lyleandscott.com — AI Site Grade

Analysis

The bare domain (lyleandscott.com) is completely inaccessible to all AI crawlers — and every browser — while the www subdomain is fully open, creating a canonical split that fragments AI understanding of the brand.

Crawler Access

The site has a two-domain split that undermines AI visibility. The bare domain lyleandscott.com returns a 403 Cloudflare JS challenge wall to every user-agent tested — GPTBot, ClaudeBot, PerplexityBot, Google-Extended, OAI-SearchBot, and even a standard browser — all blocked with the same "Verifying your connection" page. Meanwhile, www.lyleandscott.com returns 200 with full content to every bot without exception. The robots.txt (served only on the www subdomain) contains no AI-specific directives — no Disallow for GPTBot, ClaudeBot, or any AI crawler — but the bare domain's challenge wall effectively blocks all crawlers anyway. The llms.txt exists and is well-structured, describing the Universal Commerce Protocol (UCP) endpoints, agent shopping flows, and read-only browsing paths. An agents.md file mirrors this content. A .well-known/ucp endpoint returns a full JSON merchant profile with supported UCP versions, service endpoints, and Google Pay / Shopify card payment handlers — an unusually sophisticated agent-commerce setup for a heritage fashion brand.

Cold-Knowledge Gap

The LLM knows Lyle & Scott as a British heritage knitwear brand founded in 1874, famous for cashmere sweaters and the Golden Eagle logo, with roots in golfwear and the 1980s UK casual subculture. The site's actual homepage, however, leads with a "Curated by Wavey Garms" collaboration and a Heat Reactive Jacket — a tech-forward piece that shifts color with temperature. The brand history page confirms the 1874 Hawick origins and the golden eagle story, but the homepage's emphasis on streetwear collaborations (Wavey Garms, Fly Nowhere, FWA, Harry Hall) and football "Kits for Clubs" campaigns is underrepresented in the model's prior knowledge. The model knows the brand as quiet and heritage-focused; the site presents a brand actively pushing into streetwear, football culture, and performance outerwear.

Schema Posture

Product pages carry solid Product schema with AggregateRating (4.9 from 11 reviews on the sampled jumper), multiple Offer entries with SKU-level pricing, and InStock availability. The homepage carries Organization and WebSite schema with SearchAction. However, no BreadcrumbList, FAQPage, or Article schema appears anywhere — the blog posts and the extensive brand history page (4,844 words of narrative) lack structured markup entirely. The Organization schema's sameAs links point to Facebook, Instagram, and X (Twitter) but omit TikTok and YouTube, both of which are linked in the page footer.

External Signals

DuckDuckGo web search returned zero results for queries including "lyleandscott.com", "Lyle and Scott brand", and "Lyle and Scott review" — an unusual vacuum that suggests the site has limited external citation footprint in the search index. The DNS records reveal a Shopify-hosted site (IP 23.227.38.65) with Klaviyo email marketing, Hornetsecurity for mail, and multiple Google/Apple/Facebook domain verification TXT records. The sitemap is extensive (1,364+ product URLs across multiple locale sub-sitemaps for en, de, nl, sv, fr-BE, nl-BE, de-BE), indicating a mature ecommerce operation with international ambitions.

Findings

1. Bare domain blocks all AI crawlers with Cloudflare challenge wall High

   The bare domain lyleandscott.com returns a 403 Cloudflare JS challenge wall to every user-agent tested, including GPTBot, ClaudeBot, PerplexityBot, and standard browsers, while www.lyleandscott.com is fully accessible. This creates a canonical split that fragments AI understanding of the brand.

   What to change: Configure the bare domain to redirect to the www subdomain, or remove the Cloudflare challenge wall for known AI crawler user-agents on the bare domain.

2. No AI-specific directives in robots.txt Medium

   The robots.txt on www.lyleandscott.com contains no Disallow rules for GPTBot, ClaudeBot, or any AI crawler, leaving AI access unmanaged.

   What to change: Add explicit Disallow or Allow rules for AI crawlers in robots.txt to control AI access.

3. Cold-knowledge gap: heritage brand perception vs. streetwear-focused site Medium

   LLM prior knowledge describes Lyle & Scott as a quiet heritage knitwear brand, but the site leads with streetwear collaborations (Wavey Garms, Fly Nowhere) and a Heat Reactive Jacket. This mismatch reduces AI accuracy in representing the brand.

   What to change: Ensure the homepage and key landing pages include clear, structured content that explicitly states the brand's heritage alongside its modern collaborations.

4. Missing BreadcrumbList schema on product and category pages Medium

   No BreadcrumbList structured data was found on any page, which helps AI understand site hierarchy and context.

   What to change: Add BreadcrumbList schema to all product, collection, and category pages.

5. Missing FAQPage and Article schema on blog and brand history pages Medium

   The blog and the extensive brand history page (4,844 words) lack any structured markup, reducing their discoverability and utility for AI.

   What to change: Add Article schema to blog posts and FAQPage schema to any FAQ content; consider adding Article or WebPage schema to the brand history page.

6. Organization schema sameAs omits TikTok and YouTube Low

   The Organization schema on the homepage includes sameAs links for Facebook, Instagram, and X (Twitter) but omits TikTok and YouTube, both linked in the page footer.

   What to change: Add TikTok and YouTube URLs to the sameAs array in the Organization schema.

7. Zero external search results for brand queries Medium

   DuckDuckGo web search returned zero results for queries including 'lyleandscott.com', 'Lyle and Scott brand', and 'Lyle and Scott review', indicating limited external citation footprint.

   What to change: Investigate and improve off-site SEO and brand mentions to increase the site's citation footprint.

8. llms.txt not served on bare domain Low

   The llms.txt file is only available on the www subdomain; the bare domain returns a 403, preventing AI from discovering it.

   What to change: Ensure llms.txt is accessible on the bare domain, or redirect the bare domain to www.

What's working

• Well-structured llms.txt with UCP endpoints and agent flows — The llms.txt file on the www subdomain is well-structured, describing Universal Commerce Protocol endpoints, agent shopping flows, and read-only browsing paths, enabling AI agents to interact with the site programmatically.
• Universal Commerce Protocol endpoint available at .well-known/ucp — The .well-known/ucp endpoint returns a full JSON merchant profile with supported UCP versions, service endpoints, and payment handlers (Google Pay, Shopify card), enabling sophisticated agent-commerce integration.
• Product pages include Product schema with AggregateRating and multiple Offers — Product pages carry solid Product schema with AggregateRating (4.9 from 11 reviews), multiple Offer entries with SKU-level pricing, and InStock availability, providing rich data for AI.
• Homepage includes Organization and WebSite schema with SearchAction — The homepage carries Organization and WebSite schema with SearchAction, providing basic brand and site information to AI.
• Comprehensive sitemap with 1,364+ product URLs across multiple locales — The sitemap is extensive, covering product URLs in multiple locales (en, de, nl, sv, fr-BE, nl-BE, de-BE), indicating a mature ecommerce operation with international reach.
• agents.md file mirrors llms.txt content — An agents.md file is present and mirrors the llms.txt content, providing an alternative discovery path for AI agents.
• www subdomain fully accessible to all AI crawlers — The www.lyleandscott.com subdomain returns 200 with full content to all 11 tested bots without any blocks, ensuring AI can access the site's content.