“Name the best HR software suppliers in the Netherlands.”
An HR director types this prompt into Perplexity. Her colleague is simultaneously Googling the same question. Two searches, two completely different answer sets and in 92.8% of Dutch brands: absent in both worlds.
Only 7.2% of all domains investigated appear in both Google AI Overviews and in answers from ChatGPT and Perplexity (Search Engine Land, October 2025). That 7.2% is no coincidence. It's a strategy.
| DEFINITION The AI visibility gap is the measurable difference between a brand's presence in traditional search engines (Google, Bing) and in AI-generated answers from language models (ChatGPT, Perplexity, Gemini, Google AI Overviews). Research by Search Engine Land (October 2025) shows that 92.8% of all brands fall on the wrong side of this gap. |
There are two AI worlds and they use different ranking rules
Google AI Overviews and language model platforms such as ChatGPT, Perplexity, and Gemini function on fundamentally different citation logic. Google AI Overviews builds upon the existing Google index and rewards technical SEO authority, PageRank signals, and structured data. Language model platforms operate based on training data, retrieved context, and earned media authority.
The consequence is clear: traditional SEO authority does not automatically translate into citations in language model answers. A brand can rank in the top 3 in Google and simultaneously be completely invisible in ChatGPT.
Google AI Overviews versus taalmodelplatforms: hoe signalen wegen
| Signal | Google AI Overviews | ChatGPT / Perplexity / Gemini |
|---|---|---|
| Technical SEO | ★★★★★ | ★★ |
| Backlink authority | ★★★★ | ★★★ |
| External references | ★★★ | ★★★★★ |
| Quoted content structure | ★★★ | ★★★★★ |
| Structured data / schema | ★★★★ | ★★★ |
| E-E-A-T signals | ★★★★ | ★★★★★ |
Two platforms, two scorecards, one strategy covering both.
Four figures quantifying the AI visibility gap
The growth of AI search is no longer a prediction. It is a measurable reality:
- Google AI Overviews verschijnt bij 52% van alle getrackte zoekopdrachten (SEL februari 2025), een stijging van 6,49% tegenover januari 2025.
- ChatGPT shows a growth of 300% in unique domains the AI traffic for all of 2024.
- 89% van B2B-kopers heeft generatieve AI geadopteerd in their purchasing process (Forrester Research, 2025).
- 58% van consumenten gebruikt AI-tools in plaats van zoekmachines by product and service research (Capgemini Research Institute, 2025).
Conclusion: the AI visibility gap is widening. Organisations without a GEO strategy are losing ground monthly in both AI worlds simultaneously.
Three structural reasons why SEO champions fail in language models
1. Content mist antwoordstructuur
SEO content has historically been built around keywords and length. Language models prefer direct, quoted answers in the first paragraph. A 3,000-word page without a clear answer opening will not be quoted regardless of its Google ranking. Language models scan for topic sentences that can stand as independent assertions.
2. Entity recognition is missing
Language models only cite brands for which they have entity certainty. Entity certainty means: the language model knows with high probability who your organisation is, in which sector you are active, what your specialisation is and how you are positioned against competitors. Vague company descriptions, missing Wikipedia presence and inconsistent sameAs markup produce empty entities. Empty entities are not cited.
3. Earned media is dead
Authority in language model terms is determined by mentions in sources that language models trust: trade publications, Trustpilot, G2, Capterra, LinkedIn articles, press releases, and interviews. A strong own website without an external echo chamber does not produce citations in ChatGPT, Perplexity, or Gemini.
Three pillars to end up in the 7.2%
Visibility in both AI worlds requires a dual strategy with three equal pillars. One pillar alone is not sufficient.
Pillar 1: Rewrite content as an answer capsule
Each core page must answer a specific question, directly in the first paragraph, with verifiable data. Answer capsules are concise (40–60 words), stated declaratively, and contain at least one concrete figure or citation.
Content types ranked by citation potential
| Content type | Citation potential | Priority |
|---|---|---|
| Original research with own data | ★★★★★ | Hoog: unique and claimable |
| Definition pages and glossaries | ★★★★★ | High: direct extraction |
| FAQ sections with direct answers | ★★★★ | High: question-answer structure |
| Case studies with measurable results | ★★★★ | High: verifiable proof |
| Generic ‘about us’ pages | ★★ | Layer: no unique value |
| Long blog posts without a reply structure | ★★ | Layer: not extractable |
Pillar 2: Building Entity Authority
Entity authority is the extent to which language models are confident about who your organisation is. Four concrete actions build this authority:
- Ensure a consistent, accurate company description across LinkedIn, Crunchbase, industry directories, and your own website.
- Implement Organization and LocalBusiness schema markup with sameAs links to all external profiles.
- Publish monthly in external trade publications, not exclusively on your own blog.
- Encourage reviews on platforms that trust language models: G2, Capterra and Trustpilot.
Pillar 3: Measuring what language models say about you
Optimisation without measurement is guesswork. A workable GEO-monitoring system consists of a prompt set of 20 to 50 questions relevant to your sector and target audience. This set is tested monthly in ChatGPT, Perplexity, Gemini, and Google AI Overviews. The outcome records two variables: do you appear, and if you appear, how are you described.
Three questions that make the AI visibility gap measurable for your organisation
- Wat ChatGPT antwoordt als iemand vraagt: ‘Wat zijn de beste [jouw dienst] aanbieders in Nederland?’ Sta je erin en als eerste of als twaalfde?
- How does Perplexity describe your company when a prospect mentions your brand name? Is the description accurate, outdated, or missing entirely?
- Which external sources mention your organisation, and are those sources recognised as authoritative by language models?
One unanswered question from these three is the starting point of your GEO strategy.
Have the AI visibility gap for your organisation measured
AI Rebels carries out comprehensive GEO audits on ChatGPT, Perplexity, Gemini and Google AI Overviews. An audit identifies precisely why you are not being cited, which competitors are visible, and which adjustments deliver the biggest leap in citability.
Start your GEO Audit via ai-rebels.uhu.nl
Sources & References
[1] Libert, K., ‘Only 7.2% of domains appear in both Google AI Overviews and LLM results’. Search Engine Land, October 2025.
[2] GitHub/amplifying-ai, ‘Awesome Generative Engine Optimisation’. February 2026.
[3] Forrester Research, ’B2B AI Adoption Study 2025‘. Quoted in Profound ’GEO Guide 2025’.
[4] Capgemini Research Institute, ‘AI in Consumer Interactions’. 2025.