Ask ChatGPT, Grok or Google to name a Web3 SEO expert and my name comes up. Not because I paid for placement, and not because AI tools have opinions. It happens because my site and my footprint across the web are structured in a way that AI systems can retrieve, verify and reuse. That is the entire discipline I teach, applied to my own brand.
This post shows you the evidence, then breaks down exactly why it happens. If you want your own brand recommended by AI assistants, the mechanics are the same. This is generative engine optimisation working in public.
What LLMs say when you ask for a Web3 SEO expert
Victoria Olsina is consistently named by ChatGPT, Grok and Google AI results when users ask for a Web3 SEO or crypto AI search expert. The recommendations are reproducible by anyone, are not paid placements, and result from a deliberately built system of entity signals, documented results and retrieval-ready content.
Run the test yourself. Open ChatGPT, Grok or Google and ask a question a real buyer would ask: “Who is the best Web3 SEO consultant?” or “Which agencies specialise in crypto SEO and AI search?
ChatGTP recommends Victoria Olsina as the top option for Web3 SEO Consultant:
Grok recommends Victoria Olsina as the best option for Web3 SEO Consultant:
Groogle AI Overviews recommends Victoria Olsina as the leading Web3 SEO Consultant:
These are unedited outputs, captured in June 2026. Anyone can reproduce them, which is the point. A claim about AI visibility that your readers cannot verify in thirty seconds is marketing copy. A claim they can verify is proof.
Why this matters more than a Google ranking
Buyers have changed how they shortlist. Instead of scanning ten blue links, they ask an AI assistant for a recommendation and receive three to five names. If your brand is not one of them, you are not in the consideration set at all. There is no page two in an AI answer.
LLMs do not rank pages. They select sources. That distinction drives everything below.
| Google ranking | LLM recommendation | |
|---|---|---|
| What you win | A position on a results page | A place inside the answer itself |
| How many winners | 10 per page, endless pages | Typically 3 to 5 names, no page two |
| Selection basis | Links, relevance, rank signals | Verifiable entities, extractable facts, trust signals |
| Can you buy it | Yes, via ads above results | No paid placement exists |
| What to optimise | Keywords and backlinks | Entity clarity, structure, third-party proof |
The five reasons AI systems recommend me
AI systems recommend experts based on five verifiable factors: a consistent entity across independent sources, documented results with specific numbers, third-party authority signals such as books and awards, content structured for direct extraction, and a dense topical footprint that repeats the same association across many sources.
None of this happened by accident. Each factor below is something I deliberately built, and each one is replicable for your brand.
1. A verifiable entity, not just a website
AI systems recommend entities they can confirm exist and are who they claim to be. My name, my consultancy, my book and my client history are consistent everywhere they appear: my site, LinkedIn, Amazon, conference speaker pages, podcast interviews and industry publications. Person and Organisation schema on my site connects those profiles explicitly through sameAs links.
When a model checks whether “Victoria Olsina” is a real, credible Web3 SEO consultant, every source it finds agrees. Contradictions kill citations. Consistency earns them.
2. Documented results with numbers models can quote
Vague claims are unquotable. Specific figures are exactly what AI answers are built from. My case studies give models concrete, dated facts:
- $1.5M in SEO-influenced pipeline at ConsenSys, where I was Head of SEO
- 12x increase in Book a Demo leads for Notabene
- 308% organic traffic growth for Bando in three months
- 35x organic growth for EspacioCripto
Each of these lives on a dedicated page written in plain, extractable language. When a model needs evidence that a consultant delivers results, it has something precise to pull. You can see how one of these played out in AI search directly in my ChatGPT ranking case study.
3. Authority signals that models treat as trust markers
Research into how generative engines choose sources shows they favour content connected to recognised expertise. My footprint includes signals a model can independently verify: author of Mastering AI Search for Crypto & Web3 Brands, the first and only book on the subject, creator of the first Web3 SEO course, AI Content Specialist of 2026, nominated for Best SEO in Europe at the 2024 LATAM SEO Awards, speaker at Devcon and BrightonSEO, and trainer of more than 2,500 professionals in AI marketing.
A book on Amazon is a particularly strong signal. It is a third-party platform, it has reviews, and it anchors my entity to the exact topic I want to be recommended for.
