AI Site Grade
speakindeacon.com — AI Site Grade
DataDome captcha wall blocks every AI crawler, making Deacon Jones Auto Group's $2B dealership network invisible to LLMs despite a rare llms.txt file.
SpeakinDeacon.com serves a 403 captcha wall to all AI crawlers, has no AI-bot directives in robots.txt, and its AutoDealer schema omits telephone, hours, and multi-location data, resulting in zero external web presence and a cold-knowledge gap where LLMs describe the 40-year-old, $2B enterprise as potentially 'newly established.'
- Findings
- 10
- Evidence checks
- 30
- Completed
- 30 May 2026
Analysis
DataDome Captcha Wall Blocks Every AI Crawler — Site Is Invisible to LLMs
The domain speakindeacon.com is the digital storefront for Deacon Jones Auto Group, a family-owned North Carolina dealership network operating 23 locations across 16 brands (Buick, Cadillac, Chevrolet, Chrysler, Dodge, Ford, GMC, Kia, Jeep, Lincoln, Nissan, Ram, Honda, Hyundai, Toyota, BMW) with over $2 billion in annual sales and nearly 500 employees. Yet every AI crawler — GPTBot, ClaudeBot, Google-Extended, PerplexityBot, OAI-SearchBot, Applebot-Extended, Bytespider — receives a 403 captcha wall served by DataDome (geo.captcha-delivery.com). The homepage returns a full 291KB JS-rendered page to browsers but only an 8-byte "Please enable JS" shell to bots. The robots.txt contains no AI-bot directives whatsoever — no mention of GPTBot, ClaudeBot, or any AI crawler — and the llms.txt file exists (a rare positive signal for an auto dealer) but is itself behind the same captcha wall for non-browser requests.
Cold-Knowledge Gap
A frontier LLM queried cold about "SpeakinDeacon" returned: "I do not have any verified or specific information about a car dealership named 'SpeakinDeacon.'" The model described it as potentially "a very small, independent, or newly established dealership." In reality, Deacon Jones Auto Group is a 40-year-old, $2B+ enterprise with 23 locations across North Carolina and Virginia — a scale that should be well-documented in training data. The gap between the brand's actual size and its total invisibility to AI knowledge is extreme. The domain name "speakindeacon.com" (a colloquial tagline) likely contributes to the disconnect; the operating brand name "Deacon Jones Auto Group" appears only in page titles and schema, not in the domain itself.
Schema Posture
Every page carries a single AutoDealer JSON-LD schema block with the same minimal fields: name, address (Smithfield NC), URL, geo coordinates, and a $ price range. The telephone field is empty on every page checked. No openingHoursSpecification, aggregateRating, review, makesOffer, areaServed, or department properties are present — surprising for a group with 23 distinct physical locations, each with its own phone number and hours listed in the visible HTML. The schema does not reflect the multi-location reality of the business.
External Signals
Web searches for "SpeakinDeacon," "Deacon Jones Auto Group," and related terms returned zero results across multiple queries — no reviews, no Reddit threads, no press mentions, no dealer directory listings surfaced. The site has no Wayback Machine snapshot history. The llms.txt file is well-structured (title, description, key page links with summaries) and linked from the footer, which is an advanced practice for a dealership — but its value is nullified by the captcha wall blocking AI crawlers from reaching any linked page. The homepage lists 23 locations across NC and VA, yet the AutoDealer schema references only the Smithfield address, creating a fragmented representation for any AI engine that does manage to parse the page.
Findings
DataDome captcha wall blocks every AI crawler High
Every AI crawler tested (GPTBot, ClaudeBot, Google-Extended, PerplexityBot, OAI-SearchBot, Applebot-Extended, Bytespider) receives a 403 captcha wall served by DataDome. The homepage returns only an 8-byte 'Please enable JS' shell to bots, while browsers get a full 291KB page.
What to change: Remove the captcha wall for AI crawlers or whitelist known AI bot user agents to allow access to the site's content.
robots.txt contains no AI-bot directives High
The robots.txt file has only one user-agent rule and does not mention GPTBot, ClaudeBot, or any other AI crawler, leaving them without explicit guidance.
What to change: Add explicit allow/disallow directives for AI crawlers (e.g., GPTBot, ClaudeBot, Google-Extended) in robots.txt.
llms.txt file exists but is inaccessible to AI crawlers High
The site has an llms.txt file (a rare positive signal for an auto dealer), but it is served behind the same captcha wall for non-browser requests, rendering it useless for AI crawlers.
What to change: Ensure llms.txt is accessible to AI crawlers without captcha challenges.
LLMs have no knowledge of the dealership High
A frontier LLM queried cold about 'SpeakinDeacon' returned no verified information and described it as potentially 'a very small, independent, or newly established dealership,' despite the brand being a 40-year-old, $2B+ enterprise with 23 locations.
What to change: Improve AI visibility by allowing crawlers, adding structured data, and building external citations (press, directories, reviews).
Zero external web presence for the brand High
Web searches for 'SpeakinDeacon,' 'Deacon Jones Auto Group,' and related terms returned zero results across multiple queries — no reviews, Reddit threads, press mentions, or dealer directory listings.
What to change: Build external citations by claiming dealer directory listings, encouraging reviews, and generating press coverage.
AutoDealer schema has empty telephone field Medium
Every page carries a single AutoDealer JSON-LD block with an empty telephone property, despite phone numbers being present in the visible HTML.
What to change: Populate the telephone field in the AutoDealer schema with the correct phone number.
AutoDealer schema omits hours, reviews, and multi-location data Medium
The schema lacks openingHoursSpecification, aggregateRating, review, makesOffer, areaServed, and department properties. It references only the Smithfield address, ignoring the other 22 locations.
What to change: Expand the AutoDealer schema to include openingHoursSpecification, aggregateRating, makesOffer, and separate department or place entries for each location.
New vehicles page returns 403 to all clients Medium
The /newvehicles page returns a 403 error even to browser requests, making inventory invisible to both users and crawlers.
What to change: Fix the /newvehicles page to return a 200 status with content for all clients.
No accessible sitemap found Medium
The sitemap URL returned 200 but contained zero URLs, and no other sitemap was discovered.
What to change: Generate and submit a valid sitemap with all public URLs to search engines.
Domain name does not match operating brand Low
The domain 'speakindeacon.com' is a colloquial tagline, while the operating brand is 'Deacon Jones Auto Group.' This mismatch likely contributes to the cold-knowledge gap.
What to change: Consider using a domain that matches the operating brand, or ensure strong brand signals (schema, content) tie the domain to 'Deacon Jones Auto Group.'
What's working
- llms.txt file exists and is well-structured — The site has an llms.txt file with title, description, and links to key pages with summaries, an advanced practice for a dealership that signals AI-readiness.
- AutoDealer schema present on every page — Every page carries a single AutoDealer JSON-LD block with name, address, URL, geo coordinates, and price range, providing a baseline structured data foundation.
- Informative content pages with rich text — Pages like /our-story, /about-us, /why-us, /our-community, and /reviews contain substantial, well-written content about the dealership's history, values, and community involvement.
Track speakindeacon.com across AI search
This is one snapshot. Open the interactive report to inspect evidence, or grade another site free.