mcneeslaw.com — AI Site Grade

Analysis

I have enough to write a sharp audit. Let me compile.

The cold-knowledge model thinks McNees has ~140 attorneys and no Ohio/DC offices — the site actually claims 170+ attorneys across 12 offices in 4 states and DC.

Crawler Access

Every major AI crawler — GPTBot, ClaudeBot, PerplexityBot, Google-Extended, OAI-SearchBot, ChatGPT-User, Applebot-Extended — receives a full 200 response with identical byte content (60,563 bytes) to a browser visit. No UA-based blocking, no JS shell, no Cloudflare challenge. The robots.txt contains zero AI-bot directives; the * rule only disallows /wp-admin/ and /wp-includes/. The llms.txt returns 404. The site runs on nginx behind WP Engine hosting on Google Cloud (35.203.175.30), with HSTS, X-Frame-Options SAMEORIGIN, and a 10-minute cache TTL.

Cold-Knowledge Gap

A frontier LLM queried cold describes McNees as having "over 140 attorneys" with offices in "Harrisburg, Lancaster, State College, and Scranton" — missing Pittsburgh, Radnor, Reading, Plymouth Meeting, York, Columbus, Frederick, Towson, and Washington, DC. The model also omits the firm's 170+ attorney count, its 12-office footprint across PA, OH, MD, and DC, and practices like Esports, NIL (Name/Image/Likeness), Gaming & Sports Wagering, and ESG — all listed on the site. The model's knowledge appears frozen at a pre-2020 snapshot.

Schema Posture

The homepage and all subpages carry a consistent Organization + WebSite + WebPage JSON-LD graph with sameAs links to Facebook, X, LinkedIn, and YouTube. However, no LocalBusiness or Attorney schema is present on any page examined. The 13 office location pages exist (e.g., /office/harrisburg/, /office/columbus/) but carry no LocalBusiness or GeoCoordinates schema — a missed signal for AI engines answering "where is McNees located?" The blog subdomain (palaborandemploymentblog.com) does include a LegalService schema with a single Harrisburg address, but it references http://www.mwn.com/ (a domain that does not resolve) as the url.

Content Architecture

The site is a WordPress installation with Yoast SEO, serving ~3,000+ indexed URLs across 14 sitemaps. The four topical blogs (Labor & Employment, Land Use, State & Local Tax, Public Sector) live on separate subdomains — not under mcneeslaw.com. The main site's /blogs/ page merely links out to these external domains. This means AI crawlers indexing mcneeslaw.com never encounter the substantive blog content directly. Practice area pages (e.g., Energy and Environmental) are text-rich (3,900+ words) with detailed sub-navigation, but the homepage itself is thin at 336 words and lacks an H1 heading — only H3s for "News," "Events," "Blogs," and "Best Places to Work in PA."

External Signals

The firm has a BBB profile linked site-wide. No significant Reddit threads, press coverage, or review-site presence surfaced in search. The cold-knowledge model recalled Chambers USA rankings for energy/environmental law and Marcellus Shale litigation — but the site itself does not prominently feature Chambers rankings or case highlights on the homepage or practice pages.

Findings

1. Cold-knowledge representation is outdated, missing offices and attorney count High

   A frontier LLM queried cold describes McNees as having 'over 140 attorneys' with offices only in Harrisburg, Lancaster, State College, and Scranton, omitting Pittsburgh, Radnor, Reading, Plymouth Meeting, York, Columbus, Frederick, Towson, and Washington, DC. The model also misses the firm's 170+ attorney count and practices like Esports, NIL, Gaming & Sports Wagering, and ESG.

   What to change: Publish an llms.txt file with accurate firm facts and update structured data to reinforce current office locations and practice areas.

2. Office location pages lack LocalBusiness and GeoCoordinates schema High

   The 13 office location pages (e.g., /office/harrisburg/, /office/columbus/) carry no LocalBusiness or GeoCoordinates structured data, reducing their visibility for AI-driven location queries.

   What to change: Add LocalBusiness and GeoCoordinates JSON-LD to each office page with accurate address, phone, and geolocation data.

3. llms.txt file returns 404, missing opportunity to guide AI Medium

   The site does not provide an llms.txt file, which could supply AI crawlers with accurate firm facts and preferred URLs, helping correct the cold-knowledge gap.

   What to change: Create an llms.txt file listing key firm facts, office locations, practice areas, and preferred URLs for AI consumption.

4. Blog content hosted on separate subdomains, not under mcneeslaw.com Medium

   The four topical blogs (Labor & Employment, Land Use, State & Local Tax, Public Sector) live on separate subdomains (e.g., palaborandemploymentblog.com). The main site's /blogs/ page only links out, so AI crawlers indexing mcneeslaw.com never encounter the substantive blog content directly.

   What to change: Consolidate blog content under the main domain or ensure cross-linking with proper canonical tags to pass authority.

5. Homepage is thin at 336 words and lacks an H1 heading Medium

   The homepage contains only 336 words and uses H3 headings for sections like 'News' and 'Events' without a single H1, which weakens topical signals for AI crawlers.

   What to change: Expand homepage content to at least 500 words and add a clear H1 heading that describes the firm's value proposition.

6. Blog LegalService schema references non-resolving domain mwn.com Medium

   The Pennsylvania Labor & Employment Blog includes a LegalService schema with url set to http://www.mwn.com/, which does not resolve, potentially confusing AI crawlers.

   What to change: Update the LegalService schema url to https://www.mcneeslaw.com/.

7. No Attorney or LocalBusiness schema on any page Medium

   Despite consistent Organization and WebSite schema, no page uses Attorney or LocalBusiness schema, which are standard for law firm sites and help AI understand the firm's professional services.

   What to change: Add Attorney and LocalBusiness schema to relevant pages, including attorney profiles and office pages.

8. Robots.txt has no AI-bot directives, leaving access fully open Low

   The robots.txt file contains zero AI-bot directives; the wildcard rule only disallows /wp-admin/ and /wp-includes/. While this allows access, it also means no guidance is provided to AI crawlers about which content to prioritize or avoid.

   What to change: Consider adding specific directives for AI crawlers to prioritize key pages and avoid low-value content.

9. Limited external signals from reviews, press, or social mentions Low

   Web searches for reviews, Reddit mentions, and press coverage returned zero results, indicating low external visibility that could affect AI citation and ranking.

   What to change: Encourage client reviews on Google and legal directories, and pursue press coverage or thought leadership to build external signals.

What's working

• All major AI crawlers receive full HTML content without blocking — Every major AI crawler (GPTBot, ClaudeBot, PerplexityBot, etc.) receives a 200 response with identical byte content to a browser visit, with no UA-based blocking or JS shells.
• Consistent Organization and WebSite JSON-LD across pages — The homepage and subpages carry a consistent Organization + WebSite + WebPage JSON-LD graph with sameAs links to social profiles, providing a solid foundation for entity recognition.
• Practice area pages are text-rich with detailed content — Practice area pages like Energy and Environmental contain over 3,900 words with detailed sub-navigation, providing substantial content for AI indexing.
• 14 sitemaps index ~3,000 URLs with proper organization — The site uses Yoast SEO to generate 14 sitemaps covering pages, posts, practice areas, offices, and people, ensuring broad crawl coverage.
• BBB profile linked site-wide as an external trust signal — The site includes a link to the firm's BBB profile, providing a verified external signal that AI engines can use for trust assessment.
• Server includes HSTS and X-Frame-Options for security — The site sets HSTS and X-Frame-Options SAMEORIGIN headers, indicating good security practices that indirectly support trustworthiness for AI.