Vercatus
Sign inGet started
Get started
Illustrated avatar: three stacked isometric plates with a blue block rising

Author

Norbert Szabo
Founder of Vercatus

Norbert Szabo is the founder of Vercatus, the AI visibility growth platform built in Budapest, Hungary, by Oktelo Solutions. He writes about how AI answer engines read, cite and name brands, and about telling whether a change to a site actually moved the answers.

Norbert Szabo on LinkedInAbout VercatusHow we measure
Writes about
  • AI visibility
  • Answer engines
  • AI crawler access
  • Structured data and entities
  • Content for AI answers
How he writes
What he publishes here follows the rules Vercatus holds itself to: each article states its method and its limits, leans on graded evidence from the Vercatus knowledge base, and makes no claim about a result that was not measured. Vercatus is measured the same way: its own site is analysed first, and the result is published with its date and its limits.

26 articles
on AI answers.

Guides, studies and comparisons. Each states its method, its sources and its limits, and says where a fact comes from.

  1. Flat illustration of cards side by side joined by a dashed line, standing for three tools set side by side by what each says it does.

    Article9 min read

    Ahrefs Brand Radar vs Dedicated AI Visibility Platforms

    Whether Ahrefs Brand Radar is enough to track AI mentions, or a dedicated AI visibility platform is worth adding: what each says it does, which questions to check for overlap and gaps, and how to decide.

  2. Flat illustration of a central circle with smaller circles around it, joined by lines, standing for a wrong fact at the centre that has to be corrected at its sources.

    Article6 min read

    AI Answers Show Wrong Information About Our Brand: How to Fix It

    Why AI assistants state wrong or outdated facts about a brand, how to see what each engine says, and how to correct pricing, discontinued products and funding figures at the source.

  3. Flat illustration of a path running between two pillars on a white card, standing for a crawler refused at the edge before the rule behind the refusal is found.

    Article10 min read

    AI Crawler 403 Errors on Cloudflare: Diagnose the Rule Before Changing It

    Find whether a crawler refusal comes from robots policy, an edge rule, a challenge, rate limiting or the origin. Verify identity before changing access.

  4. Flat illustration of a bar chart on a white card, one bar in colour, standing for clicks falling across a cohort of pages, to be read with care.

    Article6 min read

    AI Overviews and Organic CTR: How to Measure the Impact on Your Site

    Use Search Console’s available reports, stable query cohorts and business outcomes to investigate traffic changes without treating AI impressions as a CTR denominator.

  5. Flat illustration of cards side by side joined by a dashed line, standing for three purposes, search, training and user fetch, each with its own lane.

    Article6 min read

    AI Search Crawlers vs Training Bots: What to Allow in robots.txt

    Separate search discovery, model training and user-requested fetches. Choose crawler-specific rules and verify what your website actually serves.

  6. Flat illustration of a bar chart on a white card, one bar in colour, standing for a measured return built stage by stage, with the claim kept to what was observed.

    Article6 min read

    AI Search Visibility ROI: Connect Citations to Qualified Leads

    Model the costs and qualified outcomes of AI visibility work. Separate observed citations, referral traffic and incremental business value, with explicit assumptions.

  7. Flat illustration of a grid of rounded squares, one blue and one orange, standing for a stable panel of questions reviewed month after month for a client.

    Article6 min read

    AI Visibility Audits for Agencies: A Monthly Client Workflow

    Run a client audit from authorization and baseline to evidence-backed tasks and a monthly review. Keep access, citations and business results separate.

  8. Flat illustration of a table of rows on a white card, one row highlighted, standing for tools compared row by row on dated vendor pages.

    Article6 min read

    AI Visibility Tracking Tools Compared: Features, Pricing and Data Sources

    Compare Vercatus, Profound, Peec AI, OtterlyAI and Semrush by measurement scope, prompt limits, evidence and workflow. Includes a practical evaluation checklist.

  9. Flat illustration of a central circle with smaller circles around it, joined by lines, standing for one organisation at the centre with its profiles and pages agreeing around it.

    Article6 min read

    Brand Entity Consistency: Align Your Website, Authors and Public Profiles

    Align organization, product and author facts. Use verified identifiers and sameAs relationships, distinguish parent companies, and correct conflicting claims.

  10. Flat illustration of cards side by side joined by a dashed line, standing for two pages competing for the same question and the choice of which one owns it.

    Article6 min read

    Content Cannibalisation in AI Search: Find Overlap and Choose One Page

    Map questions to primary URLs, distinguish useful supporting pages from duplicates, and choose consolidation, differentiation or canonical repairs using evidence.

  11. Flat illustration of three stacked page cards, the top one in colour with lines of text, standing for a page built in layers to answer one reader's question completely.

    Article7 min read

    How to Make Website Content Useful as a Source for AI Answers

    Turn customer questions into complete, evidence-backed pages. Use clear answers, useful detail, consistent product facts and a repeatable review process.

  12. Flat illustration of a grid of rounded squares, one blue and one orange, standing for a set of overlapping terms, each defined by what can be observed.

