ezcater.com — AI Site Grade

Analysis

Interesting — the search page is behind Cloudflare challenge for both bots and browsers. That's a JS-gated page. Now I have enough to write the audit.

ezCater's April 2026 rebrand — from "catering marketplace" to "enterprise workplace food platform" — is invisible to AI crawlers because the site's own content and schema still describe the old positioning, creating a gap between what the brand now claims and what LLMs know.

Crawler Access

All major AI crawlers — GPTBot, ClaudeBot, PerplexityBot, Google-Extended, OAI-SearchBot, anthropic-ai, Bytespider, Applebot-Extended — receive a full 200 response with ~207KB of content from the homepage, identical to browser baseline. No UA-based blocking exists. However, the robots.txt contains no AI-specific directives at all — no mention of any AI crawler. It does disallow /catering/search/* under the wildcard rule, which blocks crawlers from the core search/ordering pages. The /llms.txt returns a 404 (Next.js error page). The standard /sitemap.xml also 404s; sitemaps live only at /lunchrush/sitemap_index.xml (blog content) and a gzipped landings sitemap on a CDN subdomain.

Cold-Knowledge Gap

The LLM knows ezCater as a "US-based online food ordering platform" and a "Catering Marketplace" founded in 2007, with the ZeroCater acquisition in 2021 and $300M+ funding. This is the old brand story. The actual site, as of April 2026, has undergone a full rebrand — new logo, new identity developed with agency Koto, sunsetting of the Relish sub-brand, and repositioning as an "enterprise-grade workplace food platform" serving Bayer, FedEx, P&G, T-Mobile, and Vanderbilt. The site now emphasizes spend management dashboards, SAP Concur integration, Slack integration, tax-exempt ordering, and procurement punchout (Coupa, Ariba, Jaggaer). The LLM knows nothing about this enterprise pivot, the rebrand, or the new product suite.

Schema Posture

The homepage carries a single Corporation schema with basic org info, address, and social links. The enterprise solution pages (/company/corporate-solutions/, /company/corporate-solutions/recurring-employee-meals/, /company/corporate-solutions/food-spend-management/) all use WebPage schema with breadcrumbs — no Product, SoftwareApplication, or Service schema on any of the product pages. The FAQ sections on the corporate solutions pages are rendered as plain HTML headings (H2 questions, body text answers) with no FAQPage schema markup, meaning AI crawlers see them as unstructured text rather than structured Q&A.

External Signals

The newsroom (/company/news/) shows active press coverage from Bloomberg Businessweek, Inc., Restaurant Business, and HNGRY through May 2026. The rebrand press release (April 8, 2026) is the most recent major announcement. DNS records reveal an anthropic-domain-verification TXT record, indicating ezCater has proactively verified the domain with Anthropic for Claude-powered features or crawler access. No negative external signals (layoffs, scandals) surfaced in search.

Findings

1. Rebrand to enterprise workplace food platform invisible to LLMs High

   ezCater rebranded in April 2026 from a catering marketplace to an enterprise workplace food platform, but the site's content and schema still describe the old positioning. LLMs know ezCater only as a catering marketplace, creating a gap between the brand's current identity and AI knowledge.

   What to change: Update homepage and key landing pages to prominently describe the enterprise workplace food platform positioning. Add structured data (e.g., SoftwareApplication schema) to reflect the new product suite.

2. Robots.txt blocks AI crawlers from search and ordering pages High

   The robots.txt disallows /catering/search/* under the wildcard rule, blocking all crawlers including AI bots from the core search and ordering functionality. This prevents LLMs from indexing the platform's primary user-facing features.

   What to change: Remove or relax the disallow rule for /catering/search/* in robots.txt, or allow specific AI crawlers to access these pages.

3. No /llms.txt file published Medium

   The /llms.txt endpoint returns a 404 error page. This file is a standard way to provide LLMs with curated information about the site, and its absence means ezCater misses an opportunity to directly influence AI knowledge about the rebrand and enterprise features.

