Structural { authority_blueprint }

Mike's Auto Body & Paint Inc.

April 30, 2026 - Structural Authority Blueprint

Proprietary Disclosure & Protocol

Executive Summary

The Bottom Line

Mike's Auto Body & Paint Inc. has been serving its market for decades with earned authority (certifications, warranties, reviews). The business has real, durable credibility in the market.

The problem is that AI systems - the same tools your customers now use to find and choose you - cannot see most of that authority clearly enough to recommend it with confidence. When a potential customer asks ChatGPT, Gemini, or Perplexity who to trust, the business is frequently absent from the answer, misidentified, or confused with a different entity.

This audit identifies the technical and visibility conditions most likely contributing to that behavior and defines the practical path to reduce those risks.

Current Score: 35 / 100 - Emerging

What this means in plain language:

The Three Critical Findings

1. Template residue and source-layer noise - 32 instances of raw placeholder tokens were detected in the live source HTML across audited pages. Hidden scaffold is also present in the source. Together, these conditions add structural noise that can reduce machine trust in the site as a citation source.
2. Missing structured data - No JSON-LD schema exists on key authority pages. This leaves key identity, location, founding-date, and certification facts dependent on unstructured text and third-party sources.
3. Identity fragmentation in the wild - Live AI testing showed geographic conflation, namesake confusion, and hallucinated founding dates.

The Opportunity

Once the technical blockers are removed, the structural foundation exists to unlock real authority and make those signals machine-readable and persistently citable.

Recommended Immediate Action

Lock the Phase 1 remediation scope this week. Remove template leakage. Deploy organization and service schema. Harden source-of-truth signals. These three coordinated actions, completed within 45–60 days, are the prerequisite for everything else.

GEO Scorecard

Maturity Band Reference

Pillar Diagnosis

The Citeable Systems model evaluates four structural pillars that determine whether a website functions as a reliable, machine-readable source of truth.

Entity Anchoring - Adjusted: 44 | Raw: 70

Entity Anchoring measures whether AI systems can identify this business as a unique, stable, and locatable entity — separate from other businesses with similar names. A score of 44 means partial identity signals exist but are not strong enough to prevent conflation with other entities.

During live testing, Gemini conflated Mike's Auto Body in Fort Myers with a shop in Spring Hill, over 130 miles away. The absence of structured location schema is consistent with this kind of geographic ambiguity and likely contributes to the risk of misrepresentation.

Factual Specificity - Adjusted: 58 | Raw: 80

Factual Specificity is the strongest pillar in this audit, reflecting real content on the site. The facts exist — founding date, certifications, service descriptions — but only in natural language copy, not in machine-readable structured data.

This leaves AI systems to infer facts from text and third-party records, which increases the risk of hallucination. The confirmed founding year of 1981 is present on the site, but live tests show ChatGPT returning 1989 and Gemini returning 2009.

Logical Citeability - Adjusted: 40 | Raw: 74

Logical Citeability measures whether the site is structured so that AI systems can isolate a piece of information, verify it, and link back to the page as a trustworthy source. With 32 instances of raw placeholder tokens across all 8 audited pages, plus hidden Yext/editor scaffold present in the source HTML, the site appears structurally noisy to machine readers.

AI systems deprioritize noisy sources because citing them risks surfacing unreliable or malformed information. Noisy source conditions can make AI systems more likely to rely on third-party sources such as Yelp, Carwise, or competitor pages instead of citing mikesautobodyandpaintinc.com directly.

Source Sovereignty - Adjusted: 43 | Raw: 100

The raw score of 100 confirms that all three machine-readable artifacts are present: robots.txt, sitemap.xml, and llms.txt. This is a genuine strength. However, the adjusted score drops to 43 because those artifacts are negated by the template pollution and schema gaps on the same pages.

Audit Scope and Evidence Sources

What Was Audited

Evidence Source:

• Live Query Audit: Gemini, Perplexity, and ChatGPT platform-level testing.

Review Boundary: This report reflects evidence available during the audit window. Live query results represent platform behavior at the time of testing. AI system behavior changes over time and may differ in subsequent tests.

