How we calculate the Site Health Score.

The overall Site Health Score is a weighted average of five category scores. Weights reflect what actually moves dealer revenue based on our internal data and Google's published ranking signals.

• ● Foundational Health: 25%. Domain configuration, sitemap hygiene, indexing, technical SEO. Without this, nothing downstream matters.
• ● Performance: 20%. Core Web Vitals on mobile and desktop, third-party tag bloat. Direct input to Google Quality Score.
• ● Visibility: 25%. Schema markup, AI search visibility, local presence. The layer that decides whether you are surfaced when buyers ask.
• ● Funnel: 20%. Conversion architecture, inventory feed health, tracking integrity. Whether traffic turns into leads.
• ● Risk: 10%. ADA accessibility, FTC compliance, reputation signals. The category that can sue you, sink you, or shred your reputation.

Performance signals are scored against Google's official Core Web Vitals thresholds. We prefer real-user CrUX field data over Lighthouse synthetic data when available, because CrUX reflects what your actual buyers experience and what Google uses to rank you.

• ● Mobile LCP: under 2.5 seconds = 100 points. 2.5 to 4.0 seconds = linear interpolation. Over 4.0 seconds = 0 points.
• ● Mobile INP: under 200ms = 100 points. 200 to 500ms = linear interpolation. Over 500ms = 0 points.
• ● Mobile CLS: under 0.1 = 100 points. 0.1 to 0.25 = linear interpolation. Over 0.25 = 0 points.
• ● Third-party tag count: under 10 = 100. 10 to 20 = 70. 20 to 30 = 40. Over 30 = 10.

Foundational and Visibility signals follow a deduction model. The category starts at 100 and we deduct points for confirmed failures. We do not deduct for signals we could not verify (for example, schema rendered client-side that our static scanner cannot read). The methodology document for each specialty audit on the Audits dropdown details the exact deductions for that audit.

Every issue surfaced on a Site Health Score report includes an estimated monthly cost. These are calculated against the dealer's reported (or estimated default) ad spend and organic traffic value, multiplied by industry-typical waste rates. Sources cited at the bottom of each report.

• ● Core Web Vitals failure: 30% of monthly Google Ads spend (Source: Shift Digital, 2025).
• ● Vehicle schema missing on inventory: 5% of monthly organic value lost to invisibility in Google Vehicle Listings.
• ● Sitemap missing: 10% of monthly organic value lost to indexing degradation.
• ● Domain configuration failures: 4% of monthly ad spend lost to Quality Score throttling.
• ● Third-party tag bloat over 25 hosts: 8% of monthly ad spend lost to performance penalties.
• ● Form friction (over 5 fields above fold): 15% of monthly conversion value lost.

The Site Health Score report includes three benchmark reference numbers. They are sourced as follows:

• ● OEM average (typically 56): Calculated from our internal sample of franchise dealer sites we have audited, segmented by OEM. The number represents the median across the OEM.
• ● Top 10 percent (typically 81): The 90th percentile across our internal sample, OEM-blind.
• ● Top dealer we work with (typically 91): The highest-scoring dealer we have actively maintained on our diagnostic and remediation engagement, OEM-blind, anonymized.

Score labels are buckets, not letter grades. Each label corresponds to a recommended action.

• ● 90 to 100: Excellent. Maintain the current trajectory.
• ● 70 to 89: Healthy. Address the top issues but no urgency.
• ● 50 to 69: Below Average. Significant gaps that are likely costing real money. Run the Full Diagnostic.
• ● 30 to 49: Needs Attention. Multiple high-severity issues. Run the Full Diagnostic and consider Implementation.
• ● 0 to 29: Needs Immediate Attention. Foundational failures across categories. Implementation tier strongly recommended.

Honest disclosure. The current scan engine has known limitations. Sites that render JSON-LD schema entirely in JavaScript (common on Dealer.com platforms) cannot be verified by our static scanner. We mark these as client-rendered rather than missing, and the paid Full Diagnostic uses a headless browser to verify what Google actually sees. We will publish the headless verification methodology when that capability ships.