   What to change: Create and publish an /llms.txt file that describes the company's current enterprise positioning, key products, and recent rebrand.

4. Enterprise solution pages lack Product or SoftwareApplication schema High

   Pages like /company/corporate-solutions/, /company/corporate-solutions/recurring-employee-meals/, and /company/corporate-solutions/food-spend-management/ use only WebPage schema with breadcrumbs. No Product, SoftwareApplication, or Service schema is present, making it harder for AI crawlers to understand the offerings as structured entities.

   What to change: Add SoftwareApplication or Product schema to each enterprise solution page, including properties like name, description, applicationCategory, and offers.

5. FAQ sections on corporate pages lack FAQPage schema Medium

   The FAQ sections on corporate solutions pages are rendered as plain HTML headings and paragraphs, without FAQPage schema markup. This means AI crawlers see them as unstructured text rather than structured Q&A, reducing the chance of being used in rich results or LLM training.

   What to change: Add FAQPage schema markup to FAQ sections, with Question and Answer properties for each Q&A pair.

6. Standard sitemap.xml returns 404 Medium

   The standard /sitemap.xml URL returns a 404 error. Sitemaps are only available at non-standard paths (/lunchrush/sitemap_index.xml and a gzipped CDN sitemap). This may hinder crawler discovery of important pages.

   What to change: Create a sitemap.xml at the standard path that includes all important pages, or redirect it to the existing sitemap index.

7. Search page behind Cloudflare challenge blocks AI crawlers High

   The /catering/search/new page returns a 403 with Cloudflare challenge for both browser and GPTBot. This JS-gated page is inaccessible to AI crawlers, preventing them from indexing the core search functionality.

   What to change: Allow AI crawlers to bypass the Cloudflare challenge for the search page, or serve a static HTML version to bots.

8. Homepage content is sparse for AI understanding Medium

   The homepage contains only 111 words of visible text, which is minimal for conveying the full scope of the enterprise platform. AI crawlers may not extract enough context about the rebrand and product suite.

   What to change: Expand the homepage copy to include a clear value proposition for the enterprise platform, key features, and client logos.

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

   The robots.txt does not mention any AI crawler by name (e.g., GPTBot, ClaudeBot). While this means they are not explicitly blocked (except by wildcard rules), it also means ezCater is not proactively managing their access or providing hints about which content to prioritize.

   What to change: Add explicit directives for AI crawlers, such as allowing them to access key pages and disallowing low-value pages.

10. Newsroom pages lack Article or NewsArticle schema Low

    The newsroom page and press releases use only WebPage schema. Adding Article or NewsArticle schema would help AI crawlers understand the content as news and potentially surface it in relevant contexts.

    What to change: Add NewsArticle schema to press releases and Article schema to news posts, including headline, datePublished, and author.

What's working

• All major AI crawlers allowed on homepage — GPTBot, ClaudeBot, PerplexityBot, Google-Extended, and other AI crawlers receive a full 200 response with ~207KB of content from the homepage, identical to browser baseline. No UA-based blocking exists.
• Anthropic domain verification TXT record present — DNS records include an anthropic-domain-verification TXT record, indicating proactive verification with Anthropic for Claude-powered features or crawler access.
• Active newsroom with recent press coverage — The newsroom shows press coverage from Bloomberg Businessweek, Inc., Restaurant Business, and HNGRY through May 2026, including the April 2026 rebrand announcement. This provides authoritative external signals.
• Corporation schema on homepage with basic org info — The homepage includes a single Corporation schema with organization name, address, and social links, providing foundational structured data.
• Blog sitemap available at /lunchrush/sitemap_index.xml — A sitemap index with 80 URLs is available for the Lunch Rush blog, helping crawlers discover blog content.
• No negative external signals found — Web searches for layoffs, funding issues, or negative reviews returned no results, indicating a clean external reputation.