GEO for people: the complete guide
Generative Engine Optimization (GEO) for people is the practice of making a professional identity findable, rankable, and accurately describable by AI systems: recruiter sourcing engines, and assistants like ChatGPT, Claude, Perplexity, and Gemini.
Updated 2026-07-04 · Surfd research
Why this exists
Two shifts happened at once. Recruiter tools stopped matching keywords and started ranking by meaning: they embed profiles into vectors and compare them against role descriptions across more than a billion indexed people. And people stopped clicking ten blue links: assistants now synthesize one answer that names a handful of people. You are in that answer or you are invisible. The overlap between what ranks on Google and what AI engines cite has collapsed from roughly 70% to under 20%, so a good Google presence no longer carries over.
How recruiter AIs rank you
Sourcing engines like SeekOut, hireEZ, and Eightfold build a unified profile of you from LinkedIn, GitHub, publications, talks, and the open web, whether you asked or not. They embed that profile and rank it by semantic similarity to a role query. Your exact keywords matter less than your evidence: projects, outcomes, corroborated skills, and recent activity. Some tools now expose this directly to assistants over MCP, so a recruiter literally asks an AI for a shortlist.
How LLMs decide who to name
Assistants cite what they can retrieve and trust. The measured levers: definition-first openings lift citation rates around 2.1x; adding specific statistics lifts them around 40%; quoting named experts has driven gains above 100% in some categories; content modified within roughly 90 days is strongly preferred for time-sensitive queries. High-trust domains dominate: being described on one page an engine already trusts beats ten pages it ignores.
The levers, in order of leverage
First, be crawlable: allow the retrieval bots (OAI-SearchBot, ChatGPT-User, GPTBot, ClaudeBot, Claude-SearchBot, Claude-User, PerplexityBot, Perplexity-User, Google-Extended, GoogleOther) and check your CDN is not silently blocking them. Second, be structured: Person schema, clear headings, dated entries, an llms.txt. Third, be specific: real numbers on real outcomes, on every surface. Fourth, be corroborated: the same true story on your site, GitHub, and LinkedIn. Fifth, be fresh: touch your surfaces inside every 90-day window. Sixth, be present where engines already look: earned mentions on trusted domains.
Common questions
What is GEO for individuals?
GEO (Generative Engine Optimization) for individuals is optimizing how AI systems find, rank, and describe a person: allowing AI crawlers, publishing structured and specific evidence, corroborating claims across surfaces, and keeping them fresh.
Does ranking on Google make me visible to AI engines?
Mostly no. The overlap between Google rankings and AI citations has collapsed from roughly 70% to under 20%. AI engines retrieve and cite differently, favoring structured, specific, fresh, high-trust sources.
Can I trick recruiter AI with hidden keywords?
No, and you should not try. Modern sourcing tools detect hidden text and prompt injection and penalize them. The durable strategy is improving real signal: evidence, structure, corroboration, and freshness.
What is an llms.txt?
A plain-text file at a site's root that points language models to its most citable pages. Adoption is early and its impact is debated, but it costs nothing and signals machine-friendliness.