sailpoint.com — AI Site Grade

Analysis

The new era of adaptive identity — but AI crawlers find a site with zero structured data and a blog that tells them to go away

SailPoint.com runs a sophisticated Next.js site on Vercel behind Cloudflare, serving all AI crawlers (GPTBot, ClaudeBot, PerplexityBot, Google-Extended, Bytespider, Applebot-Extended, OAI-SearchBot, ChatGPT-User, anthropic-ai) the same full 475KB HTML as a browser — no UA-based blocking, no JS shell, no 403s. Yet the site is structurally invisible to AI engines in a different way: every single page tested — homepage, platform, suites, agentic fabric, about, solutions — returns zero JSON-LD schema of any type. No Organization, no WebSite, no Product, no FAQPage, no BreadcrumbList. The homepage has a FAQ section rendered in visible text, but no FAQPage schema wraps it. The suites page has a comparison table of five product tiers with pricing signals — no Product or ItemList schema. An AI crawler parsing this site gets rich marketing copy but no machine-readable entity graph.

Crawler Access

The robots.txt is a single line: User-agent: * with Disallow: /web-requests and a sitemap pointer. No AI-specific bot rules exist — no GPTBot, ClaudeBot, or Google-Extended directives at all. The llms.txt returns a 404 (a Next.js error page with noindex meta). The sitemap index is healthy, spanning 10 locale-specific sub-sitemaps with over 2,000 URLs. The blog index page carries noindex,nofollow — a deliberate choice that tells search engines and AI crawlers not to index the content hub. Individual blog posts appear to be indexed (they appear in the sitemap), but the blog hub itself is excluded from AI training corpora.

Cold-Knowledge Gap

The cold LLM knows SailPoint as an identity governance company founded in 2005, taken private by Thoma Bravo in 2022 for $6.9B, with products IdentityNow and IdentityIQ. The cold knowledge is 12-18 months stale — it does not know about the 2025 IPO (NASDAQ: SAIL), the rebranding to "adaptive identity," the new Agentic Fabric product line, the Harbor Pilot conversational AI, or the $1B+ ARR milestone. The site itself has aggressively repositioned around "adaptive identity" and "securing AI agents," but the model's prior is still rooted in the legacy IGA category language. The site's homepage H1 is "The new era of adaptive identity" — a phrase absent from the model's knowledge entirely.

Schema Posture

Beyond the total absence of JSON-LD, the heading structure is flat and inconsistent. The homepage has a single H1 ("The new era of adaptive identity") followed by 6 H2 tags and 1 H3. The platform page has one H1 and 14 H2 tags with no H3 hierarchy. The about page has one H1 and 12 H2 tags. No page uses H3 or H4 for meaningful sub-hierarchy — content that is visually organized into cards and tabs is flattened into a single heading level. The FAQ sections on the platform, suites, and agentic fabric pages are rendered as visible H2-headed text blocks but lack FAQPage schema, meaning AI engines cannot extract Q&A pairs as structured data.

External Signals

The DNS TXT records reveal an unusually broad third-party verification footprint: Anthropic (anthropic-domain-verification), Apple, Canva, Cursor, Docker, Drift, Jamf, Krisp, Lovable, Miro, MongoDB, OneTrust, Palo Alto Networks, Stripe, and Wiz all have verification tokens on the domain. This suggests extensive API/integration partnerships but also a sprawling vendor attack surface. The site claims "53% of the Fortune 500" as customers and cites Gartner Peer Insights Customers' Choice, IDC Marketscape leadership, and KuppingerCole Leadership Compass — all claims that appear in visible text but have zero schema markup to substantiate them for AI consumption.

Surprising Findings

The /about URL returns a 404 (the actual about page lives at /why-us/about-us with a canonical pointing to itself — a redirect that should exist but does not). The blog post titled "Introducing SailPoint's new framework: Governing AI agents before they run wild" (linked from the blog index with a Feb 20 date) returns a 404 when accessed at the expected slug path. The site has a /web-requests path disallowed in robots.txt but no obvious reason why that single path is blocked while everything else is open. The site runs on Vercel with Next.js and Cloudflare, which means AI crawlers get server-rendered HTML — no JavaScript rendering risk — but the 475KB page weight is heavy for a single homepage.

Findings

1. Zero JSON-LD schema on any page High

   Every tested page — homepage, platform, suites, agentic fabric, about, solutions — returns no JSON-LD schema of any type. No Organization, WebSite, Product, FAQPage, or BreadcrumbList exists, making the site's entity graph invisible to AI crawlers.

   What to change: Add JSON-LD schema of types Organization, WebSite, Product, FAQPage, and BreadcrumbList to all relevant pages. Use structured data to mark up FAQ sections, product tiers, and customer claims.

