analoguewonderland.co.uk — AI Site Grade

Analysis

I have enough data now. Let me write the audit.

Analogue Wonderland has a sophisticated AI-agent commerce layer (/llms.txt, UCP/MCP endpoints, Shopify agent instructions) that is entirely unreachable by the AI crawlers it was designed for — Cloudflare's JS challenge wall blocks GPTBot, ClaudeBot, PerplexityBot, and even Google-Extended from accessing every page beyond the homepage shell, including the sitemap, product pages, and the agent-facing endpoints themselves.

Crawler Access

The robots.txt allows all AI crawlers (User-agent: * Allow: /) with no AI-specific disallow rules, and the site even publishes an llms.txt with detailed agent instructions, UCP discovery endpoints, and a Shopify skill recommendation. Yet in practice, every AI bot tested returns a Cloudflare JS challenge (403 with "Verifying your connection...") on the homepage, product pages, collections, sitemap sub-files, and the /agents.md and /.well-known/ucp endpoints referenced in the instructions. The first compare_bot_access run showed GPTBot, ClaudeBot, and PerplexityBot getting 200 with full content (~936KB), but subsequent runs — including individual fetch_as_bot calls — returned consistent 403 blocks across all bots. This intermittent behaviour suggests Cloudflare's bot management is rate-thresholding or session-dependent, not applying a stable allowlist for known AI crawler IP ranges. The site runs on Shopify behind Cloudflare (IP 23.227.38.65), with Cloudflare's "Managed Challenge" mode active.

Cold-Knowledge Gap

The LLM knows Analogue Wonderland as a film subscription box service ("WonderBox") founded in 2018 by Paul McKay, with a "Film Swap" initiative and a popular YouTube channel. The actual site has no mention of a subscription box, no "WonderBox" product, no "Film Swap" programme, and no prominent founder story. The homepage is a straightforward e-commerce catalogue selling individual film rolls, cameras, and lab development services. The subscription model that defined the brand's early identity has been either retired or deeply buried. The YouTube channel exists (linked in footer) but is not featured as a content hub. The LLM's prior is stale by at least one major product pivot — AI engines describing the brand will cite a subscription model the site no longer promotes.

Schema Posture

The homepage carries only two schema types: BreadcrumbList (single-item, pointing only to Home) and WebSite with SearchAction. There is no Organization schema (no logo, social profiles, founding date, or founder), no Product schema on the homepage product listings, no FAQPage schema despite FAQ-style content ("New to Film Photography?"), and no LocalBusiness schema. The WebSite schema's url and SearchAction target use the canonical domain correctly, but the BreadcrumbList is functionally empty. Product pages could not be fetched for schema inspection due to the Cloudflare wall.

External Signals

The homepage prominently displays "Thousands of 5* reviews" and links to Trustpilot, but the Trustpilot page was also blocked by Cloudflare during testing. The site references a Club AW loyalty programme, a 2026 Big Film PhotoWalk event, and a PURE35 film camera exclusive to the UK. Social links include Instagram, YouTube, TikTok, Facebook, and LinkedIn. The sitemap index references a sitemap_agentic_discovery.xml sub-sitemap — a Shopify-native pattern for AI crawler discovery — but this file is also behind the Cloudflare challenge wall, defeating its purpose.

Findings

1. Cloudflare JS challenge blocks all major AI crawlers High

   Every tested AI bot (GPTBot, ClaudeBot, PerplexityBot, Google-Extended) receives a 403 Cloudflare JS challenge on the homepage, product pages, collections, sitemap sub-files, and agent-facing endpoints. The robots.txt allows all crawlers, but Cloudflare's Managed Challenge mode prevents access.

   What to change: Configure Cloudflare to allowlist known AI crawler IP ranges or disable JS challenge for bot user agents. Alternatively, serve a static HTML version of key pages to bots.

2. llms.txt and agent endpoints are blocked by Cloudflare High

   The site publishes an llms.txt with detailed agent instructions, UCP discovery endpoints, and a Shopify skill recommendation, but these endpoints return 403 for AI bots, making the agent layer invisible.

   What to change: Ensure llms.txt and agent endpoints are accessible to AI crawlers by adjusting Cloudflare settings or serving them from a separate domain without JS challenge.

3. LLM knowledge describes a retired subscription model High

   The LLM knows Analogue Wonderland as a film subscription box service ('WonderBox') with a 'Film Swap' programme, but the current site is a straightforward e-commerce catalogue with no mention of subscriptions. This mismatch causes AI engines to describe a product the site no longer offers.

   What to change: Update the site's content to clearly reflect the current product offering, or add structured data and a prominent page describing the subscription if it still exists. Consider publishing a press release or blog post to signal the pivot to search engines.

4. No Organization schema on homepage Medium

   The homepage lacks Organization schema, missing logo, social profiles, founding date, and founder information. This reduces the site's ability to establish brand identity in AI knowledge graphs.

   What to change: Add Organization schema with logo, social media URLs, founding date, and founder name to the homepage.

5. BreadcrumbList schema is functionally empty Low

   The homepage's BreadcrumbList schema contains only a single item pointing to Home, providing no navigational context for AI crawlers.

   What to change: Expand the BreadcrumbList to include the actual page hierarchy, or remove it if it cannot be populated meaningfully.

6. No Product schema on homepage product listings Medium

   The homepage displays product listings but lacks Product schema, which would help AI crawlers understand the catalog and surface product information in search results.

   What to change: Add Product schema to each product listing on the homepage and collection pages.

7. No FAQPage schema despite FAQ-style content Low

   The homepage includes a 'New to Film Photography?' section with FAQ-style content, but no FAQPage schema is present, missing an opportunity for rich results and AI training data.

   What to change: Add FAQPage schema to the 'New to Film Photography?' section.

8. Agentic discovery sitemap is blocked by Cloudflare High

   The sitemap index references a sitemap_agentic_discovery.xml sub-sitemap, a Shopify-native pattern for AI crawler discovery, but this file returns 403 for all bots, defeating its purpose.

   What to change: Ensure the agentic discovery sitemap is accessible to AI crawlers by adjusting Cloudflare settings.

9. Trustpilot review page blocked by Cloudflare Medium

   The site links to a Trustpilot page for reviews, but that page is also blocked by Cloudflare, preventing AI crawlers from verifying the claimed 'Thousands of 5* reviews'.

   What to change: Ensure the Trustpilot page is accessible to AI crawlers, or embed review schema on the site itself.

10. No LocalBusiness schema for physical location Low

    The site does not include LocalBusiness schema, which would help AI crawlers understand the business's physical presence and service area.

    What to change: Add LocalBusiness schema with address and service area if applicable.

What's working

• llms.txt published with detailed agent instructions — The site publishes an llms.txt file with structured agent instructions, UCP endpoints, and a Shopify skill recommendation, demonstrating a sophisticated understanding of AI-agent commerce.
• robots.txt allows all AI crawlers — The robots.txt file has no disallow rules for AI bots, indicating an intention to be crawled.
• WebSite schema with SearchAction is correctly implemented — The homepage includes WebSite schema with a valid SearchAction target, helping AI crawlers understand the site's search functionality.
• Social media links to multiple platforms are present — The homepage footer includes links to Instagram, YouTube, TikTok, Facebook, and LinkedIn, providing signals for brand presence across platforms.
• Sitemap index is published with multiple sub-sitemaps — The sitemap index references sub-sitemaps for pages, blogs, and agentic discovery, indicating a structured approach to content organization.
• YouTube channel linked in footer — The site links to a YouTube channel, which could serve as a content hub for AI training if properly utilized.