# airanking.hu

> Measure how AI assistants cite Hungarian companies. Analyze ChatGPT, Perplexity, and Google AI Overviews for accuracy and hallucination rates.

This site tracks how often and how accurately AI assistants mention Hungarian companies. It treats citability as an infrastructure problem rather than a marketing metric.

## Author

- Zoltán Varga — AI visibility researcher and AI ecosystem architect, Budapest: https://vargazoltan.ai/
- English profile: https://zoltanvarga.ai/en/about/
- ORCID: https://orcid.org/0009-0003-1020-834X
- arXiv author: https://arxiv.org/a/varga_z_1.html
- Zenodo: https://zenodo.org/search?q=owners%3A1653886

## Publications

- Per-Entity Bias Mapping for AI Visibility: Why Brand Mentions Require Entity-Specific Calibration (arXiv:2606.21595): https://arxiv.org/abs/2606.21595
- AI Visibility in the Generative Era: A Manifesto for Entity-Level Measurement (DOI 10.5281/zenodo.20285424): https://doi.org/10.5281/zenodo.20285424
- Quality Challenges Do Not Escalate Source Fabrication in Large Language Models (RICE-P1) (DOI 10.5281/zenodo.20474050): https://doi.org/10.5281/zenodo.20474050
- Rejection-Induced Confabulation Escalation — Pilot Study 2 (RICE-P2) (DOI 10.5281/zenodo.20509755): https://doi.org/10.5281/zenodo.20509755

## Key topics

- AI visibility index Hungary
- ChatGPT company mentions
- Perplexity citation accuracy
- Google AI Overviews brands
- LLM hallucination study
- AI audit Hungary
- GEO optimization
- AEO strategy
- brand hallucination paradox
- PEBM research
- AI visibility score
- machine readable entity
- AI citation audit
- zoltan varga ai
- Hungarian B2B AI
- synthetic search results

## Questions and answers

### Does AI visibility work like traditional SEO?

No, AI visibility is stochastic rather than deterministic. Traditional SEO targets logged impressions and clicks, while AI answers create no such logs. Visibility depends on whether a model can point to a machine-readable entity.

### What is the Brand Hallucination Paradox?

Well-known brands may receive more confident but false AI attributions. Familiarity gives models more material for plausible but unsupported completions. A pre-registered protocol tests if query type or source availability drives this.

### How many fake sources appear in AI answers?

In a study of 1,400 runs on 100 Hungarian B2B companies, 45.9% of cited sources were fabricated. Search-grounded models like Perplexity showed lower rates than non-search models.

### Can AI see all pages on my website?

Not necessarily. One case study showed all 33 pages returned HTTP 200 to bots, yet Google, Bing, and Brave indexed zero pages. Indexing remains a distinct bottleneck separate from access.

### What is the PEBM method?

PEBM stands for Per-Entity Bias Map, documented in arXiv:2606.21595. It measures bias and fabrication rates across different AI models and entity types. The study analyzed 100 Hungarian B2B companies over 1,400 runs.

### Why do AI models get dosage wrong?

Without search grounding, some models stated incorrect doses for agricultural products. With search enabled, all engines matched the officially approved doses exactly. This highlights the risk of reliance on internal training data.

## Related

- AI visibility expert page: https://vargazoltan.ai/ai-lathatosagi-szakerto/
- AI visibility audit: https://vargazoltan.ai/audit/
- Research: https://vargazoltan.ai/kutatas/

