Methodology

How PARCEIT Measures AI Search Visibility

The P-Score is a 0-100 rating based on a real crawl of your website. It is not a guess, not a survey, and not a prompt-tracking estimate. Every point is tied to a specific check with a specific pass/fail result. This page documents exactly what we measure, how we score it, and how it maps to the IAB framework for AI visibility measurement.

The Four Signal Categories (P-Score)

The P-Score totals 100 points across four categories. Each category measures a different layer of the pipeline that determines whether AI engines can find, understand, and cite your content.

25 pts

Crawlability

IAB: Presence

Can AI crawlers reach your site and read its content? This is the gate. If your robots.txt blocks GPTBot, ClaudeBot, PerplexityBot, or Google-Extended, your site does not exist in those engines. No amount of content optimization changes this.

What we check

  • +Site reachable (HTTP 2xx response)
  • +robots.txt present and correctly configured
  • +AI crawler access for critical bots (GPTBot, ClaudeBot, PerplexityBot, Google-Extended)
  • +Valid XML sitemap submitted
  • +llms.txt file for LLM crawlers
  • +Live crawler accessibility verification (WAF/firewall detection)

Maps to IAB: The IAB's Presence metrics (Mention Rate, Citation Rate, Share of Voice) are all zero when crawlability fails. Parceit's Crawlability score is the leading indicator of Presence.

25 pts

Entity

IAB: Prominence + Portrayal

Can AI engines understand who you are? Entity data is the structured data (JSON-LD schema) that tells AI engines your business name, type, location, and relationships. Without it, the AI has to guess. With it, there is no ambiguity.

What we check

  • +Structured data present (JSON-LD blocks detected)
  • +Organization or Business entity type identified
  • +sameAs links to external profiles (social, directories)
  • +GeoCoordinates or location data
  • +Aggregate rating markup

Maps to IAB: Entity data feeds the IAB's Prominence (placement and ranking order) and Portrayal (accuracy, framing, hallucination rate). Inconsistent entity data across the web increases hallucination likelihood.

30 pts

Structural

IAB: Prominence

Is your HTML built so AI engines can parse it? Structural signals are the technical plumbing: title tags, meta descriptions, heading hierarchy, semantic HTML5 elements, and social card markup. These determine whether an AI engine can extract clean information from your pages.

What we check

  • +Title tag present (10+ characters)
  • +Meta description (50+ characters)
  • +Canonical URL tag
  • +Open Graph tags (title, description, image)
  • +Exactly one H1 tag
  • +H2 headings for content structure (2+ required)
  • +Semantic HTML5 elements (nav, main, footer, article)
  • +Twitter card tag

Maps to IAB: Structural signals feed the IAB's Prominence metric. Pages with clean semantic HTML and proper heading hierarchy are easier for AI engines to parse, extract, and cite prominently.

20 pts

Content

IAB: Prominence + Persuasion

Does your content have the structured markup AI engines need to build cited answers? FAQPage schema, Article schema, and content depth signals determine whether an AI engine can extract self-contained answers from your pages.

What we check

  • +FAQPage schema (question-answer pairs for AI extraction)
  • +Article, HowTo, or BlogPosting content schema
  • +Content depth (3+ H2 sections as proxy)
  • +Multiple structured data types (4+ for richness)

Maps to IAB: Content structure feeds Prominence (substantive citation vs superficial mention) and Persuasion (whether the extracted answer is specific enough to drive a click-through).

How the P-Score Maps to the IAB 4 P's

On August 3, 2026, the IAB published “Measuring Visibility in the AI Era,” introducing the 4 P's of AI Visibility. The IAB framework defines what to measure at the outcome level. PARCEIT's P-Score measures the foundational signals that determine whether those outcomes are possible.

Presence

Does the brand appear in an AI response?

Crawlability (25 pts)

If AI crawlers are blocked in robots.txt or by a WAF, the site is invisible. Presence metrics are zero regardless of content quality. Crawlability is the prerequisite gate.

Prominence

Where and how prominently does the brand appear?

Structural (30 pts) + Entity (25 pts) + Content (20 pts)

Clean semantic HTML, proper heading hierarchy, JSON-LD entity data, and FAQPage schema determine whether AI engines can extract and prominently cite your content.

Portrayal

In what context, and with what accuracy?

Entity (25 pts)

Consistent entity data (NAP, sameAs links, organization type) reduces hallucination risk. Inconsistent data across the web increases the likelihood AI engines generate false information about your business.

Persuasion

Does AI visibility drive action?

All four signals combined

A user clicks through from an AI citation when the answer is accurate, specific, and trustworthy. That requires the full pipeline: crawlable content, clear entity identification, structured answers, and factual accuracy.

Directional vs Decision-Grade: Why Methodology Matters

The IAB introduced a two-tier quality standard for AI visibility data. Here is how Parceit fits.

Directional

Identifies patterns and signals trends. Useful for early signal detection and competitive awareness. Not sufficient for budget allocation or executive strategy decisions.

Not what we sell

Prompt-tracking (typing queries into ChatGPT and counting mentions) is directional measurement. Sample sizes are small, results are not reproducible, and methodology is opaque.

Decision-Grade

Meets a higher standard of rigor across query volume, sample size, prompt type coverage, testing cadence, reproducibility, and platform coverage. Required before making budget or strategy decisions.

What Parceit delivers

Parceit's audit is reproducible. The same site produces the same score every time. The methodology is published on this page. Every check is a pass/fail with a specific point value. No black boxes.

The Scoring Formula

The P-Score is the sum of points earned across all four categories. Every check has a fixed point value. A check either passes (full points) or fails (zero points). Some checks award partial credit (semantic HTML, sameAs links, AI crawler access) based on how many sub-conditions are met.

P-Score = Structural (30) + Entity (25) + Content (20) + Crawlability (25)

Maximum score: 100. No curve, no normalization, no proprietary algorithm. The same site produces the same score every time.

90-100

Excellent

Your site is well-optimized for AI search visibility. Minor improvements may increase citation prominence.

70-89

Good

Solid foundation with identifiable gaps. Fixing failed checks will improve AI engine access and entity recognition.

50-69

Needs Work

Significant gaps that limit AI visibility. AI engines may find your site but struggle to parse or cite it accurately.

0-49

Critical

AI engines probably cannot find or read your site. Likely causes include blocked crawlers, missing structured data, and non-semantic HTML.

See Your P-Score in 60 Seconds

Run a free audit at parceit.com. No signup required. You see every check, every score, and every issue.

Run my free audit