standards for the Internet Advertising Service Industry Officially Released

On September 14, 2026, five group standards titled *Generative Engine Optimization (GEO)*, proposed and administered by the China Advertising Association of Commerce (CAAC) and organized by its AI Marketing Application Working Committee, officially passed review by the National Group Standard Information Platform and were publicly announced.

This is China’s first systematic set of group standards for generative engine optimization, filling the standardization gap in the field of generative AI marketing.

The five standards and their numbers are as follows:

No.	Standard Number	Standard Name
1	T/CAACCHINA 001-2026	Generative Engine Optimization (GEO) — Part 1: General Introduction, Terms and Definitions
2	T/CAACCHINA 002-2026	Generative Engine Optimization (GEO) — Part 2: Service Provider Evaluation Specification
3	T/CAACCHINA 003-2026	Generative Engine Optimization (GEO) — Part 3: Pricing, Evaluation and Measurement Specification
4	T/CAACCHINA 004-2026	Generative Engine Optimization (GEO) — Part 4: Trustworthy Corpus Specification
5	T/CAACCHINA 005-2026	Generative Engine Optimization (GEO) — Part 5: Service Process Compliance and Security Requirements

The five standards cover five major areas of GEO services: general terms, service provider evaluation, pricing and measurement, trustworthy corpora, and compliance and security. A total of 41 participating units and 75 industry experts took part in drafting them.

During the public comment period, 38 companies and 13 experts submitted feedback, with a total of 174 revision suggestions. Beijing Xinsi Technology Co., Ltd. was honored to be one of the companies providing feedback and deeply participated in the process of improving this industry standard.

How to obtain the full text of the standards:

Social institutions can obtain the full text through the following two entry points:

1. National Group Standard Information Platform: visit http://www.ttbz.org.cn, search for “Generative Engine Optimization (GEO),” and view the five group standards;
2. Official website of the China Advertising Association of Commerce: visit http://www.caacchina.org, go to the Notices and Announcements section, and download the complete electronic version of the Generative Engine Optimization (GEO) group standards.

In addition, the AI Marketing Application Working Committee of the China Advertising Association of Commerce will hold an offline release conference for the *Generative Engine Optimization (GEO)* group standards in Shanghai on October 16, 2026, officially releasing the standard texts and interpreting their core content. Industry colleagues are warmly invited to witness it together.

-02- Xinsi Technology’s Feedback and Observations on the GEO Standards

As a company long focused on brand digital measurement and research on AI citation data, Xinsi Technology carefully studied the texts of the five standards during the public comment period and submitted feedback from the dual perspectives of a brand-side stakeholder and a measurement technology provider.

We focused on the following four aspects:

1. Welcoming the Clarification of the Relationship Between AEO and GEO

During the revision, a definition of AEO (Answer Engine Optimization) was added, and its relationship with GEO was clarified—a revision we strongly endorse. Generative engines and answer engines differ in form and optimization logic. Clarifying terminology boundaries can effectively prevent conceptual confusion in the market and allow brands to “speak the same language” when procuring and evaluating services.

2. Trustworthy Corpus Specification Is the Cornerstone of GEO

The essence of GEO is not “optimizing for AI,” but providing AI with trustworthy corpora. Part 4, the Trustworthy Corpus Specification, directly addresses the industry’s greatest concern—corpus poisoning and contamination by low-quality content. In our feedback, we emphasized that building trustworthy corpora must return to the brand’s own source system. Authoritative sources led by the brand (official websites, technical documentation, authentic user reviews, authoritative evaluations) should become the main sources cited by AI, rather than relying on patchwork optimization methods from third parties.

3. Pricing and Measurement Specifications Need a Unified “Weights and Measures”

The biggest problem GEO services previously faced was that results could not be compared: each service provider defined its own metrics and issued reports under its own standards, making it difficult for brands to judge real effectiveness. Part 3, the Pricing, Evaluation and Measurement Specification, provides a unified framework for the industry. We suggest that the measurement system should be directly linked to the brand’s business metrics (the degree to which the brand is recommended by AI, AI citation share, etc.), so that the input-output of GEO can be quantified, tracked, and reviewed.

4. Compliance and Security Are the Industry’s Bottom Line

Part 5, Service Process Compliance and Security Requirements, along with adjustments such as uniformly revising “high-risk industries” to “industries with high compliance requirements,” reflects the prudent attitude of the standard drafters. GEO should not become a tool for speculation. The implementation of these standards will reduce vicious competition and protect companies that truly win through content and source quality.

-03- GEO Is an Important Part of Integrated AI Marketing—But Only One Part

The release of the five standards marks a new stage in which the GEO industry has “standards to follow and rules to abide by.”