## Articles

- [How a National AI Visibility Index Is Built: Sampling, Scoring, and Aggregation](https://airanking.hu/articles/building-an-ai-visibility-index/): A national AI visibility index is built by sampling thousands of public domains, auditing them against specific dimensions like entity identification and structured data, and aggregating scores by sector. This process measures how well AI agents can read, verify, and cite a business rather than just how easily humans find it via traditional search.
- [Why a Single AI Visibility Score Misleads Your Strategy](https://airanking.hu/articles/why-a-single-score-misleads/): A single AI visibility score is insufficient because it collapses two opposite failure modes—invisibility and misattribution—into one number, hiding whether a company is unseen or actively hallucinated. Accurate comparison requires a per-entity diagnostic that separates citation frequency from factual accuracy and distinguishes between parametric memory gaps and retrieval-based errors.
- [The Per-Entity Bias Map: Calibrating AI Visibility](https://airanking.hu/articles/per-entity-bias-map/): A per-entity bias map is a diagnostic framework that calibrates the unique error profile of a single brand or organization across AI platforms before any ranking occurs. It distinguishes between verified mentions and false attributions to reveal hidden risks like the Brand Hallucination Paradox or parametric memory lag. This method replaces aggregate visibility scores with entity-specific error distributions to ensure fair and accurate comparisons.
- [Classifying Companies by Their AI Error Profile](https://airanking.hu/articles/ai-error-profile-classification/): Companies are grouped by AI error profiles using a ten-dimensional audit that measures retrieval inclusion, verification rates, and parametric-retrieval lag. This method distinguishes between invisible entities, those suffering from false attribution, and those facing structural infrastructure gaps. Classifying these profiles allows for targeted remediation rather than generic content optimization.
- [How to Sample Companies for a Fair AI Visibility Ranking](https://airanking.hu/articles/sampling-companies-for-ai-ranking/): Fair AI visibility rankings require stratified sampling that matches high-salience and low-salience brands by industry and query type. You must control for platform differences and source availability to prevent familiarity from inflating false attribution rates. Without these controls, rankings reflect model bias rather than actual entity visibility.
- [AI Visibility Maturity Levels: What They Mean for Your Company](https://airanking.hu/articles/ai-visibility-maturity-levels/): AI visibility maturity levels classify a company's readiness for interaction with autonomous AI agents based on a 25-dimension audit score ranging from 0 to 100. These levels indicate whether an AI system can technically access your site, correctly identify your entity, understand your offer, and cite you as a reliable source in decision-making processes. Moving between levels requires specific structural changes to your digital presence rather than general marketing efforts.
- [How AI Visibility Differs Across Sectors in the Hungarian Market](https://airanking.hu/articles/ai-visibility-by-sector/): Marketing agencies and SaaS platforms currently demonstrate the highest AI visibility scores, while general business categories and smaller e-commerce segments lag significantly. The data indicates that content structure and entity clarity drive these differences more than technical website availability.
- [Odds Ratios and Confidence Intervals in AI Visibility Research](https://airanking.hu/articles/odds-ratios-in-ai-visibility-research/): Read an odds ratio as the relative likelihood of an event occurring in one group versus another, where a value of 1.0 indicates no difference. Check the confidence interval to see if it includes 1.0; if it does, the result is not statistically significant at the chosen level. In AI visibility, this distinction separates genuine entity-level bias from random variation in model behavior.
- [Should an AI Visibility Index Weight Platforms Equally?](https://airanking.hu/articles/weighting-platforms-in-an-index/): No, an AI visibility index should not weight platforms equally because models exhibit fundamentally different behaviors regarding source fabrication and engagement. Data from a 2026 study of Hungarian B2B entities shows one platform generated 10.9 times more citations with a 2.9 times higher hallucination rate than another. Equal weighting would distort the index by allowing high-volume, high-error platforms to dominate the score.
- [Checklist: Publishing an AI Visibility Ranking Responsibly](https://airanking.hu/articles/publishing-ai-rankings-responsibly/): Publish an AI visibility ranking responsibly by explicitly disclosing the sampling method, scoring criteria, and entity calibration process before presenting any results. You must flag small sample sizes that lack statistical power and avoid absolute claims about individual company performance or guaranteed outcomes. Clear communication guardrails prevent readers from misinterpreting exploratory data as definitive market proof.
- [Validity Threats in AI Visibility Rankings](https://airanking.hu/articles/validity-threats-in-ai-rankings/): AI visibility rankings become invalid when they fail to distinguish between genuine citation escalation and artifacts like growing context windows or base-rate differences in URL generation. Misleading results also arise from unstable URLs, unverified fabrication metrics, and statistical models that ignore structural zeros in the data. Without pre-screening entities into a 'gray zone' and applying multi-layer verification, an index measures noise rather than actual brand presence.
- [Hungarian AI Visibility Index: What It Measures and How It Works](https://airanking.hu/articles/hungarian-ai-visibility-index-faq/): The Hungarian AI Visibility Index is a forthcoming benchmark that measures how clearly and accurately AI systems can identify, interpret, and cite Hungarian companies based on their public web presence. It uses a 25-dimension audit to score entities on factors like structured data, entity clarity, and factual accuracy rather than traditional traffic metrics. Results will be published periodically to track market-wide readiness for AI-driven decision-making.