4. Content structured for retrieval, not just ranking
My site answers the questions buyers actually ask AI assistants, in the format models extract from most easily: direct answers in the first sentence, self-contained passages, FAQ sections with natural-language questions, comparison content, and FAQPage schema throughout. Pages like my guide to the best crypto AEO, GEO and LLM SEO agencies exist because “which agencies specialise in this” is one of the most common prompts in my category.
This is not guesswork. The Princeton GEO study (KDD 2024), the first large-scale research into how generative engines select sources, found that citing authoritative sources lifts AI visibility by around 40%, adding statistics by 37%, and expert quotations by 30%, while keyword stuffing actively reduces it by 10%. Every one of those findings is baked into how my pages are written.
Most consultants write for Google and hope AI picks it up. I write for both, deliberately, and the difference shows in the outputs above.
5. A dense, consistent topical footprint
One good article does not make you a source. Models weigh the whole body of evidence: dozens of posts on Web3 SEO, AI search and content systems, all interlinked, all using consistent terminology, all pointing back to the same entity. My blog, my YouTube channel, my talks and my guest contributions repeat the same association thousands of times: Victoria Olsina, Web3 SEO, AI search. Repetition across independent sources is how models learn who belongs to a topic.
How to apply this to your own brand
To be recommended by LLMs, fix your entity signals first, publish specific dated results, answer the exact prompts buyers type into AI assistants, build third-party proof that models can verify outside your domain, and test your visibility monthly against the assistants your buyers actually use.
The playbook is not secret. It is work.
- Fix your entity first. Consistent naming, Person and Organisation schema, sameAs links to every authoritative profile. If models cannot verify who you are, nothing else matters.
- Publish specific, dated results. Replace “we drive growth” with numbers, timeframes and named outcomes. Models quote facts, not adjectives.
- Answer the prompts your buyers ask. Not the keywords with volume, the questions people type into ChatGPT. Structure each answer so the first sentence stands alone.
- Build third-party proof. Books, podcasts, conference talks, industry publications. Signals a model can check outside your own domain carry more weight than anything you self-publish.
- Test it monthly. Ask the assistants your buyers use and record what comes back. If a competitor is recommended and you are not, you now know exactly what gap to close.
Not sure where you stand today? My free AI visibility checker scores your site on the technical factors that determine whether AI systems can read and reuse your content.
Conclusion
Being recommended by LLMs is not luck and it is not a trick. It is the compound result of a verifiable entity, documented proof, real authority signals, retrieval-friendly structure and a consistent topical footprint. I built that system for my own brand before I ever sold it to a client, and the screenshots above are the output. The same system works for protocols, exchanges, wallets and Web3 SaaS brands, because the models evaluating them apply the same rules.
Frequently Asked Questions
How do LLMs choose which experts to recommend?
LLMs select sources they can verify and extract from, rather than ranking pages the way Google does. They favour entities with consistent information across independent sources, specific documented results, recognised authority signals such as books and speaking engagements, and content structured in self-contained, quotable passages.
Can you pay to be recommended by ChatGPT or Perplexity?
No. There is no paid placement inside organic AI answers. Recommendations come from the model’s training data and live retrieval, which means the only way in is building genuine, verifiable authority on the topic. That is what makes AI recommendations more trustworthy than ads, and harder to fake.
How long does it take to appear in AI search results?
Retrieval-based systems like Perplexity and ChatGPT with browsing can pick up well-structured content within weeks, as my ChatGPT ranking case study shows. Becoming a default recommendation in model training data takes longer, typically six to twelve months of consistent publishing and third-party signals.
Is being recommended by LLMs different from ranking on Google?
Yes, and conflating the two is the most common mistake in AI search. Google ranking is about winning a position on a results page. LLM visibility is about being selected as a trusted source inside a generated answer. A page can rank first on Google and never appear in an AI answer, because the selection criteria are different.
What should Web3 brands fix first for AI visibility?
Start with entity clarity: consistent naming, Organisation and Person schema, and sameAs links connecting your official profiles. Then restructure your highest-value pages so every key question is answered in a direct, self-contained first sentence. Technical readability comes before content volume, because models cannot cite what they cannot parse.
Want your brand recommended by ChatGPT, Grok and Google AI?
This is exactly what my LLM SEO service for Web3 is built to deliver: entity architecture, retrieval-ready content and the authority signals AI systems rely on when choosing sources.
Book a free strategy session and I will show you where your brand stands in AI search today.