But we also want to remind the industry: GEO’s relationship to AI Marketing is just like SEO’s relationship to digital marketing—it is an important specialized component, but by no means the whole.

Looking back at the digital marketing era: SEO solved the problem of “whether search engines can find you and how they present you,” but ranking first in search results does not equal marketing success. What truly led digital marketing was the continuous management of digital brand assets and the overall coordination of product, price, place, and promotion (the 4Ps).

Only when SEO is embedded in brand asset management and the 4P framework can it generate sustained commercial value.

The same applies in the AI Marketing era. GEO solves the problem of “how generative engines understand you, cite you, and recommend you,” but whether a brand can succeed in the AI era depends on a higher level of integration capability:

• Source building: Does the brand possess an authoritative content system that AI can trust and cite?
• Content management: Can the brand’s knowledge, products, and reputation be structured and “fed” into the AI ecosystem?
• Products and services: Can AI feedback flow back into new product development and new service development?
• Metric leadership: Is there a North Star Metric that gives all AI marketing actions a unified measurement standard?

This is precisely the next level of thinking we advocate.

-04- Using IPS as the North Star Metric to Lead Automakers’ AI Marketing

Just as the digital marketing era needed digital brand assets to lead the overall picture, the AI Marketing era needs a new North Star Metric. We recommend: IPS (AI Recommendation Index).

IPS measures the degree to which a brand is understood and recommended by generative AI—when users ask AI “What SUV should I buy with a budget of 300,000 yuan?” or “Which new energy light truck is more trustworthy?”, is your brand mentioned by AI, placed on the recommendation list, and given positive reasons?

IPS + NPS (Net Promoter Score) form a dual metric system covering both AI feedback and human feedback: NPS measures the human heart, while IPS measures the “machine mind”—AI’s judgment of your brand.

In IPSmoto’s multi-market LLLM monitoring practice, we cover 16 markets and continuously monitor 11 mainstream large models. This allows us to directly observe differences in how auto brands are recommended across different markets and different models—this itself is the data foundation for automakers to calibrate their AI marketing strategies.

For automakers, AI Marketing led by IPS should focus on four levers:

1. New Product Development: Let AI Feedback Enter Product Definition

AI aggregates massive amounts of real user voices and public opinion. Monitoring IPS attribution data (which models are recommended, and for what reasons they are or are not recommended) is equivalent to continuously obtaining a dynamic “product reputation census.”

Automakers should incorporate AI feedback into new product definition and configuration decisions—the range anxiety, driver-assistance concerns, and after-sales experiences that users complain about to AI are precisely the questions the next generation of products must answer.

2. Service Development: Use AI Feedback to Optimize Service Experience

Purchase consultations, maintenance knowledge, troubleshooting—a large number of service touchpoints are migrating to AI conversations. Whether a brand’s service capabilities (charging network, maintenance policy, rescue system) can be accurately understood by AI and used as reasons for recommendation directly affects the “AI visibility” of its service advantages. At the same time, service shortcomings exposed in AI feedback are direct inputs for service process optimization.

3. Source Building: Brands Must Become Authoritative Sources Cited by AI

This is the corporate-side response to Part 4 of the GEO standards, the Trustworthy Corpus Specification. Automakers should systematically build an AI-friendly source system: official technical documentation, safety test reports, authentic owner reviews, and authoritative media evaluations, forming a structured, verifiable, and continuously updated brand source library.

Let AI “have evidence to check and trustworthy sources to cite” when it needs to cite—this is both the compliance path for GEO and the way brand assets are accumulated in the AI era.

4. Content Management: From “Content Read by Humans” to “Content Co-Read by Humans and Machines”

Content management in the AI era requires content to be friendly to both human readers and generative engines: structured product parameters, clear comparison data, and traceable factual statements.

Automakers need to establish a management mechanism of “content as source,” so that every content asset can enter the citation cycle of the AI ecosystem.

Conclusion：Standards as the Foundation, Metrics as the Guide

The release of the five GEO group standards provides the industry with a “bottom line and confidence”; while the “height and direction” of AI Marketing need to be led by a North Star Metric.

For automakers, our recommendation can be summed up in one sentence: Use IPS as the North Star to measure the effectiveness of AI marketing, use source building and content management to solidify the foundation of GEO, and use AI feedback to feed back into new product and new service development—just as the digital marketing era used digital brand assets to lead SEO and 4P coordination, the AI Marketing era uses IPS to lead GEO and the integration of the entire marketing chain.

With standards to follow and metrics to guide. The era of integrated AI Marketing is arriving.