balderton.com — AI Site Grade

Analysis

Balderton.com — AI-Visibility Audit

ClaudeBot receives a 403 from Cloudflare on the homepage while every other major AI crawler — GPTBot, Google-Extended, PerplexityBot, OAI-SearchBot, anthropic-ai — gets a full 200 with identical content to a browser, making this one of the most selectively blocked VC sites in European venture.

Crawler Access

The robots.txt is minimal (47 bytes) with no AI-bot directives — only Disallow: /search/ and /?s= for the wildcard rule. The llms.txt exists and is substantial (275 KB), auto-generated by All in One SEO v4.9.6.2, containing a full index of posts and pages. The sitemap index is healthy with 13 sub-sitemaps covering posts, pages, team members, founders, resources, and categories. The site runs on Cloudflare with WordPress hosting via ExactDN CDN. ClaudeBot is the sole AI crawler blocked (403, 146 bytes) while anthropic-ai passes at 200 — a Cloudflare WAF rule targeting the ClaudeBot UA string specifically. All other bots return the full 97 KB HTML payload with no JS-rendering dependency. Security headers are absent (no HSTS, CSP, X-Frame-Options), which is unusual for a financial-services firm.

Cold-Knowledge Gap

The LLM prior knows Balderton as a London-based VC spun out of Benchmark Capital in 2007, focused on early-stage European tech, with portfolio highlights including Revolut, Darktrace, Citymapper, and Depop. It mentions a €1.3B fund raised in 2023 and the "Balderton Build" operational support team. The site itself contradicts or supersedes this prior in several ways: the firm now claims $5.7B raised across eight funds (not €1.3B), describes itself as investing "from seed to IPO" (not just Series A), and the "Balderton Build" brand is entirely absent from the live site — replaced by "Portfolio Services" and "Founder Wellbeing and Performance." The prior also omits major current portfolio companies like Wayve ($8.6B valuation), The Exploration Company, Proxima Fusion, and Dash0 (new unicorn). The site's positioning as "Europe's leading venture firm focused exclusively on European-founded tech companies" is sharper and more exclusive than the prior's generic description.

Schema Posture

Every page carries consistent Organization schema with name, URL, logo, and description. Pages also include BreadcrumbList and WebPage schemas. News articles use BlogPosting with author, publisher, and datePublished/dateModified. No FAQPage, Product, InvestmentFund, or FinancialService schema types are used anywhere — the site has no structured data for its funds, investment thesis, or portfolio companies as individual entities. The portfolio page lists 80+ companies with investment stage, location, and sector but encodes none of this in schema. The team page lists 40+ people with roles but uses no Person schema with worksFor relationships.

Content & Signals

The homepage is a strong narrative page (614 words) with founder testimonials, portfolio logos, and a clear value proposition. The about page states "26 years embedded in European tech, $5.7B raised, 250+ companies backed." The portfolio page is a comprehensive filterable directory with status (live/exited), location, category, and investment year for each company. The resources section contains 20+ long-form guides on AI, ESG, sales, and internationalisation. The site has no FAQ content, no comparison tables, and no investment-thesis page that explains sector focus or check sizes — details an AI crawler would surface for a founder researching potential investors. The news section is active with 2026 articles, showing the site is well-maintained. The dateModified on the homepage reads 2026-04-02, suggesting forward-dated content management.

External Signals

DNS TXT records show domain verification tokens for OpenAI, Anthropic, Perplexity, Dust, Langdock, Lovable, Manus, Attio, and Miro — indicating the firm actively verifies its domain across multiple AI platforms and SaaS tools. The anthropic-domain-verification and openai-domain-verification records confirm intentional AI-platform integration. No negative external signals were found in search results. The firm's Medium blog and LinkedIn are linked from every page, providing additional off-domain content for AI training data.

Findings

1. ClaudeBot selectively blocked by Cloudflare WAF High

   ClaudeBot receives a 403 response from Cloudflare while all other major AI crawlers (GPTBot, Google-Extended, PerplexityBot, OAI-SearchBot, anthropic-ai) get full 200 responses with identical content to a browser. This selective blocking creates an inconsistent AI-visibility posture.

   What to change: Remove the Cloudflare WAF rule that blocks the ClaudeBot user-agent string, or ensure ClaudeBot is allowed access like other AI crawlers.

