reybold.com — AI Site Grade

Analysis

AI crawlers see the full site — but what they find is a brand the LLM world barely recognizes

Every major AI crawler — GPTBot, ClaudeBot, PerplexityBot, Google-Extended, OAI-SearchBot, ChatGPT-User — receives a 200 with full HTML content (206KB, identical to browser baseline) from reybold.com. Only Bytespider (ByteDance) gets a 403 from Cloudflare. The robots.txt is a bare-bones wildcard rule (Disallow: /wp-admin/) with no AI-bot-specific directives, no crawl-delay for bots, and no llms.txt (404). The site runs on WP Engine behind Cloudflare with HSTS preload and strong security headers. No JS-rendering risk: all pages return rich server-rendered HTML.

Cold-Knowledge Gap

The LLM cold-knowledge prior describes Reybold Group as a "privately held real estate development and construction company" operating across Delaware, Maryland, and Pennsylvania, with named communities like "The Reserve at Glasgow" and "Whitehall." None of those communities exist on the current site. The actual portfolio is entirely different: St. Andrews, Meridian Crossing, Ascot Landing, Dodson Court, Lincoln Center, Columbia Place (55+), and manufactured home communities under MH Reybold. The cold model also references a "Reybold Homes" brand for new construction — the site uses no such brand; the construction arm is simply /build/. The model knows nothing about the self-storage, RV/boat storage, coworking (PLY), or corporate housing lines of business, which are prominent on the site. The model's knowledge is stale by at least one portfolio generation.

Schema Posture

The site has robust JSON-LD on every page, anchored to a HomeAndConstructionBusiness + Organization type at #organization. The schema includes Place, WebSite (with SearchAction), BreadcrumbList, WebPage, Article/BlogPosting, and on the St. Andrews page, VideoObject and FloorPlan types. The Columbia Place page includes a Residence type. However, no FAQPage, Product, ItemList, or LocalBusiness (per-community) schema exists — each community page is typed as Article rather than ApartmentComplex or RealEstateListing. The Article type wrapping is a structural mismatch for property listing pages.

Content & Answer Signals

The homepage is a 515-word marketing splash with no FAQ, no comparison tables, and no structured data beyond the organization schema. The blog (20+ posts from late 2025 through mid-2026) is the strongest answer-format content: posts like "Renting vs Owning in Retirement: A Delaware Cost Comparison" and "Defining Modular, Manufactured, and Mobile Homes" use comparison language and definition patterns. The St. Andrews community page is 2,524 words with detailed floorplan listings (square footage, bed/bath counts) but no FAQPage schema for common renter questions. The meta description on the St. Andrews page is broken — it contains raw JavaScript (document.addEventListener('DOMContentLoaded', function ()...) instead of a human-readable summary.

External Signals

Web searches returned zero results for Reybold Group reviews, Reddit threads, or apartment complaints — the brand has effectively no organic third-party footprint in searchable public forums. The only external signals are the social media profiles linked in schema (Facebook, Instagram, Twitter/X) and the Diamond State Apartment Association awards page. This absence of external mentions means AI engines have almost no corroborating signals to reinforce the brand's claims.

Surprising Findings

The dateModified field on the homepage reads 2026-04-22 — a date in the future relative to the current year, suggesting a WordPress plugin or caching layer is injecting an incorrect timestamp. The /build/ page has only 161 words of content and was last modified in September 2024, making it the thinnest major section. The /commercial/ page is similarly thin at 315 words. The site has no pricing data anywhere — no rent ranges, no price lists — which is a gap for any AI engine trying to answer "how much does a Reybold apartment cost." The author entity in schema is "Quattro Tech" (the web agency), not a Reybold employee, which dilutes the organization's authorship signal.

Findings

1. LLM cold knowledge is stale — references communities not on the site High

   The LLM prior describes communities like 'The Reserve at Glasgow' that no longer exist on the site, and misses current lines of business (self-storage, coworking, corporate housing). This means AI engines have outdated or incomplete information about the brand.

