Skip to content
7.18Advanced9 min

The Five Pillars of GEO Skepticism: A Critical Analysis

Lucas BlochbergerUpdated 11 June 2026
Summarize with AIChatGPTClaudePerplexity

Opens the chat with a prepared prompt.

Definition

GEO skeptics argue on five levels: Measurement is fundamentally broken (SparkToro), GEO is repackaged SEO (Sullivan, Beus), popular tactics are being penalized (Lily Ray), the zero-click paradox undermines the economic logic, and the market remains far smaller than claimed.

Key Takeaways

  • SparkToro: Less than 1% chance of identical brand list, rankings meaningless
  • Danny Sullivan: Good SEO is good GEO
  • Lily Ray documented five dangerous GEO tactics that Google penalizes
  • Glenn Gabe coined Mt. AI for pages with scaled AI content
  • Google processes 373x more searches than ChatGPT
  • Gartner's 25% forecast did not materialize; Google grew 21.64%
  • Nuanced middle: GEO is layer on SEO, not replacement

The strongest criticisms of GEO deserve serious investigation rather than dismissal. A balanced analysis of five pillars.

Measurement is fundamentally broken

SparkToro's January 2026 study (2,961 responses, 600 volunteers) proved there's less than a 1-in-100 chance that ChatGPT delivers the same brand list twice, and approximately 1-in-1,000 for the same order. Rand Fishkin estimates that 100+ million USD annually is spent on meaningless AI-ranking metrics.

Repackaged SEO

Danny Sullivan, Johannes Beus, and John Mueller all argue that GEO's core recommendations are identical to SEO best practices. Lily Ray documented coordinated GEO disinformation campaigns from emerging tool providers.

Tactics fail

Lily Ray documented in February 2026 that Google actively penalizes self-promotional listicles. Glenn Gabe coined the term Mt. AI for pages that published scaled AI content: traffic rose, then crashed. Five dangerous tactics: Scaled AI production, self-promotional listicles, white text as prompt injections, fake astroturfing, schema without quality.

The market is smaller than claimed

Google processes 373 times more searches than ChatGPT. Gartner's forecast of a 25-percent decline has not materialized: Google grew 21.64 percent instead. 95 percent of ChatGPT users continue to use Google.

The nuanced middle

GEO represents real changes. Brand mentions correlate 3x stronger than backlinks. AI Overviews increasingly cite beyond the Top-10. Content freshness is dramatically more important. But these are extensions of SEO, not a new paradigm. The brands that win in AI have genuine authority and broad third-party presence.

Data & Statistics

GEO ist größtenteils umetikettiertes SEO: 80 Prozent von GEO sind gutes, fundamentales SEO (Jeremy Moser, uSERP-CEO); GEO ist eine Erweiterung von SEO, keine eigene Disziplin

Digiday (2026)

Die Princeton-GEO-Studie ist ein Preprint ohne Peer-Review mit simulierter statt produktiver Engine (GPT-3.5 plus Template-Prompts); Kritik wörtlich: ernste Verzerrungen, Fehler und Fragezeichen in der Methodik

SandboxSEO - GEO Targeted: Critiquing the Generative Engine Optimization Research (2024)

ChatGPT-Traffic konvertierte mit 1,81 Prozent vs. 1,39 Prozent bei non-branded organischer Suche (plus 31 Prozent); 9,46 Mio. organische vs. 135.000 ChatGPT-Sessions; 474.000 USD ChatGPT-Umsatz vs. 32,1 Mio. USD organisch = 1,48 Prozent des organischen Umsatzes; 94 E-Commerce-Marken

Visibility Labs via Search Engine Land (2026)

61,7 Prozent Ghost Citations (zitiert ohne Markennennung), 38,3 Prozent der Erscheinungen mit Markennennung; 62 Prozent der KI-Zitate führen zu keiner Markennennung; 3.981 Domain-Erscheinungen, 115 Prompts, 14 Länder, 4 KI-Engines

Semrush und Kevin Indig (Growth Memo), Ghost Citations Study (2026)