2. No InvestmentFund or FinancialService schema for funds Medium

   The site uses Organization, BreadcrumbList, and WebPage schema but lacks InvestmentFund, FinancialService, or Product schema types to describe its funds, investment thesis, or portfolio companies as structured entities. This limits AI crawlers' ability to understand the firm's financial products and investment focus.

   What to change: Add InvestmentFund or FinancialService schema to pages describing funds, and use Product or InvestmentProduct schema for portfolio companies with relevant properties.

3. Team page lacks Person schema with worksFor relationships Medium

   The team page lists 40+ people with roles but uses no Person schema to encode individual profiles or their worksFor relationship to the organization. This prevents AI crawlers from associating team members with Balderton Capital in structured data.

   What to change: Add Person schema markup to each team member listing, including name, jobTitle, and worksFor properties pointing to the Organization schema.

4. Portfolio companies not encoded in structured data Medium

   The portfolio page lists 80+ companies with investment stage, location, and sector but encodes none of this in schema. AI crawlers cannot extract structured information about portfolio companies, their investment stages, or sectors.

   What to change: Add schema markup for each portfolio company, using Product or Organization schema with properties for investment stage, sector, and location.

5. No dedicated investment-thesis or sector-focus page Medium

   The site lacks a page that explains sector focus, check sizes, or investment criteria — details an AI crawler would surface for a founder researching potential investors. This limits the site's ability to attract inbound founder interest via AI search.

   What to change: Create a dedicated page outlining investment thesis, sector focus, check sizes, and stage preferences, and link it from the main navigation.

6. No FAQ content for common founder questions Low

   The site has no FAQ page or FAQ schema markup. Common questions founders might ask (e.g., 'How to get a meeting?', 'What stages do you invest in?') are not addressed in a structured format that AI crawlers can surface as rich results.

   What to change: Add an FAQ page with common founder questions and implement FAQPage schema markup.

7. Security headers absent on homepage Low

   The homepage lacks HSTS, CSP, and X-Frame-Options security headers, which is unusual for a financial-services firm. While not directly impacting AI visibility, it may affect trust signals for AI crawlers that evaluate site security.

   What to change: Add HSTS, Content-Security-Policy, and X-Frame-Options headers to all pages.

8. LLM prior knowledge outdated on fund size and portfolio Medium

   The LLM prior knows Balderton as having raised €1.3B, but the site states $5.7B raised across eight funds. The prior also omits major current portfolio companies like Wayve, The Exploration Company, Proxima Fusion, and Dash0. This gap means AI crawlers relying on training data may surface outdated information.

   What to change: Ensure key financial figures and portfolio highlights are prominently featured on the site and in llms.txt to override outdated training data.

What's working

• Comprehensive llms.txt file (275 KB) with full content index — The llms.txt file is auto-generated by All in One SEO and contains a full index of posts and pages, providing AI crawlers with a structured entry point to site content.
• Healthy sitemap index with 13 sub-sitemaps — The sitemap index is well-structured with 13 sub-sitemaps covering posts, pages, team members, founders, resources, and categories, ensuring comprehensive crawl coverage.
• Consistent Organization schema on all pages — Every page carries Organization schema with name, URL, logo, and description, providing a solid structured data foundation for brand recognition by AI crawlers.
• Domain verified across multiple AI platforms — DNS TXT records show verification tokens for OpenAI, Anthropic, Perplexity, Dust, Langdock, Lovable, Manus, Attio, and Miro, indicating intentional AI-platform integration and trust signals.
• Homepage provides strong narrative with testimonials and portfolio logos — The homepage (614 words) includes founder testimonials, portfolio logos, and a clear value proposition, offering rich content for AI crawlers to summarize the firm's positioning.
• Portfolio page is a filterable directory with detailed company data — The portfolio page lists 80+ companies with status, location, category, and investment year, providing rich content for AI crawlers to extract portfolio information.
• News section regularly updated with 2026 articles — The news section contains recent articles (e.g., Wayve raises $1.5B Series D), demonstrating active content management and providing fresh content for AI crawlers.
• Resources section with 20+ long-form guides — The resources section contains 20+ long-form guides on AI, ESG, sales, and internationalisation, providing in-depth content that AI crawlers can surface for founders.