getgarner.com — AI Site Grade

Analysis

The $2.74B Company That AI Models Have Never Heard Of

Garner Health operates a domain bifurcation (getgarner.com redirects to garnerhealth.com) that creates a cold-knowledge vacuum: a frontier LLM queried on "getgarner.com" reports zero awareness of the company, despite the site itself describing a $2.74 billion valuation, $200M annual revenue, 2.5M members, and a $100M Series E closed May 2026.

Crawler Access

All major AI crawlers — GPTBot, ClaudeBot, PerplexityBot, Google-Extended, OAI-SearchBot, ChatGPT-User, Applebot-Extended, anthropic-ai — receive a full 200 response with identical byte size (77,083 bytes) to a browser baseline. The sole exception is Bytespider (TikTok/ByteDance), which gets a Cloudflare 403 block. The site runs on Cloudflare with Webflow hosting (x-wf-region header). The robots.txt contains only a sitemap directive and no AI-bot rules whatsoever — no disallow, no allow, no crawl-delay. This is permissive but leaves the site without any governance over which AI crawlers can scrape content or how frequently.

llms.txt — A Standout Asset

Garner Health has one of the most sophisticated llms.txt files observed. It includes a brand summary, core facts, explicit "Important Notes for LLMs" (instructing models not to describe Garner as insurance, not to call rankings paid placements, and to distinguish between employer benefit, health plan solution, provider data tools, and member app), and a full linked index of primary pages, audience solutions, and resources. This file is a model implementation of the emerging standard and directly addresses the cold-knowledge gap.

Cold-Knowledge Gap

The disconnect is stark. The LLM queried on "getgarner.com" returned: "I do not have specific, verifiable information about getgarner.com... The site may be a small or niche business, a newly launched project, or a domain with limited public footprint." In reality, Garner Health serves 700+ organizations including USA Today, Paylocity, Archer-Daniels-Midland, and the University of Oklahoma; has saved customers over $1 billion; and was covered by Bloomberg (linked from the Series E blog post). The cold model has no awareness of any of this. The llms.txt is designed to fix this, but the model's training cutoff predates the file's creation or the file has not been indexed.

Schema Posture

JSON-LD is present on every page examined with Organization, WebSite, WebPage, BreadcrumbList, HowTo, FAQPage, Service, and BlogPosting types. The homepage schema includes foundingDate: 2019, founder: Nick Reber, and sameAs links to LinkedIn and X. The "How It Works" page adds a HowTo with three steps and an inline FAQPage. The blog uses BlogPosting with datePublished and author. Missing: Product or SoftwareApplication schema for the mobile app, Review schema for the 12+ member testimonials displayed on the homepage, and LocalBusiness for any physical presence.

Content and JS-Rendering Risk

The site is built on Webflow (server-rendered HTML, not a JS shell). The homepage delivers 820 words of visible text from a plain GET. However, the FAQ page (/frequently-asked-questions) returns only 15 words of extracted text — the Q&A content is loaded client-side via accordion interactions, meaning AI crawlers that do not execute JavaScript see an empty shell. This is a significant gap: the FAQ page is explicitly linked from the llms.txt as a key resource, but its substantive content is invisible to crawlers.

External Signals

External search results returned zero organic results for "Garner Health reviews employer healthcare benefit" and "getgarner.com OR garnerhealth.com AI healthcare" — the domain has minimal indexed third-party coverage outside the Bloomberg article linked from the blog. The DNS TXT records reveal integrations with HubSpot, Zendesk, Greenhouse, Salesforce, SendGrid, Freshservice, and Outlook — a complex martech stack — plus an anthropic-domain-verification record confirming active Claude/Anthropic API usage.

Findings

1. Frontier LLMs have zero awareness of Garner Health despite $2.74B valuation High

   A query on 'getgarner.com' returned no verifiable information, describing the site as possibly a small or niche business. In reality, Garner Health serves 700+ organizations, has saved over $1 billion, and closed a $100M Series E. The llms.txt file is designed to address this, but the model's training cutoff or indexing gap prevents it from being effective.

   What to change: Ensure the llms.txt file is actively submitted to search engines and AI crawlers, and consider adding structured data markup that explicitly states the company's valuation, revenue, and member count.

