---
title: "Schema / JSON-LD checklist for GEO"
lang: "en"
canonical: "https://winin.ai/en/guides/schema-geo/"
alternate: "https://winin.ai/zh/guides/schema-geo/"
datePublished: "2026-09-12"
dateModified: "2026-09-12"
section: "guides"
---

# Schema / JSON-LD checklist for GEO

**Definition (site FAQ)**: structured data is machine-readable markup (such as Schema or JSON-LD) that declares what something is, its price, availability, etc. It makes explicit facts easier to recognise; whether and how AI or search uses it still needs observation.

## Recommended types (by page)
| Page type | JSON-LD suggestion |
|-----------|--------------------|
| Site-wide | Organization (name, url, email) |
| Product / facts | SoftwareApplication or Product |
| FAQ | FAQPage |
| Guide | HowTo (verifiable steps, no fake performance promises) |
| Comparison category | ItemList + Article |
| Glossary term | DefinedTerm |
| Case | Article + link to metrics disclaimer |
| Every page | BreadcrumbList; unique canonical |

## Organization minimum fields (Winin example semantics)
- name: Winin
- url: https://winin.ai
- email: contact@winin.ai

## Implementation steps
1. Choose the canonical URL; avoid multiple URLs for one product with differing JSON-LD.
2. Keep fields consistent with visible copy (especially price and availability).
3. Validate syntax with rich-result tools (does not guarantee AI citation).
4. Add to sitemap; keep robots open for crawlers you want to reach.
5. Watch with Edith-class monitoring whether answers improve — **Schema ≠ automatic shortlist**.

## What Schema is not
- Not black-hat keyword stuffing.
- Not a guarantee of recommendation share.
- Do not mark up unapproved or false certifications.

## With fact pages and the loop
Approve in Homer first → write fact pages and JSON-LD → Friday advances tasks → Edith retests. See [fact-pages](/en/guides/fact-pages/) · [geo loop](/en/facts/loop-find-govern-close-retest-operate/).

## FAQ
**Q: Is JSON-LD enough without content work?** A: No. GEO still depends on verifiable statements, comparison reasons and retesting.

**Q: What to watch with FAQPage?** A: Q&A must match visible content; avoid hidden keywords.

## Common consistency errors
1. JSON-LD price differs from the visible price.
2. Product and FAQ specs contradict each other.
3. Organization email or URL is wrong, conflicting with the contact fact page.
4. One SKU across several URLs declares different availability.

## Minimum launch pack
- Site-wide Organization
- Product/SoftwareApplication for core products
- FAQPage for key FAQs
- Article + BreadcrumbList for fact pages
- Sitemap inclusion and a crawl spot-check

**Source**: https://winin.ai FAQ + public technical practice · 2026-09-11