cbwebsitedesign.co.uk — AI Site Grade

Analysis

Cold-Knowledge Gap

A frontier LLM queried cold about CreativeWeb / cbwebsitedesign.co.uk returned zero substantive knowledge — it could not confirm services, clients, reputation, or even that the agency exists. This is the central finding. The site itself claims to be the No.1 ranked UK web design agency on Clutch out of 7,000+ providers, with a 4.9 rating from 68 reviews, over 500 bespoke websites built since 2013, and high-profile clients including the NATO Innovation Fund (a EUR 1B+ venture capital fund backed by 24 NATO allies). The gap between the brand's actual market standing and what AI models know about it cold is near-total.

Crawler Access

All 11 tested AI crawlers — GPTBot, OAI-SearchBot, ChatGPT-User, ClaudeBot, anthropic-ai, PerplexityBot, Perplexity-User, Google-Extended, Bytespider, Applebot-Extended — receive a 200 status with full content identical to browser baseline (427 KB). No UA-based blocking, no Cloudflare challenge, no JS shell. The robots.txt is a bare Yoast default (User-agent: * Disallow:) with no AI-bot-specific rules — no GPTBot, no ClaudeBot, no PerplexityBot directives at all. The site hosts an llms.txt (rare for a web agency), auto-generated by Yoast SEO v26.7, listing pages, posts, portfolio items, team members, and categories. The sitemap index contains 7 sub-sitemaps covering ~175+ URLs. The site runs on Cloudflare (DNS, CDN, WAF) with WordPress CMS.

Schema Posture

Every page carries rich, consistent JSON-LD: Organization with logo, address, telephone, sameAs (Facebook, X, LinkedIn, Instagram, Clutch), and an AggregateRating (4.9, 68 reviews). The homepage also embeds LocalBusiness nodes for both London and Guildford offices, plus a VideoObject and BreadcrumbList. The website-design page adds a Product schema with the same aggregate rating. The @id references are consistent across all pages. This is a mature, well-maintained schema implementation — one of the strongest seen for a UK agency of this size.

Content & AI-Relevance Signals

The site has a dedicated GEO (Generative Engine Optimisation) service page and a /geo-ai/ blog category with three substantial articles published November–December 2025, including a 2,458-word guide "How to Rank on ChatGPT and Bing Copilot in 2026" that explicitly advises opening OAI-SearchBot and GPTBot in robots.txt — advice the site itself does not follow. The homepage contains a FAQ section and a website cost calculator. Blog content is fresh (most recent post February 2026), with multiple authors and named bylines. The site uses question-led H2s, answer blocks, and FAQ schema signals — all patterns AI engines prefer.

External Signals

The DNS TXT records show verification tokens for Google, Facebook, Apple, Pinterest, Trustpilot, and HubSpot — indicating active presence across those platforms. The schema references a Clutch profile (clutch.co/profile/creativeweb) which returns a 403 JS-wall when fetched directly. No external review or press mentions surfaced in web searches, suggesting limited off-domain citation footprint — a likely contributor to the cold-knowledge gap. The legal entity is Creative Brand Design Ltd (registered in England No.08757535), trading as CreativeWeb, with the domain cbwebsitedesign.co.uk as the primary web address.

Findings

1. Zero cold AI knowledge despite claimed market leadership High

   A frontier LLM queried cold about CreativeWeb returned no substantive information — it could not confirm services, clients, reputation, or the agency's existence. This contradicts the site's claim of being the No.1 ranked UK web design agency on Clutch with a 4.9 rating from 68 reviews.

   What to change: Build a stronger off-domain citation footprint by earning mentions on reputable tech and business publications, securing backlinks from high-authority domains, and actively managing profiles on platforms like Clutch, Trustpilot, and Google Business Profile.

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

   The robots.txt file contains only a bare Yoast default with no rules for GPTBot, ClaudeBot, PerplexityBot, or other AI crawlers. While this does not block them, it misses the opportunity to signal explicit permission and may cause some crawlers to be cautious.

   What to change: Add explicit Allow directives for GPTBot, OAI-SearchBot, ChatGPT-User, ClaudeBot, anthropic-ai, PerplexityBot, Perplexity-User, Google-Extended, Bytespider, and Applebot-Extended in robots.txt.

3. GEO blog advises opening AI bots but site does not follow Medium

   A 2,458-word guide on ranking in ChatGPT and Bing Copilot explicitly advises opening OAI-SearchBot and GPTBot in robots.txt, yet the site's own robots.txt does not include these directives.

   What to change: Update robots.txt to include Allow directives for OAI-SearchBot and GPTBot, aligning the site's configuration with the advice published on the site.

4. Limited off-domain citation footprint High

   Web searches for the agency on Clutch, reviews, and press mentions returned zero results. The Clutch profile referenced in schema returns a 403 JS-wall. This lack of external citations is a likely contributor to the cold-knowledge gap.

   What to change: Earn backlinks from reputable industry publications, secure press coverage, and ensure the Clutch profile is publicly accessible without a JS wall.

5. Clutch profile returns 403 JS wall Medium

   The Clutch profile referenced in the site's schema (clutch.co/profile/creativeweb) returns a 403 status with a JavaScript challenge, making it inaccessible to AI crawlers and reducing the credibility of the aggregate rating claim.

   What to change: Contact Clutch to ensure the profile is publicly accessible without a JS wall, or provide an alternative verifiable source for the rating.

6. llms.txt is present but not optimized for AI agents Low

   The site hosts an llms.txt auto-generated by Yoast SEO, which is rare and positive, but it lists all pages generically rather than curating the most important content for AI consumption.

   What to change: Manually curate the llms.txt to highlight key pages (services, portfolio, about, contact) and remove low-value entries to improve AI agent efficiency.

What's working

• All 11 tested AI crawlers receive full content — Every major AI crawler receives a 200 status with full HTML content identical to the browser baseline, with no UA-based blocking or JS shells.
• Mature, consistent JSON-LD schema across all pages — Every page carries Organization, LocalBusiness, AggregateRating, and BreadcrumbList schema with consistent @id references, plus Product schema on service pages. This is one of the strongest schema implementations seen for a UK agency.
• Dedicated GEO service page and blog category — The site has a GEO service page and a /geo-ai/ blog category with three substantial articles published in late 2025, including a detailed guide on ranking in ChatGPT and Bing Copilot.
• llms.txt file is published — The site hosts an llms.txt file (rare for a web agency), auto-generated by Yoast SEO, providing a structured summary of site content for AI agents.
• Fresh blog content with multiple authors — Blog content is regularly updated (most recent post February 2026) with named bylines and question-led H2s, FAQ sections, and answer blocks — patterns preferred by AI engines.
• Sitemap index with 7 sub-sitemaps covering 175+ URLs — The sitemap index contains 7 sub-sitemaps covering pages, posts, portfolio items, team members, and categories, ensuring comprehensive crawl coverage.
• DNS verification tokens for Google, Facebook, Apple, Pinterest, Trustpilot, and HubSpot — TXT records show verification tokens for multiple platforms, indicating active presence and management across key business directories.