yao-geo-page-blueprint

Yao GEO Page Blueprint

Use Cases

• Turn the target content, target questions, brand knowledge base, user paths, and conversion goals into a GEO-friendly page plan.
• The page type is a product page, topic page, article page, listicle page, comparison page, explainer page, knowledge base page, case study page, or FAQ page.
• The page must simultaneously serve scraping by domestic AI platforms, answer extraction, real user reading, and conversion.
• Deliverables need to include HTML semantic structure, Schema recommendations, a CMS field list, and reports in four formats.

Not Applicable Scenarios

• Site-wide diagnosis of brands, competitors, sources, and AI answer samples; should instead use yao-geo-panorama-audit。
• GEO backend effect attribution, tracking, or a closed tracking loop; should instead use yao-geo-tracking。
• Writing only articles, titles, WeChat official account posts, or content rewrites, without needing a page implementation plan.

Required Reading Materials

• references/page-blueprint-method.md
• references/research-foundation.md
• references/cn-ai-platform-adaptation.md
• references/schema-html-cms-contract.md
• references/real-data-workflow.md
• references/artifact-layout.md
• references/quality-gates.md
• references/output-file-workflow.md

Execution Workflow

1. Choose the page type: product page, topic page, listicle page, comparison page, explainer page, knowledge base page, case study page, or FAQ page.
2. Determine the data mode:provided-source-pack、assisted-web-research、ai-answer-sample or synthetic-planning; when real data is needed, follow references/real-data-workflow.md to establish a source pack.
3. Align the target questions, user paths, brand facts, conversion goals, and compliance boundaries.
4. Establish a table of inputs, assumptions, and boundaries, clarifying known facts, design assumptions, gaps, compliance restrictions, and items to be confirmed.
5. Establish a plan for real data access and verification, distinguishing official sources, client knowledge bases, third-party materials, AI answer samples, and plan assumptions.
6. Establish a research basis mapping: translate GEO visibility, Google generative AI search, query fan-out, long-context position bias, RAG evidence traceability, structured features, Schema, and WCAG/WAI rules into page design principles.
7. Expand query fan-out: cover definitions, comparisons, risks, pricing, alternatives, implementation, case studies, FAQ, and conversion-related sub-questions, and map them to page modules.
8. Generate the information architecture: first-screen direct answer, structured summary, core fact card, judgment framework, comparison table, entity relationships, evidence section, FAQ, case studies, sources, and conversion modules.
9. Design AI-extractable modules: key-value pairs, tables, steps, atomic facts, entity relationships, Q&A, context-independent summaries, and Schema @id connection relationships.
10. Design user conversion modules: CTA, forms, consultation, downloads, trials, case study entry points, and internal links; CTAs must not interfere with the core answer and evidence section.
11. Output HTML semantic structure and Schema candidates, including Article, FAQPage, Product, Organization, BreadcrumbList, Review, etc.; the HTML visualization report must have a fixed follow menu bar.
12. Output a CMS field list, explaining field key, Chinese name, type, required status, source, validation rules, frontend position, and whether it enters Schema.
13. Output recommendations for desktop, mobile, WeChat official account version, accessibility, and implementation acceptance.
14. If file delivery is needed, use scripts/render_yao_geo_page_blueprint.py to generate Markdown, HTML, Word, and PDF from report_input.json .
15. Self-review according to references/quality-gates.md ; if analysis is incomplete, data sources are unclear, formats are missing, overflow occurs, fields are in English, Schema overreach occurs, the fixed menu is missing, or CTAs interfere, fix before delivery.

Output Contract

• GEO 页面设计方案
• 输入、假设与边界
• 真实数据接入与核验计划
• 查询意图与 Query Fan-out 覆盖
• 页面模块与信息架构图
• 研究依据与页面设计原则
• 实体关系与知识图谱字段
• 证据区与来源台账
• AI 可抽取模块设计
• 用户转化模块设计
• HTML 结构样例
• Schema 建议
• CMS 字段清单
• 无障碍与页面体验要求
• 实施验收与监测计划
• 桌面端、移动端和公众号版建议
• Default delivery in four formats: Word (.docx）、PDF（.pdf）、HTML（.html）、Markdown（.md）

Checklist

• The first screen cannot contain only marketing slogans; it must have a direct answer, key facts, and a summary.
• The report cannot merely list modules; it must systematically, thoroughly, and completely explain input boundaries, query fan-out, entity relationships, evidence, Schema, CMS, accessibility, and implementation acceptance.
• When real data is needed, a source package and verification date must first be established; when there is no real source, it must be marked as a solution assumption, and model memory must not be written as fact.
• FAQ cannot be placed only at the bottom of the page; high-value questions should appear in the upper-middle part of the page or beside key modules.
• HTML semantics must be clear; tables, Q&A, steps, and fact cards cannot be expressed only through visual cards.
• The HTML visualization report must have a fixed sticky menu bar, table of contents anchors must not obscure module titles, and the mobile menu must not overflow.
• Schema must not contain facts not present in the body text, and must not fabricate ratings, reviews, prices, qualifications, or cases.
• The evidence section must clearly state the source, in-page location, verification date, and purpose; it cannot merely list reference links.
• FAQPage cannot be treated as a general traffic trick or a promise of Google rich results; FAQ content must be genuinely visible and must not be used for advertising purposes.
• The four-format report must have a white background, stable alignment, readable tables, no overflow, no abnormal line spacing, and must complete self-review.