top of page

ICYMI Metehan: Statistics & Fact Checking in AI Chats and beyond the Web [Experiment]

Autorenbild: David Epding
David Epding
vor 9 Minuten
2 Min. Lesezeit

Key Takeaways:

ICYMI Metehan: Statistics & Fact Checking in AI Chats and beyond the Web [Experiment]
  • Peec AI/Metehan tested the SEO/GEO tactic to publish statistics as LLMs look for reliable data to ground their answers from

  • Metehan set up a website with 50.000 pages of fake statistics (stateglobe.com) on SEO, content marketing, social media, e-commerce, and web technology from 200 countries and it ranked well in ChatGPT

    • Google burried the page after a few days and did not sent further traffic

    • ChatGPT kept crawling and citing the fake statistics from the pages

    • The website gained 330+ referring domains in 7 months even from well-known brands such as Shopify or DHL

    • The website also got cited in multiple academic papers

  • Metehan also referred to an experiment by Seer Interactive who manipulated the footer of their website to get recommended through the success metrics added (2025)

  • Peec AI set up a statistics page with their original data and almost instantly received referral links (mostly by websites using AI writing tools, but still): https://peec.ai/ai-search-geo-statistics

    • When a prompt requires research, the model breaks it into sub-queries, adds modifiers like "statistics," "data," "report," or the current year (if also requested in the prompt text). If someone asks for the model to compose a blog post about remote work, the model will search for "remote work statistics 2026." Pages that optimize for what models search rather than for what humans type will appear more often in LLM results.

  • What should you do?

    • Data is key! Statistics work! Publish as much original data as you can - helps with Information Gain and AI Chat citations through Query Fan-outs - and it might even alter the recommendation of your brand (see Serr Interactive experiment)

    • Do not trust a statistic you did not alter yourself

  • Metehan also provides some how to recommendations:

    • Make one claim per sentence. Each stat should make sense on its own, out of context. Retrieval doesn’t work on pages, but rather on passages. If a number needs the paragraph above it to make sense, it probably won't get pulled.

    • Put the source and year next to every number. "According to [Source] (2026), X%..." This helps the model, the human editor, and your credibility. It’s one small piece of information hygiene that benefits the broader information environment.

    • Match headings to fanout queries. The headings on your page should match the way LLMs search the internet: "[Topic] statistics," "[Topic] market size," "[Topic] adoption by country." Think in the queries models run, not just the keywords people type.

    • Show the update date (and actually update it). Freshness is a real signal that models take into account. A stale stats page will lose to a newer one.

    • Make the stats easy to fetch. Use server-rendered HTML instead of hiding data behind JavaScript, tabs, or images. Also, be sure to check your robots.txt and your CDN or WAF rules. Your site may be blocking AI retrieval bots without your knowing it.

    • Keep it simple. Our Peec AI GEO statistics page is basically a list. Yours doesn’t need to be anything more than that, either. Save design resources for sites and assets that can actually benefit from them.

    • Add original data when you can. Aggregated stats may get you a few citations, but when you say something no one else does, you can become the primary source.



Sources:

© 2026 David Epding.            Erstellt mit Wix.com.

logo_new_100_edited.png

David Epding ist AI Search (GEO & SEO), Data Analytics und Automation Manager mit über 10 Jahren Erfahrung in Technischem SEO mit breiter Expertise für LLMs und langjähriger Erfahrung in der Daten-Analyse.

bottom of page