jaratrading.co.uk — AI Site Grade

Analysis

The live site is invisible to every AI crawler — and to every browser too

The entire domain jaratrading.co.uk sits behind a Cloudflare JS challenge wall that returns HTTP 403 to all traffic, regardless of user-agent. Every AI crawler tested — GPTBot, ClaudeBot, PerplexityBot, Google-Extended, OAI-SearchBot, Bytespider, Applebot-Extended, anthropic-ai, ChatGPT-User, Perplexity-User — receives the same "Verifying your connection..." JS shell with zero visible content. Even a standard browser UA gets blocked. The site has no robots.txt accessible (returns 403), no llms.txt (returns 403), and no sitemap.xml (returns 403). The only content available is via the Wayback Machine, whose most recent snapshot (January 2026) shows the site as it existed before the Cloudflare wall went up.

Cold-Knowledge Gap

A frontier LLM queried cold about "Jara Trading UK" returns no verifiable information — the model cannot confirm what the company sells, what brands it owns, or what industry it operates in. This is a complete knowledge vacuum. The site itself, per the archived snapshot, describes a London-based ecommerce private-label aggregator founded in 2018 by co-founders Raph and Jake (Raphael Englard and Jake Balkin), operating three consumer brands — ebake.co.uk (baking), greenthumbz.co.uk (gardening), and volila.co.uk — selling across Amazon, eBay, and B&Q. The gap between the LLM's blank slate and the site's actual positioning as a multi-brand ecommerce operator is total.

Schema Posture

The archived homepage and all subpages (Privacy Policy, Cookie Policy, Terms & Conditions, Equality Statement) contain zero JSON-LD structured data of any type. No Organization, WebSite, Product, FAQPage, or BreadcrumbList schema exists. The site uses a single-page layout with anchor-based navigation (#about, #brands, #network, #team, #careers, #buy, #Contact), meaning each section lacks its own URL, title tag, or meta description — making them invisible to search engines and AI crawlers even if the Cloudflare wall were removed.

External Signals

The brand has zero detectable external footprint. Searches for "Jara Trading" across web, reviews, LinkedIn, Companies House, and news return no results. The three sub-brands (ebake.co.uk, greenthumbz.co.uk, volila.co.uk) are also behind identical Cloudflare JS walls. The only external traces are Amazon product reviews quoted on the archived homepage (for what appear to be kitchen/garden products) and links to Amazon UK storefronts, eBay stores, and a B&Q product page. The registered office listed in the privacy policy (Churchill House, 137-139 Brent Street, London NW4 4DJ) and the address in the cookie policy (Athene House, 86 The Broadway, London NW7 3TD) differ, suggesting a change of registered address that is inconsistently reflected across pages.

Infrastructure Contradiction

DNS resolves to 23.227.38.65 (Cloudflare) with DigitalOcean nameservers, and the Cloudflare cf-ray headers confirm active WAF/JS challenge enforcement. The site claims to be "Jara Trading LTD, a company incorporated in England and Wales" but no Companies House registration number appears anywhere on the site. The copyright footer reads "Jara Trading 2023" — the most recent snapshot is from January 2026, suggesting the site may be unmaintained. The Cloudflare wall blocks all programmatic access, meaning AI engines cannot retrieve even the basic company description that the Wayback Machine preserves.

Findings

1. Cloudflare JS challenge wall blocks all AI crawlers and browsers High

   The entire domain returns HTTP 403 with a JS challenge to every user-agent, including all major AI crawlers (GPTBot, ClaudeBot, PerplexityBot, etc.) and standard browsers. No content is accessible programmatically.

   What to change: Remove the JS challenge for known AI crawler user-agents or serve static HTML content to bots. Allow GPTBot, ClaudeBot, and other AI crawlers through the WAF.

2. Robots.txt, sitemap, and llms.txt all return 403 High

   The robots.txt, sitemap.xml, and llms.txt files are all blocked by the Cloudflare wall, returning 403 errors. AI crawlers cannot discover allowed paths or site structure.

   What to change: Serve a robots.txt that allows AI crawlers and a sitemap.xml listing all pages. Create an llms.txt file with a company description and key URLs.

3. No JSON-LD structured data on any archived page High

   The archived homepage and all subpages (Privacy Policy, Cookie Policy, Terms & Conditions, Equality Statement) contain zero JSON-LD structured data. No Organization, WebSite, Product, or other schema types are present.

   What to change: Add Organization, WebSite, and Product JSON-LD schemas to all pages. Include company name, description, founding date, founder names, and brand URLs.

4. Single-page layout with anchor navigation hides sections from crawlers High

   The site uses a single-page design where sections like #about, #brands, #team, and #buy are only accessible via anchor links. Each section lacks its own URL, title tag, or meta description, making them invisible to search engines and AI crawlers.

   What to change: Convert each major section into a separate, crawlable page with unique URLs, title tags, and meta descriptions. Use internal linking between pages.

5. Frontier LLM has no verifiable information about Jara Trading High

   A cold query to a frontier LLM about 'Jara Trading UK' returns no verifiable information. The model cannot confirm the company's products, brands, or industry, despite the archived site describing a multi-brand ecommerce aggregator.

   What to change: Publish a company Wikipedia page or other authoritative sources. Ensure the live site is crawlable and contains structured data to feed AI knowledge bases.

6. Zero external footprint across web, reviews, social, and company registries High

   Searches for 'Jara Trading' across web, reviews, LinkedIn, Companies House, and news return no results. The three sub-brands (ebake.co.uk, greenthumbz.co.uk, volila.co.uk) are also blocked by Cloudflare. No external signals exist to validate the company's existence.

   What to change: Create and maintain profiles on LinkedIn, Crunchbase, and other business directories. Encourage customer reviews on Trustpilot or Google. Ensure sub-brand sites are crawlable.

7. Inconsistent registered addresses across privacy and cookie policies Medium

   The Privacy Policy lists 'Churchill House, 137-139 Brent Street, London NW4 4DJ' while the Cookie Policy lists 'Athene House, 86 The Broadway, London NW7 3TD'. This inconsistency may confuse crawlers and users.

   What to change: Standardize the registered address across all legal pages. Update the Privacy Policy to reflect the current address.

8. No Companies House registration number displayed on site Medium

   The site claims incorporation in England and Wales but does not display a Companies House registration number, which is a legal requirement for UK companies and reduces trustworthiness for AI crawlers.

   What to change: Add the Companies House registration number to the footer and legal pages.

9. Copyright footer shows 2023 while snapshot is from 2026 Low

   The copyright footer reads 'Jara Trading 2023' but the most recent Wayback Machine snapshot is from January 2026, suggesting the site may be unmaintained. This can signal to AI crawlers that the site is stale.

   What to change: Update the copyright year to the current year or use a dynamic year.

What's working

• Wayback Machine snapshot preserves company description and brand details — The archived snapshot from January 2026 contains a full company description, brand listings, and team information that can be used to reconstruct the site's content for AI crawlers once access is restored.
• Privacy, cookie, and terms pages exist and are well-structured — The archived site includes Privacy Policy, Cookie Policy, Terms & Conditions, and Equality Statement pages with substantial text content, providing a foundation for compliance and trust signals.
• DNS resolves correctly and Cloudflare is properly configured — The domain resolves to a valid Cloudflare IP (23.227.38.65) with proper DNS records, indicating the infrastructure is functional even if access is restricted.