Top Forensic Findings

The most score-relevant issue in this finding is the presence of unresolved placeholder tokens such as {{placeholder_retargeting_pixel}}, {{placeholder_dpni}}, and {{placeholder_footer_reserve1}}. These are hidden from normal visitors but fully visible in the raw HTML read by crawlers, AI indexers, and citation systems.

Separately, the live source also contains hidden Yext/editor scaffold blocks such as editorOnlyInfo and wrongScriptError. Those blocks are not the same as valid Yext business-data fields. Address, phone, hours, and email hooks delivered through Yext are normal. The issue is the additional hidden scaffold and unresolved placeholder residue left in production HTML.

Together, these conditions create source-layer noise. The raw placeholder tokens are the strongest evidence of template leakage and remain the main suppressor in scoring. The hidden Yext/editor scaffold strengthens the cleanliness concern but should not be confused with legitimate business data integration.

JSON-LD (JavaScript Object Notation for Linked Data) is the mechanism by which websites communicate machine-readable facts directly to AI and search systems. It is a structured identity card that tells a machine who the business is, what it does, where it is located, what certifications it holds, and what evidence supports its authority claims.

• Gemini: Conflated the Fort Myers shop with a Mike's Auto Body in Spring Hill / Brooksville (Pasco County). Directly told Fort Myers users to visit a different shop. Root cause: no machine-readable location anchoring.
• Perplexity: Borrowed certification and service data from mikesautobodynorth.com — a different shop in Massachusetts with a similar name. Corrected the location but described the wrong business. Root cause: identity signals too thin to distinguish the Fort Myers entity.
• ChatGPT: Returned 1989 as the founding year — erasing 8 years of seniority. The only platform to read the Carwise review volume (1,500+) but still lacked a stable historical anchor. Root cause: founding year not encoded in schema.

Position Authority Interpretation

Position Authority is how this report defines the relationship between what Mike's Auto Body actually is in the market and what AI systems currently believe, represent, and communicate about it. The gap between those two things is the strategic problem this audit exists to solve.

What the Market Actually Is

What the AI Systems Currently See

Across three major AI platforms, the live query results reveal a consistent and damaging pattern:

• Category searches ("best auto body shop in Fort Myers") — Mike's is absent. Competitors with fewer years of operation and fewer reviews are consistently recommended instead.
• Direct lookup queries ("tell me about Mike's Auto Body") — Mike's is mentioned, but the facts returned are partially or significantly wrong. Founding dates are hallucinated. Location is misrepresented. Credentials are borrowed from other businesses.
• Comparison prompts ("Mike's vs. Beach Fender Mender", "Mike's vs. Maaco") — performance is mixed. Against Maaco, recognition is strong. Against local competitors, results are inconsistent and unreliable.

The Five Visibility Failure Modes

What Needs to Change

• Phase 1 — Remove the blockers: Eliminate template leakage. Deploy structured data. Harden the owned-site source-of-truth layer. This stops the active damage and creates a stable foundation.
• Phase 2 — Encode the authority: Make the 1981 founding date, the I-CAR certification, the lifetime warranty, and the Carwise review volume machine-readable and persistently citable. Turn dark data into structured proof.
• Phase 3 — Sustain and defend: Monitor AI platform behavior, track citation patterns, and adapt as platforms evolve. Maintain the competitive authority gap that Phases 1 and 2 create.

Live Query and Prompt Result Analysis

Platform-Level Observations

Gemini

Gemini's behavior represents the most severe visibility failure observed in this audit. In the Beach Fender Mender comparison prompt, Gemini conflated Mike's Fort Myers location with a Spring Hill / Brooksville-area shop and told users in Lee County to choose Beach Fender Mender. In other tested responses, Gemini attributed experience and certification-style trust signals to Absolute Collision while omitting Mike's equivalent signals.

Perplexity

Perplexity identified the correct geographic location but used service and certification details from a Massachusetts business with a similar name. This is consistent with weak entity disambiguation: the platform found stronger or more accessible corroborating data for a namesake and used it to fill gaps.

