convexin.com — AI Site Grade

Analysis

Convexin.com — AI-Visibility Audit

The site's homepage dateModified is set to 2026-05-28 — a date in the future — which means every AI crawler consuming the JSON-LD schema receives a timestamp that has not yet occurred, undermining temporal credibility signals.

Crawler Access

All 11 tested AI bot user-agents (GPTBot, ClaudeBot, PerplexityBot, Google-Extended, OAI-SearchBot, ChatGPT-User, anthropic-ai, Bytespider, Applebot-Extended, Perplexity-User) receive a 200 status with identical 142,854-byte payload — no UA-based blocking, no Cloudflare challenge, no JS shell. The robots.txt uses a Yoast-generated User-agent: * with Disallow: (empty, meaning allow all) and contains zero AI-bot-specific directives. No llms.txt exists (404). The site has a perplexity-ai-domain-verification TXT record, indicating active engagement with Perplexity's publisher program, but no corresponding llms.txt to feed it structured content.

Cold-Knowledge Gap

The LLM prior knows Convex as a 2019-founded specialty insurer/reinsurer co-founded by Stephen Catlin, with $1.7bn initial capital, focused on complex risks across energy, marine, aviation, property, casualty, and cyber. The site confirms this but reveals materially more current information the model lacks: Convex now has $5.9bn GWP (2025), a record $711m net income, an A (Excellent) AM Best rating (upgraded from A-), an A S&P rating, a new Lloyd's syndicate (1984), a new ownership structure with Onex as majority shareholder, and Stephen Catlin stepping down as Chairman with Bobby Le Blanc succeeding him. The cold model also mentions Warburg Pincus as a backer — the site shows Onex, not Warburg Pincus, as the current majority owner.

Schema Posture

Every page carries Yoast-generated JSON-LD with WebPage, WebSite, and BreadcrumbList types. However, the site is a specialty insurer and reinsurer — a financial services business — yet no page uses InsuranceProduct, InsurancePolicy, FinancialService, or Organization schema with proper insurance-specific properties. The WebSite schema includes a SearchAction but no potentialAction for quote requests or claims. The homepage's dateModified value of 2026-05-28 is a future date (as of the time of audit), which will cause schema validators and AI crawlers to flag or distrust the timestamp.

Content Architecture

The site has ~156+ indexed URLs across 9 sitemaps, with content spanning 2019 through May 2026. The blog/news section is active and well-maintained with recent posts (Q1 2026 results, Lloyd's syndicate expansion, Catlin's departure). However, the homepage and all subpages display stale sidebar "Related Articles" linking to 2021-2022 posts (Solvency reports, employee interviews) that are years old, creating a mixed freshness signal. The internal link sample reveals broken URL patterns like https://convexin.com/convexin.com/about-us/ (doubled domain) appearing in the page markup, which redirect to the correct page but represent a templating error that AI crawlers may treat as a soft signal of poor maintenance.

External Signals

The DNS TXT records show integrations with 20+ third-party services (Atlassian, Canva, Cursor, Docker, Dropbox, Klaviyo, LaunchDarkly, Miro, PandaDoc, Parallels, Slack, 1Password, Jamf, Remarkable, H1) — an unusually high number that signals a complex tech stack but also a fragmented verification footprint that AI crawlers could interpret as noise. No Reddit threads, Glassdoor reviews, or press citations surfaced in search results for the domain specifically. The site is hosted on Cloudflare with strong security headers (CSP, HSTS preload, frame-ancestors 'none'), which is appropriate for a financial services domain.

Findings

1. Homepage JSON-LD dateModified set to future date 2026-05-28 High

   The homepage's WebPage schema includes a dateModified value of 2026-05-28, a date that has not yet occurred. This causes schema validators and AI crawlers to flag or distrust the timestamp, reducing temporal credibility signals.

   What to change: Update the dateModified field in the JSON-LD schema on the homepage to reflect the actual last-modified date.