2. llms.txt file returns 404 Medium

   The llms.txt file at /llms.txt returns a 404 error page with a noindex meta tag. This file is a standard way to guide AI crawlers to key content, and its absence means AI agents have no curated entry point.

   What to change: Create an llms.txt file listing the most important pages for AI consumption, such as the platform, agentic fabric, and about pages.

3. Blog index page has noindex,nofollow Medium

   The blog index page at /blog carries a noindex,nofollow meta tag, telling search engines and AI crawlers not to index the content hub. While individual posts may be indexed, the hub itself is excluded from AI training corpora.

   What to change: Remove the noindex,nofollow directive from the blog index page to allow AI crawlers to discover and index the content hub.

4. Cold LLM knowledge is 12-18 months stale High

   The cold LLM knows SailPoint as a pre-IPO IGA company with IdentityNow and IdentityIQ, but lacks knowledge of the 2025 IPO, adaptive identity rebranding, Agentic Fabric product, Harbor Pilot conversational AI, and $1B+ ARR milestone. The site's current positioning is absent from the model's prior.

   What to change: Publish structured data and authoritative content (e.g., press releases, Wikipedia updates) to help AI models refresh their knowledge. Consider submitting to AI training data sources.

5. Flat and inconsistent heading hierarchy Medium

   Pages use a single H1 followed by many H2 tags with no H3 or H4 sub-headings. Content visually organized into cards and tabs is flattened into one heading level, reducing semantic structure for AI parsers.

   What to change: Implement a proper heading hierarchy with H2 for major sections and H3/H4 for sub-sections, especially in FAQ areas and product comparison tables.

6. FAQ sections lack FAQPage schema Medium

   The platform, suites, and agentic fabric pages contain visible FAQ sections with questions and answers, but none are marked up with FAQPage schema. AI crawlers cannot extract Q&A pairs as structured data.

   What to change: Add FAQPage schema markup to all FAQ sections, wrapping each Q&A pair in a Question and Answer structure.

7. /about URL returns 404 Low

   The /about path returns a 404 error. The actual about page lives at /why-us/about-us. No redirect is in place, causing broken links and poor user and crawler experience.

   What to change: Set up a 301 redirect from /about to /why-us/about-us.

8. Blog post about governing AI agents returns 404 Low

   A blog post titled 'Introducing SailPoint's new framework: Governing AI agents before they run wild' linked from the blog index returns a 404 at its expected slug. This suggests a broken link or unpublished content.

   What to change: Fix the broken link or publish the post at the expected URL. If the post is removed, update the blog index to remove the link.

9. No AI-specific bot rules in robots.txt Low

   The robots.txt file contains only a single User-agent: * rule with a Disallow for /web-requests. No directives for GPTBot, ClaudeBot, Google-Extended, or other AI crawlers exist, leaving their access unmanaged.

   What to change: Consider adding explicit rules for AI crawlers if any sections should be blocked or allowed differently.

10. Homepage HTML weight is 475KB Low

    The homepage HTML is 475KB, which is heavy for a single page. While server-rendered, this size may slow down crawling and increase bandwidth costs.

    What to change: Optimize the homepage HTML size by reducing inline scripts/styles or deferring non-critical content.

What's working

• AI crawlers receive full server-rendered HTML — All tested AI crawlers (GPTBot, ClaudeBot, etc.) receive the same 475KB HTML as a browser, with no UA-based blocking, JavaScript shells, or 403 errors. This ensures content is fully accessible to AI engines.
• Sitemap index is healthy with 10 locale-specific sub-sitemaps — The sitemap index contains 10 sub-sitemaps covering over 2,000 URLs, ensuring comprehensive discovery of site content.
• Modern tech stack with server-side rendering — The site runs on Next.js on Vercel behind Cloudflare, providing server-rendered HTML that is easily crawlable without JavaScript execution risks.
• Broad third-party domain verification footprint — DNS TXT records show verification tokens for Anthropic, Apple, Canva, Cursor, Docker, Drift, Jamf, Krisp, Lovable, Miro, MongoDB, OneTrust, Palo Alto Networks, Stripe, and Wiz, indicating strong integration partnerships.
• Dedicated Agentic Fabric product page with rich content — The Agentic Fabric page provides 1571 words of detailed content about identity security for AI, positioning SailPoint for AI-related queries.
• Dedicated solution page for non-human identities — The /solutions/security-non-human-identities page provides 1178 words of content addressing machine identity security, a growing AI-related topic.
• About page exists at /why-us/about-us with canonical — The about page at /why-us/about-us is well-formed with a self-referencing canonical tag, providing a proper source for company information.