ChatGPT

ChatGPT demonstrates both the problem and the opportunity clearly. It surfaced the Carwise review volume in some tests, showing that the signal can be found. But it also returned a 1989 founding year in another result, suggesting that Mike's historical authority is not consistently anchored across the sources ChatGPT used.

Phase 1 Action Plan

Phase 1 remediates the two critical blockers that are actively suppressing trust and preventing reliable AI citation. These three actions are the prerequisite for all authority-building work. They do not amplify authority — they remove what is currently suppressing it.

Remove Template Residue and Harden Page Architecture (30-45 days)

Problem: 32 instances of unresolved placeholder template variables are present across audited pages. Actions:

• Audit the CMS template system and identify the source of unfilled placeholder variables.
• Remove or correctly fill all detected {{placeholder}} token instances.
• Remove or suppress hidden editor/debug scaffold from production HTML.

Deploy Organization and Service Schema (45-60 days)

Problem: No JSON-LD schema on 6 key authority pages. Actions:

• Deploy LocalBusiness schema on the homepage with NAP, founding year, hours, and geo-coordinates.
• Add Service schema to collision repair and auto body pages.
• Encode certifications and review counts in structured data.
• Add Organization schema at the site level to stabilize entity identity.

Harden Source-of-Truth Signals (45-60 days)

Problem: The owned site declares itself as authoritative but cannot yet be trusted consistently. Actions:

• Review and update the llms.txt file to reflect correct descriptions and locations.
• Verify robots.txt allows citation-relevant crawlers.
• Ensure sitemap.xml is current.
• Confirm NAP data is consistent across all pages and matches schema declarations.

Measurement Plan

Progress should be measured against three questions: Were the blockers actually removed? Did the owned-site trust improve? Did those changes produce more consistent AI citation behavior in live tests?

Primary Metrics

Retest Cadence

• Phase 1 validation retest: 30-45 days after all three actions are confirmed complete.
• Full prompt pack retest: 60 days post Phase 1 completion.
• Quarterly ongoing monitoring: Track score movement and citation behavior every 90 days.

Validation Checklist Before Next Report

Recommended Next Steps

Phase 1: 30–60 Days

• Remove or suppress hidden Yext/editor scaffold from production HTML where feasible.
• Preserve valid Yext-delivered business data fields during cleanup.
• Deploy LocalBusiness, Service, and Organization schema across homepage and all authority pages.
• Encode 1981 founding year, I-CAR certification, lifetime warranty, and geo-coordinates in structured data.
• Update llms.txt with accurate business description and service list.
• Run validation retest at 30 days to confirm blocker removal.

Phase 2 Preview: 60–120 Days

Phase 2 shifts from removing suppressors to encoding the full authority layer:

• Build a review aggregation schema to surface the 1,500+ Carwise reviews as machine-readable social proof.
• Add FAQPage schema to high-traffic service pages to capture conversational query formats.
• Encode insurance partnership and OEM affiliation signals if applicable.
• Develop an authority content layer that supports third-party citation and link-building.

The Strategic Horizon

The current validated remediation blueprint has a realistic near-term ceiling of 52-54 on the Citeable Systems score: the Grounded band. That is enough to make the site structurally cleaner and materially easier for AI systems to identify, summarize, and cite. Moving beyond that range toward Citeable (65+) would require an expanded authority program beyond the current validated blueprint: deeper review-proof encoding, stronger external corroboration, sustained citation monitoring, and ongoing content/entity reinforcement.

Appendix A: Citeable Systems Scoring Method

The Four Pillars

• Entity Anchoring: Can AI systems identify this as a unique, stable business entity?
• Factual Specificity: Are concrete, verifiable facts encoded in machine-readable form?
• Logical Citeability: Is the page structure clean and trustworthy enough for AI citation?
• Source Sovereignty: Does the owned website function as the primary authority?

Risk-Adjusted Scoring - The Suppressor Effect The score is not a simple count of what is present. Severe technical issues - called Suppressors - reduce the effective score even when positive signals exist elsewhere. Common suppressors include unresolved placeholder residue and missing JSON-LD schema.

Structural Authority Blueprint - client-shared