2. No insurance-specific schema types used on any page High

   Despite being a specialty insurer and reinsurer, the site uses only generic WebPage, WebSite, and BreadcrumbList schema. No InsuranceProduct, InsurancePolicy, FinancialService, or Organization schema with insurance properties are present, limiting rich snippet eligibility and AI understanding of the business.

   What to change: Add InsuranceProduct, InsurancePolicy, and Organization schema with appropriate properties to relevant pages.

3. LLM prior knowledge significantly outdated compared to current site content High

   The LLM prior knows Convex as a 2019-founded insurer with $1.7bn capital and Warburg Pincus backing, but the site shows $5.9bn GWP, record $711m net income, A (Excellent) AM Best rating, new Lloyd's syndicate 1984, Onex as majority owner, and Stephen Catlin stepping down. This gap means AI models may present outdated facts when citing the site.

   What to change: Publish an llms.txt file and ensure key facts (ratings, financials, leadership) are prominently featured in structured data and on high-level pages.

4. Sidebar 'Related Articles' link to 2021-2022 posts, creating mixed freshness signals Medium

   The homepage and subpages display related articles from 2021-2022 (Solvency reports, employee interviews) that are years old, which may dilute the freshness signal from newer content.

   What to change: Update the related articles widget to prioritize recent posts or dynamically show the latest content.

5. Internal links with doubled domain (convexin.com/convexin.com/) appear in markup Medium

   URLs like https://convexin.com/convexin.com/about-us/ appear in the page markup, which redirect to the correct page but represent a templating error that AI crawlers may treat as a soft signal of poor maintenance.

   What to change: Fix the templating logic to prevent doubled domain paths in internal links.

6. No llms.txt file despite Perplexity publisher program participation Medium

   The site has a perplexity-ai-domain-verification TXT record indicating active engagement with Perplexity's publisher program, but no llms.txt file exists (404). This misses an opportunity to feed structured content to AI crawlers.

   What to change: Create an llms.txt file with a summary of the site and links to key pages.

7. Robots.txt lacks AI-bot-specific directives Low

   The robots.txt uses a generic User-agent: * with Disallow: (empty) and does not name any AI bots. While this allows all bots, it also means the site cannot fine-tune access for specific AI crawlers if needed.

   What to change: Consider adding explicit directives for AI bots if selective blocking or rate-limiting becomes necessary.

8. High number of third-party DNS verification records may be interpreted as noise Low

   The DNS TXT records include verifications for 20+ third-party services (Atlassian, Canva, Cursor, etc.), which is unusually high and could be seen as fragmented or noisy by AI crawlers analyzing domain signals.

   What to change: Consolidate or remove unused verification records to reduce noise.

What's working

• All 11 tested AI bots receive full 200 OK access with no blocking — Every major AI crawler (GPTBot, ClaudeBot, PerplexityBot, etc.) receives a 200 status with identical payload, no UA-based blocking, and no Cloudflare challenges. This ensures content is fully accessible to AI crawlers.
• Blog/news section actively maintained with recent posts through 2026 — The site publishes regular news articles, including Q1 2026 results, Lloyd's syndicate expansion, and leadership changes, providing fresh content for AI crawlers.
• Every page includes Yoast-generated JSON-LD with WebPage, WebSite, and BreadcrumbList — The site consistently uses Yoast SEO to output structured data, providing a baseline of schema markup that helps AI crawlers understand page types and site structure.
• Sitemap with 80 URLs and 9 sitemap indexes provides full site coverage — The sitemap is well-structured and includes all major pages, ensuring AI crawlers can discover the full breadth of content.
• Cloudflare hosting with robust security headers (CSP, HSTS preload, frame-ancestors 'none') — The site uses Cloudflare with Content Security Policy, HSTS preload, and frame-ancestors 'none', which are appropriate for a financial services domain and build trust signals.
• Perplexity publisher program verification record present — The DNS includes a perplexity-ai-domain-verification TXT record, indicating active engagement with Perplexity's publisher program, which can improve visibility in Perplexity's AI search.