   What to change: Update the site's content to reflect the current portfolio and lines of business, and consider publishing an llms.txt file to provide authoritative brand information to AI crawlers.

2. Zero third-party mentions in searchable forums High

   Web searches for Reybold Group reviews, Reddit threads, and apartment complaints returned no results. AI engines have no corroborating external signals to reinforce the brand's claims.

   What to change: Encourage residents to leave reviews on Google, Yelp, and ApartmentRatings; engage in local forums and social media to build an organic third-party footprint.

3. Community pages use Article schema instead of ApartmentComplex or RealEstateListing High

   Pages like St. Andrews and Columbia Place are typed as 'Article' in JSON-LD, which is a structural mismatch for property listing pages. This reduces the chance that AI engines understand them as apartment communities.

   What to change: Replace the Article type with ApartmentComplex or RealEstateListing schema on community pages, including properties like floor plans, amenities, and pricing.

4. St. Andrews page has a broken meta description containing raw JavaScript High

   The meta description on the St. Andrews page contains raw JavaScript code instead of a human-readable summary, which harms search snippet quality and AI comprehension.

   What to change: Replace the meta description with a concise, human-readable summary of the community, e.g., 'St. Andrews offers luxury apartments in Bear, DE with 1-3 bedroom floor plans.'

5. No pricing data anywhere on the site Medium

   The site contains no rent ranges, price lists, or pricing information for any community. AI engines cannot answer 'how much does a Reybold apartment cost' from the site's content.

   What to change: Add pricing information to community pages, either as static content or via structured data (e.g., offers or priceRange).

6. Build and Commercial pages are very thin Medium

   The /build/ page has only 161 words and the /commercial/ page has 315 words, providing minimal content for AI engines to understand these business lines.

   What to change: Expand these pages with detailed descriptions, case studies, and relevant structured data to better represent the construction and commercial leasing services.

7. Homepage dateModified is set to a future date Medium

   The homepage's dateModified field reads 2026-04-22, which is in the future relative to the current year. This may confuse AI engines about content freshness.

   What to change: Investigate the WordPress plugin or caching layer causing the incorrect timestamp and ensure dateModified reflects the actual last modification date.

8. Schema author entity is the web agency, not Reybold Low

   The author property in JSON-LD is set to 'Quattro Tech' (the web agency) rather than a Reybold employee, diluting the organization's authorship signal.

   What to change: Update the author property to reference the Reybold organization or a named employee to strengthen brand authorship.

9. No llms.txt file published Low

   The site returns a 404 for /llms.txt, missing an opportunity to provide authoritative brand information directly to AI crawlers.

   What to change: Create an llms.txt file with a summary of the business, key pages, and lines of business to guide AI crawlers.

10. No FAQPage schema on any page Low

    Despite having blog posts that answer common questions, the site does not use FAQPage schema, which could help AI engines surface answers directly.

    What to change: Add FAQPage schema to pages that answer common renter questions, such as the blog posts or a dedicated FAQ page.

What's working

• All major AI crawlers receive full HTML content — Every major AI crawler (GPTBot, ClaudeBot, PerplexityBot, etc.) gets a 200 with full server-rendered HTML, with no JS-rendering risk.
• Comprehensive JSON-LD schema across the site — Every page includes JSON-LD with Organization, WebSite, BreadcrumbList, WebPage, and Article types, providing a solid structured data foundation.
• Blog contains answer-format content with comparisons and definitions — Blog posts like 'Renting vs Owning in Retirement' and 'Defining Modular, Manufactured, and Mobile Homes' use comparison and definition patterns that align with AI answer formats.
• St. Andrews page has extensive floorplan details — The St. Andrews community page contains 2,524 words with detailed floorplan listings including square footage and bed/bath counts, providing rich content for AI.
• Site uses Cloudflare, HSTS preload, and strong security headers — The site is behind Cloudflare with HSTS preload and robust security headers, ensuring safe and reliable access for crawlers.
• Sitemap is available and contains 80 URLs — The sitemap is accessible and lists 80 URLs, helping crawlers discover all pages.