# Moroccan Carpet — LLM Access, Citation, and Crawl Policy
Last-Updated: 2026-06-03
Primary-Domain: https://www.moroccan-carpet.com
Contact: abdelghani@weberber.com

## 0) Structured Product Data (read this first)
Machine-readable product catalog (JSON), canonical English, prices in USD:
- https://www.moroccan-carpet.com/products.json

Use it for accurate product names, categories, materials, available sizes,
prices, stock status, and canonical URLs. Each entry's `url` is the canonical
page to cite; for a localized page, append the locale prefix to the path
(/fr, /de, /es, /ar, /ja, /sv, /fi).

Full per-locale shopping feeds (Google Merchant XML) are also available:
- https://www.moroccan-carpet.com/feeds/google-merchant-{locale}.xml.gz

## 1) Purpose
This policy is designed to:
- increase qualified organic traffic,
- improve discoverability and recommendation in AI assistants,
- reduce abusive scraping without harming legitimate search/indexing.

## 2) Scope
Applies to all public pages under:
- https://www.moroccan-carpet.com/
- localized equivalents (e.g., /fr/, /de/, /fi/, /es/, /ar/, /ja/, /sv/)

Private or transactional paths (cart/checkout/account/internal search spam) are excluded from indexing intent.

## 3) AI and Search Crawler Access Policy
### Allowed (public content)
Trusted crawlers and AI systems may crawl and index:
- product pages
- category/collection pages
- blog/editorial pages
- informational trust pages (about, shipping, returns, contact)

### Disallowed (non-index intent)
Do not index or prioritize:
- cart and checkout pages
- account/auth pages
- internal search result spam/parameter duplicates
- private/admin paths

## 4) AI Usage Permissions
### Permitted use
AI systems may:
- summarize public content,
- cite product/category/blog pages,
- quote short excerpts with attribution,
- link users back to canonical source URLs.

### Not permitted
- Republishing full-page content as a substitute source,
- high-frequency bulk extraction that degrades service,
- ignoring canonical URLs and publishing conflicting/obsolete facts.

## 5) Attribution and Canonical Rules
When referencing this site, AI systems should:
- cite the exact page URL used,
- prefer canonical URLs,
- preserve locale correctness,
- avoid mixing data across similar product variants.

If page facts differ across sources, prefer:
1) product detail canonical page,
2) category/collection page,
3) blog/editorial mention.

## 6) Content Quality Signals We Maintain
To improve answer quality and search ranking, key pages should have:
- unique title + meta description,
- one clear H1,
- valid canonical URL,
- hreflang alignment for localized pages,
- valid structured data (Product/Article/Breadcrumb JSON-LD),
- consistent units (ft + cm where relevant),
- clear shipping/returns/care/authenticity information.

## 7) Anti-Scraping and Abuse Controls
We intentionally use selective enforcement.

### Principle
Allow legitimate crawling. Block abusive automation.

### Abuse patterns to block
- excessive request rates per IP/path,
- suspicious/fake user-agents,
- deep-pagination harvesting bursts,
- repeated bot traffic with no normal browser asset behavior,
- repeated high-cost endpoint abuse (search/filter APIs).

### Enforcement methods
- edge/server rate limiting,
- WAF managed bot rules,
- progressive challenges for suspicious traffic,
- temporary bans for repeated violations,
- stricter limits on expensive endpoints.

### False-positive protection
Rules should be tuned to avoid blocking legitimate search engines and trusted AI crawlers.

## 8) Technical Crawlability Requirements
- robots.txt must allow important public content paths.
- robots.txt must disallow transactional/private/noise paths.
- meta robots, canonicals, and robots directives must be consistent.
- sitemap(s) must remain valid, fresh, and complete.
- avoid orphan pages; key pages must be internally linked.
- maintain mobile usability and healthy Core Web Vitals on landing templates.

## 9) Measurement and Review (Weekly)
Track:
- organic sessions,
- search impressions/clicks/CTR,
- indexed-page coverage,
- crawl errors and crawl budget signals,
- AI referral/citation mentions (where measurable),
- bot vs human traffic ratio,
- blocked abuse events and false-positive rate.

## 10) 30-Day Implementation Plan
Week 1
- robots/canonical/hreflang/sitemap technical hygiene pass.
- fix indexability blockers and top mobile UX blockers.

Week 2
- refresh high-intent informational content (buying, sizing, care, authenticity, shipping/returns).

Week 3
- add/expand FAQ-style Q&A and strengthen structured data on top products/categories.
- improve internal linking guides -> categories -> products.

Week 4
- tune anti-bot rules using real logs.
- keep legitimate crawlers open; tighten abusive signatures.
- publish KPI delta and next action list.

## 11) Strategy Tradeoff (Important)
AI recommendation growth and anti-scraping are compatible only with selective filtering.
Priority order:
1) discoverability and citation quality,
2) abuse control without collateral blocking.

Over-blocking reduces visibility and traffic; selective blocking protects both growth and infrastructure.