Klick auf klassisches Suchergebnis: 8 Prozent mit KI-Zusammenfassung vs. 15 Prozent ohne; Klick auf Link in der KI-Zusammenfassung: 1 Prozent; n=900 US-Erwachsene, 68.879 Suchanfragen, März 2025

Pew Research Center (2025)

Minus 58 Prozent organische CTR für Position 1 (Datensatz Dezember 2025), zuvor minus 34,5 Prozent (April 2025); 300.000 Keywords (150k mit AIO / 150k ohne AIO)

Ahrefs Blog (2026)

In Österreich schätzen 73 Prozent ihr KI-Wissen als gering oder nicht vorhanden ein; 31 Prozent haben generative KI-Tools schon einmal verwendet; nur 4 Prozent nutzen sie täglich oder fast täglich

Statistik Austria, IKT-Erhebung in Haushalten 2024 (Pressemitteilung 13.605-098/25) (2025)

“I believe the methodology outlined in the paper contains serious biases, flaws and question marks.”

— SandboxSEO, Autor der Analyse GEO Targeted: Critiquing the Generative Engine Optimization Research

“80 percent of GEO is good, fundamental SEO.”

— Jeremy Moser, CEO, uSERP (zitiert in Digiday)

FAQ

Is GEO just a new name for SEO?
Mostly yes, but not entirely. According to Digiday, around 80 percent of GEO is fundamental SEO, and LLMs predominantly cite high-ranking, authoritative content. However, the remaining 20 percent differs: AI systems prefer front-loaded, extractable answer blocks and evaluate brand mentions differently than backlinks. GEO is therefore a layer on top of SEO, not a replacement.
How reliable is the Princeton GEO study with the plus 40 percent?
Limited. According to SandboxSEO, it is a preprint without peer review. The study did not test a production search engine but fed each query with the top five pages via a template prompt into GPT-3.5, essentially a simulated engine. The percentage gains are an indication of possible mechanisms, but not a reliable business foundation.
Is GEO worthwhile for an Austrian B2B company today?
That depends on maturity. According to Visibility Labs, AI traffic converts at 1.81 percent, 31 percent higher than non-branded organic search, but accounts for only 1.48 percent of organic revenue. With such a small channel, a dedicated large GEO budget is rarely justified. It makes more sense to treat GEO as a byproduct of a strong SEO and content foundation.
Why is GEO success so difficult to measure?
Because reliable metrics are lacking, LLM ranking is a black box, and results cannot be guaranteed. Added to this is the ghost citation problem: according to Semrush, 61.7 percent of AI citations are ghost citations, meaning citations without brand mention. Such effects generate neither clicks nor visible brand perception and do not appear in any traditional analytics report.
What do GEO skeptics overlook?
The measurable click collapse in traditional search. Pew Research measures 8 instead of 15 percent clicks once an AI summary appears, and Ahrefs documents minus 58 percent CTR for the position-1 page with AI Overviews. Even if AI traffic generates little revenue today, the traditional organic click as a source is eroding. This justifies GEO as preparation.
Do you need an llms.txt file for GEO?
There is no evidence for it. No relevant AI search engine confirms that it considers llms.txt. The file costs little and does no harm, but does not replace substantive content. Marketing it as mandatory is one of the patterns that GEO critics describe as monetizing uncertainty.
When will GEO become mandatory for DACH companies?
As soon as their target audience actively uses AI systems for research and pre-selection and the traditional click noticeably erodes. In the DACH region, this point shifts backward because according to Statistik Austria, 73 percent rate their AI knowledge as low and only 4 percent use generative AI daily. The smart approach is to build the foundation so that GEO runs alongside, and scale precisely when the target audience follows.

How does your website perform?

Get a free, AI-powered SEO report of your website by email: technical SEO, on-page, keywords & competitors. No obligation.

Get a free SEO audit →
Previous← Zero-Click and AI Cannibalization: The Traffic ParadoxNextAgentic Commerce: AI Agents as Purchase Mediators →