2. FAQ page content is invisible to non-JS crawlers High

   The /frequently-asked-questions page returns only 15 words of extracted text; the Q&A content is loaded client-side via accordion interactions. AI crawlers that do not execute JavaScript see an empty shell, despite the page being linked from llms.txt as a key resource.

   What to change: Server-render the FAQ content or use progressive enhancement so that the full Q&A text is present in the initial HTML response.

3. No Product or SoftwareApplication schema for the mobile app Medium

   The site uses Organization, WebSite, WebPage, and other schema types, but the mobile app (mentioned in llms.txt) lacks Product or SoftwareApplication schema. This limits AI models' ability to understand and surface the app as a distinct offering.

   What to change: Add Product or SoftwareApplication schema to the app landing page with properties like name, description, operating system, and application category.

4. Homepage testimonials lack Review schema markup Medium

   The homepage displays 12+ member testimonials, but none are marked up with Review schema. This prevents AI crawlers from extracting and potentially surfacing these social proof signals.

   What to change: Add Review schema to each testimonial block, including author, review body, and rating if applicable.

5. robots.txt lacks any AI-bot rules Low

   The robots.txt file contains only a sitemap directive and no rules for AI crawlers (no disallow, allow, or crawl-delay). While permissive, this leaves the site without governance over which AI crawlers can scrape content or how frequently.

   What to change: Add explicit rules for AI crawlers, such as allowing all but setting a crawl-delay to manage load.

6. Bytespider (TikTok/ByteDance) is blocked by Cloudflare Low

   The Bytespider crawler receives a 403 block from Cloudflare, while all other major AI crawlers are allowed. This may limit visibility on ByteDance's platforms.

   What to change: If visibility on ByteDance platforms is desired, allow Bytespider access via Cloudflare or robots.txt.

7. Minimal indexed third-party coverage in search results Medium

   Searches for 'Garner Health reviews employer healthcare benefit' and 'getgarner.com OR garnerhealth.com AI healthcare' returned zero organic results. The domain has limited indexed third-party coverage beyond a Bloomberg article linked from the blog.

   What to change: Increase PR and content marketing efforts to generate more third-party mentions and backlinks from reputable sources.

8. Domain bifurcation creates cold-knowledge vacuum Medium

   The primary domain getgarner.com redirects to garnerhealth.com, but the LLM queried on 'getgarner.com' had no awareness of the company. The redirect may confuse crawlers and dilute brand recognition.

   What to change: Consolidate all content under a single domain (garnerhealth.com) and ensure getgarner.com has proper canonical tags and consistent branding.

What's working

• Exemplary llms.txt file with brand summary and instructions for LLMs — The llms.txt file includes a brand summary, core facts, explicit notes for LLMs (e.g., not to describe Garner as insurance), and a full linked index of primary pages. This is a model implementation of the emerging standard.
• JSON-LD schema present on all pages with multiple types — Every page examined includes JSON-LD with Organization, WebSite, WebPage, BreadcrumbList, HowTo, FAQPage, Service, and BlogPosting types. The homepage schema includes foundingDate, founder, and sameAs links.
• All major AI crawlers allowed with full 200 responses — GPTBot, ClaudeBot, PerplexityBot, Google-Extended, OAI-SearchBot, ChatGPT-User, Applebot-Extended, and anthropic-ai all receive a full 200 response with identical byte size to a browser baseline.
• Server-rendered HTML on most pages avoids JS shell issues — The site is built on Webflow with server-rendered HTML. The homepage delivers 820 words of visible text from a plain GET, avoiding the JS shell problem common with SPAs.
• Blog posts use BlogPosting schema with datePublished and author — The blog page and individual posts include BlogPosting schema with datePublished and author properties, aiding AI crawlers in understanding content freshness and authorship.
• Anthropic domain verification record indicates active Claude API usage — DNS TXT records include an anthropic-domain-verification record, confirming active use of Claude/Anthropic API, which may indicate internal AI integration.
• HowTo and FAQPage schema on How It Works page — The How It Works page includes a HowTo schema with three steps and an inline FAQPage, providing structured guidance to AI crawlers.