Every GAIO Score we publish is the output of a standardized, repeatable audit process — not an opinion. This page explains exactly how we collect data, what we measure, and how we calculate the numbers you see in our reports.
What We Measure
AI visibility is not a single metric — it is the intersection of three distinct signals. AngryRobot Scanner measures all three and combines them into a single GAIO Score.
Structured data, schema markup, crawlability, PageSpeed, Core Web Vitals. The technical signals that allow AI engines to parse and understand your site.
Citation rate, brand mention frequency, query coverage across Gemini and Perplexity. How often AI engines reference your brand when answering relevant queries.
A weighted composite of SEO and GEO signals, calibrated against category benchmarks. Your single number for AI visibility readiness.
How We Collect Data
Step 1 — Technical Crawl
We crawl the target domain using a headless browser, extracting structured data (JSON-LD, OpenGraph, schema.org types), HTTP headers, canonical tags, PageSpeed metrics via Google Lighthouse, mobile usability signals, and internal link structure. This takes between 43 seconds and 4 minutes depending on site complexity.
Step 2 — AI Engine Queries
We run a standardized test suite of localized search queries against Gemini and Perplexity — the two AI engines with the highest market penetration in European markets. Queries are designed to reflect real user intent across three categories:
- Brand queries — direct brand name lookups and reputation queries
- Category queries — generic product/service category searches where the brand could appear
- Competitor comparison queries — "best X in [country]" style prompts
We record whether the brand is cited, how prominently, and what sentiment is attached to the citation.
Step 3 — AI Perception Analysis
Six AI perception dimensions are scored independently using structured prompts across both engines: Digital Presence, Innovation, Market Leadership, Future Readiness, AI Adoption, and Brand Sentiment. Scores are normalized to a 0–100 scale and averaged across engine responses.
Step 4 — Score Calculation
The GAIO Score is a weighted composite:
GEO is weighted higher because AI engine citation is the emerging competitive frontier — technical SEO is table stakes. Scores are then benchmarked against category peers to produce a relative ranking.
AI Engines We Query
We do not query ChatGPT directly. Our engine selection is reviewed quarterly and updated as market share data evolves.
How to Interpret Scores
Honest Limitations
- AI engine responses vary. We run multiple queries and average results, but individual responses can differ.
- Scores reflect the state of your site and AI engine training data at the time of the scan — they are not static.
- We scan the homepage by default on single-page scans. Multi-page plans scan deeper content layers.
- PageSpeed data is collected under standardized network conditions and may differ from real-world user experience.
This methodology page describes the AngryRobot Scanner audit process as of July 2026. Methodology is reviewed and updated quarterly. For questions, contact us at scan.angryrobot.eu.