    Article6 min read

    GEO, AEO and LLMO: An AI Search Glossary with Measurement Definitions

    Understand AI search terminology and distinguish access, coverage, retrieval, mentions, citations and business outcomes. Includes Vercatus's own terms: Answer Share and verdicts.

  13. Flat illustration of one node with lines fanning out to several nodes, standing for one question fanning out into several related searches.

    Article6 min read

    Google AI Mode Query Fan-Out: What It Means for Content Planning

    Map a complex buyer task to useful supporting answers without inventing hidden engine queries or publishing a page for every wording variation.

  14. Flat illustration of a central circle with smaller circles around it, joined by lines, standing for a cluster of pages linked around one primary page.

    Article6 min read

    Internal Linking for Content Clusters: Connect Answers to the Right Page

    Give each question a primary page and connect guides, evidence and product documentation with descriptive HTML links. Find orphan pages and maintain links after changes.

  15. Flat illustration of cards side by side joined by a dashed line, standing for two language versions that must point at each other and answer for their own market.

    Article6 min read

    Multilingual AI Search Visibility: Diagnose Language and Locale Mismatches

    Why an AI assistant answers a question in one language with another market's facts, and how to align localized content, canonical URLs and hreflang, then test each language and market separately.

  16. Flat illustration of steps rising left to right to a green flag, standing for a page's real revisions over time, step by step.

    Article6 min read

    Does Updating a Page’s Date Improve AI Citations?

    Distinguish cosmetic date changes from useful content updates. Keep visible dates, article schema and sitemap dates honest, then measure comparable observations.

  17. Flat illustration of a table of rows on a white card, one row highlighted, standing for product facts and variants laid out as data a shopper can compare.

    Article6 min read

    Product Pages for AI Shopping: Facts, Variants, Prices and Feeds

    Publish product facts that answer buying questions and keep variant pages, offers and channel feeds consistent. Covers Google shopping and Amazon separately.

  18. Flat illustration of one tall bar with a flag beside shorter bars, standing for a study standing on its method and data, with its limits next to it.

    Article6 min read

    How to Publish Original Research That Others Can Check and Cite

    Design a useful study, publish the sampling and method, release inspectable data, and make claims easy to verify without promising links or AI citations.

  19. Flat illustration of a path running between two pillars on a white card, standing for a publisher choosing which crawlers pass, by purpose and by content area.

    Article6 min read

    Should Publishers Allow AI Search Crawlers? A Practical Decision Framework

    Choose publisher access policies by crawler purpose, content rights and measurable outcomes. Verify implementation without treating visits as citations or revenue.

  20. Flat illustration of a table of rows on a white card, one row highlighted, standing for plans, prices and conditions kept together in one readable table.

    Article6 min read

    How to Structure a SaaS Pricing Page for Search and AI Answers

    Make plan names, billing conditions, allowances and exceptions readable together. Test the buying questions your SaaS pricing page should answer.

  21. Flat illustration of a viewfinder frame on a white card with a square inside it, standing for a fixed panel of questions observed through the same lens each time.

    Article8 min read

    How to Track AI Citations, Brand Mentions and Share of Answer

    Build a repeatable AI visibility panel. Separate citations from mentions, use explicit denominators and compare the same questions across answer engines.

  22. Flat illustration of a viewfinder frame on a white card with a square inside it, standing for looking for why a site does not appear, one check at a time.

    Article8 min read

    Why Your Website Is Not Showing Up in ChatGPT Search

    Diagnose missing ChatGPT citations: check search access, discovery, readable content and question coverage, then measure the same questions again.

  23. Flat illustration of one node with lines fanning out to several nodes, standing for several alternatives, each fitting a different job.

    Comparison7 min read

    Vercatus Alternatives: When to Choose a Tracker, SEO Suite or Agency

    Choose an alternative to Vercatus by the work you need: a prompt tracker, broad SEO research, implementation capacity or a custom evidence pipeline.

  24. Flat illustration of cards side by side joined by a dashed line, standing for two platforms compared by the job each is built to do.

    Comparison6 min read

    Vercatus vs Profound: Counting AI Answers or Changing Them

    Compare Vercatus and Profound by the job your team needs done: measuring AI answers, explaining why a competitor is picked, making the change, and judging honestly whether it worked.

  25. Flat illustration of three stacked page cards, the top one in colour with lines of text, standing for the page as served in HTML and the page after scripts run, two layers that are not the same.

    Article7 min read

    Do AI Crawlers Run JavaScript? How to Test Your Website's Content

    Compare initial HTML with the rendered page, separate rendering gaps from crawler blocks, and make your essential answers readable before JavaScript runs.

  26. Flat illustration of a bar chart on a white card, one bar in colour, standing for a lab measurement and the field measurement beside it, which do not have to agree.

    Article7 min read

    Lab vs Field Core Web Vitals: Diagnose a Fast Test and Slow Users

    Understand why lab and field performance disagree, separate TTFB from Core Web Vitals, and investigate a slow-user experience without assuming its cause.

Evidence, not mentions.

Create your account, add your site, and see what Vercatus already reads on your domain before you choose a plan.

Get started

Cancel any time · Data stored in the EU