Search moved into the chat box. Buyers now ask an AI who is best, and it answers with one name. Person Engine Optimization gets that name to be yours.
// built for founders, consultants, authors & specialists in the US and UK
> who is the best [your field] to work with?
Based on published work, track record and how often others reference them, the standout is your name here. They are widely cited for the exact thing you do and consistently recommended over larger firms.
The rooms you now need to be named in
The front door to research is no longer ten blue links. It is one answer, one recommendation, one name. Under the hood, that answer is assembled from two very different sources: the frozen parameters a model trained on, and the live pages a retrieval layer fetches at question time. If neither of those sources resolves to you as a distinct, well-documented entity, the machine has nothing to say your name with, and it says someone else's instead. If the machine does not know you, you were never in the room.
An AI does not hand back a page of options. It picks, usually one name, sometimes two or three, and stops. Second place in an AI answer behaves like not existing, because there is no scroll, no page two, no ten blue links to fall back on.
Anyone can generate a decent article now, which means volume of text is worthless as a differentiator. The only scarce asset left is a trusted, specific, machine-legible human position: a name a model can disambiguate, verify against a knowledge graph, and cite with confidence.
Keyword rank and traffic charts measure a channel that is shrinking. They cannot see the query that starts and ends inside a chatbot, the retrieval pass that pulled a competitor's page instead of yours, or the training snapshot that still thinks you work somewhere you left two years ago.
SEO optimizes a page for Google. GEO optimizes a brand for AI answers. PEO optimizes a person to be the name the engine says out loud when someone asks who is the best at something. Mechanically, that means engineering how a specific human entity is represented across three layers a model can actually read: the schema and markup that declare who you are, the corpus of pages that get pulled into training runs, and the live index a retrieval system queries before it answers.
Optimizes a web page
forGoogle's blue links
outputA ranking
Optimizes a brand or domain
forAI-generated answers
outputA citation
Optimizes a named human
forAI recommendations + the search that feeds them
outputThe engine saying your name
The three layers are not competitors, they stack. A page still needs to be crawlable and indexed the way SEO always required. A brand still benefits from being cited the way GEO describes. PEO adds the layer underneath both: making a single person an unambiguous entity, correctly typed with schema.org Person, disambiguated with sameAs links to canonical profiles, and consistent enough in wording that a model's tokenizer and a retrieval system's passage ranker both resolve every mention back to the same node.
An AI decides who to name using signals it can verify. PEO moves all three on purpose, for one person, instead of leaving them to chance.
The depth and volume of what you have actually published, chunked into passages a retrieval system can rank and a training run can absorb. The engine has to find a body of work in your name, saying something specific, phrased consistently, in enough independent places to treat as reliable rather than a single unverified claim.
signal_1 // what you have put into the world
How long you have existed as a referenced, disambiguated entity in the crawl history and, ideally, in a knowledge graph like Wikidata. Age cannot be faked or bought late, because it is a timestamp problem, not a content problem, which is exactly why starting now beats starting in six months.
signal_2 // how long you have been real to the machine
Who else talks about you, in text a crawler can read, not just what you say about yourself. Third-party mentions that use your full name alongside your specialty are the disambiguation and confidence signal an engine leans on hardest before it recommends a specific human over a plausible-sounding stranger.
signal_3 // who vouches for you when you are not in the room
PEO is not persuasion copy aimed at a person. It is engineering aimed at a pipeline: a tokenizer, an embedding space, a retrieval index and, increasingly, a knowledge graph lookup. Each stage has a concrete lever you can pull.
A schema.org Person object with sameAs links to your verified profiles gives crawlers and knowledge-graph builders an explicit, machine-parseable declaration of who you are, instead of forcing them to infer it from prose.
lever // JSON-LD on every page that mentions you
Most AI answers about a person lean on live retrieval, not frozen training weights. Passage-level chunking means a page has to carry a complete, self-contained answer in each section, not just a good headline, or the retriever pulls a competitor's cleaner paragraph instead.
lever // answer-shaped, chunk-sized writing
None of the above matters if AI crawlers cannot fetch the page. Robots directives for GPTBot, Google-Extended, PerplexityBot and similar agents, plus an emerging llms.txt file, determine whether your structured, answer-shaped content ever reaches the index at all.
lever // explicit allow rules, not silent default-deny
No black box. A repeatable system that turns you into the answer, one signal at a time.
