medimpact.com — AI Site Grade

Analysis

MedImpact — AI-Visibility Audit

The site's press release listing page advertises at least six recent press releases that all return 404 errors, including a GLP-1 obesity care announcement and an acquisition — meaning AI crawlers that index the listing page will find broken content paths for the site's most newsworthy material.

Crawler Access

All major AI crawlers (GPTBot, ClaudeBot, PerplexityBot, Google-Extended, OAI-SearchBot, Bytespider, Applebot-Extended) receive 200 status with identical byte-size content as a browser on the homepage and all accessible pages. No UA-based blocking exists. The robots.txt contains a generic User-agent: * rule with no AI-bot-specific directives — GPTBot, ClaudeBot, and others are neither explicitly allowed nor disallowed, leaving them subject to the default Disallow: /admin/, /search/, /user/* rules. The llms.txt returns 404, a missed opportunity for a company that positions itself as a technology-forward PBM.

Schema and Structured Data Posture

Zero pages across the entire site contain any JSON-LD structured data. The homepage, solution pages, press releases, and "who we are" page all return "schema": {"types": [], "jsonld": []}. No Organization, WebSite, FAQPage, Article, or BreadcrumbList schema exists anywhere. This is a critical gap: AI engines performing entity extraction or knowledge graph construction receive no machine-readable signal about MedImpact being a PBM, its headquarters, its leadership, or its 20+ million member count.

Cold-Knowledge Gap

The LLM prior knows MedImpact as "one of the largest privately held PBMs in the U.S.," headquartered in San Diego, founded in 1989, with a "MedImpact Direct" transparent pricing model. It also recalls a class-action lawsuit over MAC pricing. The actual site never mentions "MedImpact Direct" by name on any fetched page, never addresses the lawsuit, and describes itself as serving "more than 20 million members" — while the cold knowledge says "55 million consumers globally" (a figure that appears in a 2019 press release but not on current pages). The site's positioning as "the largest independent PBM" is consistent, but the discrepancy between 20M (site) and 55M (older press release / LLM prior) creates confusion for AI engines reconciling conflicting signals.

Content Fragmentation and Broken Press Infrastructure

The resources listing page (/clients/resources/press-releases) displays six press release teasers with "READ MORE" links, but at least four of those links resolve to 404 pages (the GLP-1 obesity solutions announcement, the Sav-Rx acquisition, the New York expansion, and the women's health program). Only the URAC accreditation press release (March 2026) and the NCQA reaccreditation (2019) actually load. The 404 pages share a canonical URL pointing to http://www.medimpact.com:443/node/94 — a Drupal node path that itself returns a 404. The canonical tags across the entire site use http://www.medimpact.com:443/ (with port 443 in the URL), a malformed canonical that differs from the actual https://www.medimpact.com served URL.

Leadership Page Blocked

The /clients/Ourleadership page returns 403 Access Denied to all user-agents, including all AI bots. This blocks AI crawlers from accessing executive names, titles, and bios — information that would normally populate entity knowledge graphs and authoritativeness signals.

Findings

1. Four of six press releases on listing page return 404 errors High

   The press release listing page advertises six recent announcements, but four links resolve to 404 pages, including a GLP-1 obesity care solution and an acquisition. AI crawlers indexing the listing page encounter broken content paths for the site's most newsworthy material.

   What to change: Fix the broken press release URLs or remove the teasers from the listing page. Ensure all linked press releases resolve to 200 status.

2. No JSON-LD structured data on any page High

   Every page tested, including the homepage, solution pages, press releases, and 'who we are' page, returns zero JSON-LD schemas. No Organization, WebSite, Article, or BreadcrumbList schema exists, depriving AI engines of machine-readable entity signals.

   What to change: Add JSON-LD structured data across the site, starting with Organization schema on the homepage and Article schema on press releases.

3. Leadership page returns 403 Access Denied to all user-agents High

   The /clients/Ourleadership page returns a 403 status for all AI crawlers and browsers, blocking access to executive names, titles, and bios that would populate entity knowledge graphs and authoritativeness signals.

   What to change: Remove the access restriction on the leadership page or serve a public version to crawlers.

4. Canonical tags use malformed URL with port 443 Medium

   Across the site, canonical tags reference http://www.medimpact.com:443/ (with port 443 in the URL), which differs from the actual served URL https://www.medimpact.com. This malformed canonical can confuse search engines and AI crawlers about the preferred URL.

   What to change: Update canonical tags to use the correct https://www.medimpact.com URL without the port number.

5. llms.txt file returns 404 Medium

   The site does not serve an llms.txt file, missing an opportunity to provide AI crawlers with a curated list of important pages and context about the company.

   What to change: Create an llms.txt file at the root that lists key pages and provides a brief company description.

6. Members page returns connection reset error Medium

   The /members URL returns a ReadError (connection reset by peer), preventing AI crawlers from accessing member-facing content. This may be a transient issue but blocks crawlers entirely.

   What to change: Investigate and fix the connection reset issue on the /members page to ensure it returns a proper response.

7. Member count discrepancy between site and LLM prior Medium

   The site states 'more than 20 million members' while the LLM prior recalls '55 million consumers globally' from a 2019 press release. This inconsistency can confuse AI engines reconciling conflicting signals.

   What to change: Update the site to consistently reflect the current member count and consider adding a press release or page that explains the change.

8. Transparent pricing model 'MedImpact Direct' not mentioned on site Low

   The LLM prior recalls a 'MedImpact Direct' transparent pricing model, but no fetched page on the current site mentions this offering by name, creating a gap between AI knowledge and site content.

   What to change: Add a page or section that describes the MedImpact Direct transparent pricing model to align with AI knowledge.

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

   The robots.txt file contains only a generic User-agent: * rule with no explicit allow/disallow for GPTBot, ClaudeBot, or other AI crawlers. While they are not blocked, the lack of explicit directives leaves ambiguity.

   What to change: Add explicit allow rules for AI crawlers in robots.txt to ensure they can access key pages.

What's working

• All major AI crawlers receive 200 status on accessible pages — The homepage and all accessible pages return 200 status with identical content to all tested AI crawlers, with no UA-based blocking.
• Sitemap available with 77 URLs — The site serves a sitemap with 77 URLs, providing a clear index for crawlers to discover content.
• Consistent positioning as largest independent PBM — The site consistently describes MedImpact as 'the largest independent PBM' across multiple pages, aligning with the LLM prior and reinforcing brand identity.
• Press release listing page exists and is accessible — The /clients/resources/press-releases page is accessible and lists recent press releases, providing a hub for news content.
• URAC accreditation press release loads correctly — The press release about MedImpact earning URAC PBM accreditation (March 2026) returns 200 status with full content, providing a positive signal for AI crawlers.
• NCQA reaccreditation press release loads correctly — The press release about NCQA reaccreditation (2019) returns 200 status with full content.