Article, 6-minute read
How does ChatGPT choose which businesses to recommend?
ChatGPT and other AI assistants recommend businesses they find clear, specific information about in sources they trust: their website, Google Maps, directories and articles. Research shows that pages with numbers, sources and quotations appear in answers up to 41% more often.
Where does ChatGPT get its information?
When you ask ChatGPT, Perplexity or Gemini for “the best dentist in Amman”, the assistant searches the web, reads a small number of pages, then writes an answer naming a few businesses and their sources. So the competition is no longer only about rank, but about being one of the few pages it reads and quotes. Google itself says there are “no additional requirements” to appear in its AI Overviews, and that SEO basics are what count1.
Which pages do answers cite?
The GEO-16 researchers analysed more than 1,700 citations from three engines and found that pages hitting 12 of 16 pillars were cited 78% of the time2. The strongest signals were a clear update date, structured HTML, structured data and cited sources. The GEO study found that adding quotations raised visibility by about 41%, and adding statistics by about 30%3.
“Support claims with specific, verifiable data, statistics or examples.”
- Metadata & freshness
- Semantic HTML
- Structured data
- Evidence & citations
- Authority & trust
- Internal links
- External links
- Readability
- Claim accuracy
- Direct answers
- Transparency
- Content depth
- Visuals & media
- Engagement
- Canonical & crawlability
- Topic focus
A pillar counts as hit at 2 or more
What can you do today? Seven steps
- Answer first: the first two sentences of every page answer the main question with a specific fact.
- Give numbers: price, number of sessions, duration and years of experience, with a source for each.
- Add a quote: one sentence from the doctor, engineer or expert, in their name.
- Cite your sources: at least two links to official bodies or studies.
- Show the date: “last updated” on the page and in the code.
- Complete your Google profile: answers often cite maps and reviews.
- Test it yourself: ask ChatGPT and Gemini your customers’ questions more than once and note who is named.
What are the common mistakes?
- Keyword stuffing: lowered visibility by about 8% in the GEO study3.
- Promotional language: phrases like “the best” and “number 1” with no evidence; AI assistants prefer neutral language4.
- Pages with no date or author: Google advises making it clear who wrote the content5.
- Hidden text and instructions aimed at AI: they can work briefly in experiments6, but they break Google’s policies7.
What does the research not say?
There is no way to guarantee ChatGPT names you every time. Most studies are on English content, some on experimental engines, and answers change from day to day. So repeated measurement matters more than any trick. We explain our approach in our method and apply it in the GEO service.
FAQs
What is GEO, or visibility in AI answers?
GEO stands for Generative Engine Optimization: improving content so that ChatGPT, Gemini, Perplexity and Google’s AI Overviews quote it when they answer a question. The name comes from a study published at KDD 2024, which found that specific edits to a text raise its visibility in answers by up to 40%3.
Can you guarantee ChatGPT will recommend me?
No, and nobody can honestly promise that, because answers change from one run to the next and from one assistant to another. What we guarantee is the work: rules proven in research, and a monthly report showing how often you are named, who is named instead and why.
Sources
- Google Search Central, AI features and your website, updated December 2025. developers.google.com
- Kumar & Palkhouski, AI Answer Engine Citation Behavior: the GEO-16 framework, 2025. arxiv.org
- Aggarwal et al., GEO: Generative Engine Optimization (KDD 2024), 2024. arxiv.org
- Wu, Zhong, Kim & Xiong, What Generative Search Engines Like and How to Optimize Web Content Cooperatively (AutoGEO), 2025. arxiv.org
- Google Search Central, Creating helpful, reliable, people-first content, updated October 2026. developers.google.com
- Kumar & Lakkaraju, Manipulating Large Language Models to Increase Product Visibility, 2024. arxiv.org
- Google Search Central, Spam policies for Google web search, 2026. developers.google.com