lazyoaf.com — AI Site Grade

Analysis

Cloudflare JS Challenge Blocks All AI Crawlers From Every Page

Every AI crawler tested — GPTBot, ClaudeBot, PerplexityBot, Google-Extended, ChatGPT-User, OAI-SearchBot, Applebot-Extended, Bytespider, and anthropic-ai — receives a 403 response from Cloudflare's JS challenge wall on every URL tested, including the homepage, robots.txt, sitemap.xml, llms.txt, and all product/collection/pages paths. The site returns zero visible content to any non-JS-capable client.

Crawler Access

The site runs on Shopify (Shopify IP 23.227.38.32, Shopify-style URL structure with /collections/, /products/, /pages/ paths) behind Cloudflare with a managed JS challenge (cf-mitigated header present). The robots.txt and llms.txt endpoints return the same Cloudflare challenge wall as the homepage — no actual rules file is ever served. The sitemap.xml is equally blocked. The help subdomain (help.lazyoaf.com, hosted on Gorgias) is accessible to plain HTTP clients and contains 16 FAQ articles, 5 shipping articles, and 5 returns articles, but this subdomain is not linked from any discoverable robots.txt or sitemap.

Cold-Knowledge Gap

The LLM model knows Lazy Oaf as a London-based streetwear brand founded in 2001 by Gemma Shiel, with a flagship Soho store, collaborations with Disney/Hello Kitty/Coca-Cola, inclusive sizing up to UK 22, and a "weirdo" aesthetic. The actual site content behind the Cloudflare wall cannot confirm or contradict any of this. The help subdomain confirms the brand voice ("Oaf Support", "V.I.O — very important Oaf", "LO-yalty scheme") and that the brand describes itself as "a small and independent business." The model's knowledge of the brand's founding story, collaborations, and flagship store location is entirely derived from off-domain sources — the site itself contributes zero structured data or text to AI training pipelines.

Schema Posture

No JSON-LD schema of any type was found on any accessible page. The homepage meta description ("Lazy Oaf create irreverent streetwear clothing for men and women, inspired by youth nostalgia, teenage rebellion and a spirit of non-conformity") is visible only via the Wayback Machine snapshot. The help subdomain pages contain no schema markup either — no FAQPage, Organization, WebSite, or Product schema exists anywhere on the domain.

External Signals

The help subdomain references social media accounts (Twitter/X, Instagram, Facebook) and a press contact ([email protected]). The brand operates a loyalty program (LO-yalty), a mobile app (iOS/Android), and offers Klarna/Clearpay payment options. The returns policy states "as a small and independent business we're unable to offer free returns." No external review sites, press mentions, or Reddit threads were surfaced through search — the brand's off-domain footprint appears fragmented and difficult for AI engines to triangulate.

Findings

1. Cloudflare JS challenge blocks all AI crawlers from every page High

   Every AI crawler tested receives a 403 response from Cloudflare's JS challenge wall on all URLs, including homepage, robots.txt, sitemap.xml, and product pages. The site returns zero visible content to non-JS-capable clients.

   What to change: Configure Cloudflare to allow AI crawler user agents (e.g., GPTBot, ClaudeBot) through the JS challenge, or serve static HTML versions of key pages to bots.

2. Robots.txt and sitemap.xml are blocked by Cloudflare challenge High

   The robots.txt and sitemap.xml endpoints return the same Cloudflare challenge wall as the homepage, so no actual rules or sitemap are ever served to crawlers.

   What to change: Ensure robots.txt and sitemap.xml are served without JS challenge, or host them on a separate domain that is accessible to crawlers.

3. No llms.txt file available Medium

   The llms.txt endpoint returns a 403 error, meaning no AI-specific content guidance is provided.

   What to change: Create an llms.txt file with a summary of the brand and links to key pages, and serve it without JS challenge.

4. No JSON-LD schema on any accessible page High

   No JSON-LD schema of any type was found on the homepage, help subdomain, or any other accessible page. Missing Organization, WebSite, Product, and FAQPage schema.

   What to change: Add JSON-LD structured data for Organization, WebSite, Product, and FAQPage on relevant pages.

5. Help subdomain not linked from main site's robots.txt or sitemap Medium

   The help subdomain (help.lazyoaf.com) is accessible to plain HTTP clients but is not referenced in any discoverable robots.txt or sitemap, making it hard for crawlers to find.

   What to change: Add the help subdomain to the main site's sitemap and robots.txt, or include links from the main site.

6. No external search results or reviews surfaced for the brand Medium

   Multiple web searches for the brand name, domain, and reviews returned zero results, indicating a fragmented off-domain footprint that is difficult for AI engines to triangulate.

   What to change: Encourage customer reviews on third-party platforms and ensure the brand is listed on relevant directories and social media.

7. Site content invisible to AI training pipelines High

   The Cloudflare challenge prevents AI crawlers from accessing any page content, so the site contributes zero structured data or text to AI training pipelines. LLM knowledge of the brand comes entirely from off-domain sources.

   What to change: Allow AI crawlers through the Cloudflare challenge to let the site's content be indexed and used for AI training.

8. No structured data on help subdomain pages Medium

   The help subdomain pages (FAQs, shipping, returns) contain no FAQPage or other schema markup, missing an opportunity to provide direct answers in AI search results.

   What to change: Add FAQPage schema to FAQ articles and Organization schema to the help subdomain.

What's working

• Help subdomain accessible to plain HTTP clients — The help subdomain (help.lazyoaf.com) is not behind the Cloudflare challenge and serves content to plain HTTP clients, including FAQ, shipping, and returns articles.
• Brand voice consistent on help subdomain — The help subdomain uses a consistent brand voice ('Oaf Support', 'V.I.O', 'LO-yalty scheme') and confirms the brand describes itself as a small independent business.
• Wayback Machine snapshot available for homepage — A Wayback Machine snapshot from 2026 shows the homepage meta description, providing some historical content visibility.