What Is an AI Visibility Score? The Three Pillars Explained
When we talk about AI visibility, we're not talking about search rankings, domain authority, or keyword positioning. We're talking about something more fundamental: can AI systems accurately identify and represent your business?
The AI Visibility Score is a single number — 0 to 100 — that answers that question. It measures the completeness, consistency, and accessibility of the signals that AI systems use to understand who you are. And unlike most scores in the digital marketing world, every point is fully explainable.
Why a new score matters
The digital landscape already has plenty of scores. SEO tools offer domain authority, page authority, keyword difficulty. Google provides Core Web Vitals. Various platforms offer "brand health" metrics. None of them measure what AI systems actually need.
Search engine optimisation is fundamentally about ranking. AI visibility is fundamentally about identity. A website can rank on page one of Google for its target keywords and still be completely misrepresented by ChatGPT, because the signals that drive search rankings are different from the signals that drive AI understanding.
How the score works
The scoring system is deliberately simple in principle, even though the evidence collection behind it is comprehensive. Your total score is the weighted sum of three pillar scores. Each pillar starts at 100 and is reduced by specific, named penalties called reason codes. Every reason code maps to an observable, binary signal — something is either present or it isn't.
Total Score = (Identity Clarity × 0.40) + (Structural Support × 0.30) + (Evidence Consistency × 0.30)
Pillar 1: Identity Clarity (40%)
Identity Clarity carries the highest weight because it answers the most fundamental question: does your website explicitly declare who you are in a format machines can read?
This pillar evaluates the JSON-LD structured data on your homepage — Organization schema, sameAs properties, logo declaration, and Website schema. When any of these signals is missing, a specific reason code fires with a defined point penalty.
| Signal | Condition | Reason Code | Penalty |
|---|---|---|---|
| Organization schema | Missing | ID_NO_ORG_SCHEMA | −40 |
| sameAs properties | Missing | ID_NO_SAMEAS | −25 |
| Logo declaration | Missing | ID_NO_LOGO | −10 |
| Website schema | Missing | ID_NO_WEBSITE_SCHEMA | −15 |
Pillar 2: Structural Support (30%)
Structural Support evaluates whether AI crawlers can actually reach and parse your content. You might have excellent structured data, but if your robots.txt blocks crawlers, your sitemap is missing, or your HTML is malformed, AI systems may never see it.
| Signal | Condition | Reason Code | Penalty |
|---|---|---|---|
| JSON-LD present | Missing | STRUCT_NO_JSONLD | −50 |
| robots.txt | Blocking | STRUCT_ROBOTS_BLOCKING | −20 |
| Sitemap | Missing | STRUCT_NO_SITEMAP | −15 |
| HTML validity | Invalid | STRUCT_INVALID_HTML | −15 |
Pillar 3: Evidence Consistency (30%)
Evidence Consistency looks beyond your website. It checks whether external sources — Google's search results, business directory listings, rich results — corroborate the identity your website claims. This pillar uses evidence from three Google surfaces queried with your brand name in your specific geographic market.
| Signal | Condition | Reason Code | Penalty |
|---|---|---|---|
| Brand in SERP | Not found | EVID_NO_BRAND_SERP | −35 |
| Knowledge panel | Not present | EVID_NO_KP | −20 |
| Rich results | Not detected | EVID_NO_RICH_RESULTS | −15 |
| NAP consistency | Inconsistent | EVID_INCONSISTENT_NAP | −20 |
Risk labels
The total score maps to a plain-language risk label that helps non-technical stakeholders understand the urgency.
Critical (0–39)
AI systems have almost no reliable identity data. Misrepresentation is likely.
High (40–59)
Significant gaps exist. AI may represent your business inaccurately or incompletely.
Medium (60–79)
Foundation exists but gaps remain. Some AI representations may be incomplete.
Low (80–100)
Strong AI visibility. Identity is well-established across signals and sources.
What makes this different
Three properties distinguish the AI Visibility Score from other metrics in the market.
First, it's deterministic. The same inputs always produce the same score. There's no model drift, no probabilistic weighting that changes over time, no opaque algorithm. If your score is 62, you can trace exactly which reason codes are active and calculate the number yourself.
Second, it's explainable. Every deduction has a name, a pillar, and a specific point value. There's no "improve your score" without actionable specifics. Each reason code maps directly to a remediation action.
Third, it's geo-scoped. Your score reflects what AI sees in your market, not globally. A business in Sydney might have different evidence consistency than the same business queried from San Francisco, because search results and business listings vary by location.