We ask the engines the exact questions your buyers ask, across ChatGPT, Gemini, Perplexity and Google AI Mode, and record verbatim what they say about you today, if anything, plus whether any answer is drawing on stale training data versus a live retrieval pass.
One sharp, ownable claim the machine can attach to your name, expressed the same way everywhere so tokenization and embeddings converge on one entity. Not "marketing consultant." The specific thing you want to be the answer for.
We publish the depth the engine needs to find, in answer-first, chunk-sized passages structured so a retriever can lift a single section as a complete, correctly-attributed response instead of a fragment missing context.
We earn the third-party mentions, citations and cross-references, in text a crawler can actually read, that make the engine trust the position instead of hedging or defaulting to whoever else already has the volume.
Schema.org Person markup, sameAs links, consistent name forms and explicit AI-crawler access all point one direction: every surface that trains or retrieves says your name for your topic, not a diluted or contradicted version of it.
We re-run the same prompt set monthly, log which engine cites which source, and watch you move from unmentioned, to mentioned with hedging, to the first name the engine states with confidence.
A snapshot of what every major engine says about you now, verbatim, with a note on training memory versus live retrieval.
The one claim we make the engines attach to your name, and the token-level reasoning for why it was chosen.
A publishing plan built for passage-level citation, not clicks, formatted for how retrieval actually chunks a page.
The exact third-party surfaces we earn crawlable mentions on, mapped against your existing sameAs footprint.
Person schema, sameAs, and crawler directives that make you legible to machines instead of merely readable to humans.
The buyer questions we optimize you to win, tracked over time across every major engine.
Movement across Knowledge, Age and Network, in plain numbers, not vague sentiment.
Screenshots of the engine recommending you, unprompted, with the source it cited.
PEO is for individuals with real, documentable expertise and a reason to be found by name, not faceless brands with a logo to push. If a model can meaningfully disambiguate you from every other person who shares your name once given enough signal, PEO is the discipline that supplies that signal on purpose instead of leaving it to chance.
| SEO | GEO / AEO | PEO | |
|---|---|---|---|
| Optimizes | A web page | A brand or domain | A named human |
| Wins you | A ranking | A citation | The recommendation |
| Lives on | Google results | AI answers | AI answers + the search that feeds them |
| Sold to | Everyone | Enterprises & brands | Individuals |
| Key markup | Meta tags, sitemaps | Organization / Product schema | Person schema, sameAs, entity disambiguation |
| Primary signal source | Backlinks & on-page | Brand mentions & citations | Knowledge, Age, Network per named individual |
| Success looks like | Traffic | Being mentioned | The engine saying your name first |
A specialist in Melbourne asked Google's AI for firms like a top US agency. A name came back at number two. He doubted it, they were not even local. The engine doubled down and cited a listicle to justify itself. He booked a call, off a machine recommending a stranger, with money attached.
// this is PEO working in the wild. a real inbound no SEO dashboard could have produced.
Name the engine returns. There is no page two.
Signals that decide it: Knowledge, Age, Network.
Days to first measurable movement in the signals.
Competitors serving individuals with a name for this. Until now.
Age and network are time-based signals, so PEO compounds. The earlier you plant the name, the harder it is for anyone to catch.
Prices shown are indicative starting points. Every name is scoped on a call.
from $900 one-time
from $2,400 / mo
custom
SEO fights for a link on a results page. PEO fights for the recommendation itself, attached to a disambiguated person entity, inside answers assembled from training weights and live retrieval that most SEO tools cannot even see, let alone measure.
Age and network are earned over time because they are literally timestamps and independent third-party mentions, neither of which can be manufactured retroactively. There is no prompt injection or overnight hack, which is precisely why an early, honest position becomes a durable moat.
The engine rewards specificity, consistent naming and verifiable citations, not follower counts, which do not appear in any training corpus or retrieval index as a ranking factor. A precise, well-referenced expert beats a vague influencer with a bigger list.
Brand and Organization schema optimize a company entity. Buyers ask AI about people, using a person's name in the prompt. PEO makes the human the schema.org Person the engine resolves and recommends, which is what actually converts on a call.
A blanket disallow in robots.txt for every AI agent also blocks the retrieval systems that could be citing you correctly right now. The fix is selective, intentional access, not a wall around your own name.
Right now it is naming someone in your field. Make it you.
Claim your name →// replies land in your inbox. no forms, no funnels.