Home › Guides › How do AI answers choose which businesses to name?
How do AI answers choose which businesses to name?
An AI answer names the businesses it can assemble a confident description of. Between the buyer's question and the written answer, the engine retrieves sources, resolves them into a record of each candidate business, and composes the answer from the records it trusts. A business with a consistent, corroborated, citable record gets named. A business with a thin or contradictory record gets skipped, and the answer names a competitor whose record lines up. This guide walks through that selection, step by step, because the selection is the thing the AI Visibility Audit measures.
What happens between the question and the answer?
Four things, in order. First, retrieval: the engine fetches candidate pages and profiles for the question from its indexes and the live web. Second, resolution: it works out which mentions belong to the same business, the same person, the same place. Third, corroboration: it compares the resolved records across sources and settles on the facts the sources support. Fourth, composition: the language model writes the answer, naming the businesses whose settled records fit the question.
Notice what is missing from that sequence: a ranking, of the kind the old results page showed. A business can sit in the retrieved set, even rank well as a link, and still never reach the answer, because composition works from records, not links. The question the buyer asked is answered from whatever the engine could assemble, and assembly is where most businesses are lost.
What is an entity record?
Your entity record is everything the engine has gathered and attributed to your business as one entity: its name, its category, its services, its location and service area, its hours, its phone number, its prices where they are published, the people who run it, its profiles, its reviews, and the mentions of it on other sites. The record exists whether or not you have ever looked at it. It was assembled from your website, your Google Business Profile, your listings, and third-party pages, in whatever order the crawlers found them.
Entity schema is how a business takes part in its own record. Schema with a stable identifier and sameAs links declares that this website, this profile, and this listing are the same entity, and states the entity's facts in machine-readable form that matches the visible page. The GEO guide describes the verification pipeline in full. Without that declaration, resolution is guesswork, and the engine's guesses about which mentions belong to you are exactly as good as the mess it was handed.
Why does consistency decide who gets named?
Because corroboration is a vote, and consistency is how a fact wins it. When your website, your Google Business Profile, and your listings disagree, the answer engine goes with the version it can corroborate. Three sources stating the same category, the same services, and the same phone number outweigh one louder source stating something else. The engine is not judging which version is true. It is reporting which version it could verify.
The failures are concrete. The website calls the business a plumber, the profile calls it a handyman service, and an old listing calls it a general contractor: three blurred records instead of one clear one. The address changed in one place and not the others. The founding service, the one buyers ask about, appears on a brochure site and nowhere the crawlers go. Each blur costs confidence, and the answer's names are the businesses that cost it the least. This is also why the fix is unglamorous. Consistency is a ranking signal precisely because it is work most businesses never finish.
What makes a business easy to cite?
Citability is a property of how facts are written down. A canonical page that states what the business is, in the category language buyers search, gives the engine one authoritative source to resolve everything else against. Direct answers in buyer language, stated at the top of the page section that answers them, can be lifted into an answer without rewriting. Published prices, where the business can publish them, settle questions competitors leave vague. FAQ content answers the follow-up questions. An llms.txt file hands the whole record to language models in one plain text document.
None of this asks the engine for a favor. The engine is under pressure to answer quickly and defensibly, and it reaches for the sources that cost it the least effort and the least risk. A business whose facts arrive pre-assembled, consistently, in quotable sentences is simply the easiest correct answer available. Being good at the work matters after the buyer calls. Getting named is a clerical contest held before that, and it is won on the quality of the paperwork.
What makes an AI answer skip a business?
The skip list, from the audits, is repetitive:
- Crawlers blocked: robots.txt or a firewall default keeps the AI crawlers out, so the pages never enter the sources.
- No plain statement of category: no page says, in so many words, what the business is and what it sells, so retrieval finds nothing to resolve.
- Contradictory facts: the website, profile, and listings tell different stories, and corroboration fails on the details that matter.
- A broken profile: a suspended, duplicate, or miscategorized Google Business Profile corrupts the record at one of its primary sources.
- Facts locked in unreadable forms: the key statements live inside images, scripts, or documents the engines read poorly, if at all.
- No third-party corroboration: nothing anywhere repeats the business's own claims, so every fact rests on a single source.
Read the list again and notice what it does not contain: the quality of the business's work. Skips are record problems. The engine is not reporting that a competitor is better at your trade. It is reporting that a competitor's paperwork assembled more cleanly, and paperwork can be fixed.
How do you find out what the answers say about you?
You run the questions yourself, before you fix anything. Take the searches that bring you revenue, ask them in Google AI Overviews, ChatGPT, Perplexity, and Gemini in the words your buyers use, and write down what comes back: named or skipped, described rightly or wrongly, and who was named in your place. Then compare the answers against your record, find the disagreements and the gaps, fix them in priority order, and re-run the same questions monthly to confirm the answers moved.
That test, run properly and written down, is the AI Visibility Audit on the Gerris services page. It is $750 flat. We test how your business appears across Google AI Overviews, ChatGPT, Perplexity, and Gemini for the searches that bring you revenue, check your entity and name, address, and phone consistency, and deliver a written report with every fix in priority order. The report is yours to keep regardless of what follows.
Frequently asked questions
Why does ChatGPT recommend my competitor instead of me?
Usually because the competitor's record is easier to assemble: their category, services, and location are stated plainly on a page the engine can read, and the same facts appear on their Google Business Profile and listings. The engine names the business it can describe with confidence. The fix is rarely one trick. It is making your record as consistent and citable as theirs.
What is an entity in AI search?
An entity is one real-world thing the engine tracks as a single record: a business, a person, a place, a service. Your business's entity record is everything the engine has gathered about it from your website, your profiles, and other sites' mentions. Entity schema with sameAs links tells the engine which pages and profiles belong to the same entity.
Can a business pay to be named in an AI answer?
Not in the organic answers this guide is about. There is no placement to buy inside the written answer. What a business can do is fix the record the answer is built from: consistent facts, a canonical page, readable answers to buyer questions, and profiles that agree. That is the whole of the method, and it is also the product Gerris sells.
How much does it cost to find out what AI answers say about my business?
The AI Visibility Audit is $750 flat. Gerris tests how your business appears across Google AI Overviews, ChatGPT, Perplexity, and Gemini for the searches that bring you revenue, and delivers a written report with every fix in priority order. The report is yours to keep regardless of what follows.
Read next: What is answer engine optimization (AEO)? and What is generative engine optimization (GEO)?