Delivered Project 01 / SuperPDR
SuperPDR: Turning a Dent Repair Tool Catalog into Citable Knowledge
The project had extensive product information and industry experience, but key facts were scattered, product fields were inconsistent, and knowledge content, FAQs, and trust materials lacked clear connections. Zhihe Growth reorganized products, questions, and evidence into a crawlable, understandable, and traceable website knowledge system.
17%Test-Phase Website Citation Rate Compared with the project's initial 0%–3% range, the website began appearing more consistently as a source in relevant AI answers.
692 public product entries in a discoverable catalog
37 product categories organized by tool and intended use
28 + 16 28 long-form technical articles, 16 FAQ Hubs, and 2 Trust resources
17%Test-phase website citation rate within a fixed question scope
Website Snapshot (Public site verified on August 20, 2026)
The project's public website now connects its product catalog, technical knowledge, short Q&As, and evidence resources. The catalog provides entity coverage, while knowledge articles and FAQs address buying, use, and risk-assessment questions.
- The public product interface contains 692 product entries, distributed across 37 product categories.
- The Knowledge section contains 28 long-form technical articles, and the FAQ system contains 16 topic Hubs.
- Key topics include repairability assessment, tool selection, rod access, glue pulling, reflection-board interpretation, hail damage repair, and final quality inspection.
- References and the Update Log retain sources and update dates to support verification of high-risk facts.
Business and structural issues before optimization
- Despite a large catalog, product names, models, applications, kit contents, and specifications lacked a consistent factual basis.
- Product pages functioned mainly as listings and could not independently answer decision questions such as what can be repaired, which tools are needed, or when work should stop.
- Technical expertise was scattered across pages, while FAQs, references, and update records were not linked to specific product facts.
- AI could easily confuse kits, tool types, repair limits, and third-party descriptions.
Zhihe Growth's Implementation
- Established a single source of truth (SSOT) for 300+ core product fields and mapped it to visible page content, Product semantics, and relevant Schema.
- Organized 28 long-form technical articles and 16 FAQ Hubs around real questions, with distinct roles for in-depth explanations and short answers.
- Added assessment matrices, operating steps, stop conditions, counterexamples, and applicability limits, rather than describing product advantages alone.
- Connected product pages, knowledge pages, FAQs, References, and the Update Log with bidirectional internal links.
Why the Content Is Easier for AI to Use as a Source
- Question-led titles closely match buyers' wording, and opening paragraphs provide standalone, citable answers.
- Models, applications, kit contents, operating conditions, and risk limits can be read together on the page.
- Assessment matrices and stop conditions define clear limits, reducing overgeneralization in complex repair scenarios.
- References and update dates provide a traceable route for checking high-risk facts.
Example Explanation Chain: Can This Type of Dent Be Repaired?
The project separates assessment, explanation, verification, and next-step decisions into linked pages instead of expecting one product page to answer everything. Each page provides a self-contained answer, with links to fuller evidence.
01 Starting Question FAQs answer focused repairability questions and make clear that not every dent is suitable for paintless repair.
02 Assessment Framework Long-form technical articles explain assessment criteria such as paint condition, location, depth, metal stretch, and tool access.
03 Limits and Counterexamples Stop conditions include damaged paint, severe stretching, and obstructed access, preventing applicability statements from becoming absolute promises.
04 Product verification Consistent product fields explain tool applications, kit contents, and suitable scenarios to support tool selection.
05 Evidence Review References and the Update Log document sources, review status, and the latest update date.
| Delivery layer | Public Deliverables | Role in AI Answers |
|---|---|---|
| Product knowledge layer | 692 public product entries and 37 product categories; governance completed for 300+ core fields. | Broadens entity coverage while reducing the chance that AI confuses models, kit contents, or applications. |
| Content layer | 28 long-form technical articles, 16 FAQ Hubs, and 2 Trust resources covering repairability, tool selection, operation, and risk limits. | Provides distinct resources for technical explanations, short answers, and evidence verification. |
| Facts and Schema | More than 90% of high-priority facts were brought under governance; facts included in Schema were mapped from confirmed information. | Keeps visible content consistent with machine-readable fields. |
| Question Coverage | Content was built around 50 common industry questions, with direct-answer coverage reaching 65%–75%. | Turns product pages from catalog entries into starting points for answers. |
Result Limitations: The 17% figure is a retest result for this project's fixed question scope and recorded period. Platform retrieval methods, model versions, regions, and question wording affect answers. It is not a long-term guarantee or a promise of fixed rankings.