How Google and AI answers pick a business

What ChatGPT is doing when it names a company

ChatGPT, Claude, Gemini, Perplexity and Google's AI Overviews (the summary Google now prints above the blue links) all do the same thing. Instead of handing over ten links, they write a paragraph. Somewhere in that paragraph a company gets named, usually with a small link beside it. The job is being the company named in that paragraph.

Where the answer comes from

The answer comes from two places. You can only change one of them.

The first is training. The model read an enormous amount of text months before anyone asked it anything. You cannot edit that, and it may be out of date about you, or wrong about you.

The second is a live look at the web. When someone asks for a local company, the engine runs a search, fetches a handful of pages, reads them, and writes the answer from what it just read. Those fetched pages are what the citations point at. That step happens after your site changed, which makes it the one that can reflect your business as it is today.

Where each one looks

They do not all look in the same places, which is why one fix rarely shows up everywhere at once.

• ChatGPT runs its own web search. Since July 2026 it has used Yelp's reviews, ratings and business details under a license. Your Google reviews are not part of what it sees.
• Gemini can use Google Search and Google Maps, so it can see your Google listing.
• Claude searches the web on some questions, not on every question.
• Perplexity searches the web for every answer.
• Google's AI Overview sits on top of Google's own results. Google says the same basics that get a page found in search apply to its AI features, and that no special markup or file is needed.

How often people ask, and how often they click

In BrightLocal's Local Consumer Review Survey, published March 2026, 45% of US adults said they had asked an AI tool to recommend a business in the past year. Most of them said they check what it tells them, and about half said they look at whether the reviews are real. When they check, your site, your reviews and your listings have to match.

On Google, a Pew Research Center study published July 2025 found people clicked the ordinary results on 8% of visits when an AI summary appeared, and on 15% of visits when none did. Links inside the summary were clicked on 1% of visits. That is Google's summary only. It describes what people did, not why they did it.

If the search robots are blocked, they cannot read your pages

Those fetches are made by named robots. The names matter because they do different jobs.

• The ones that decide whether you can be found and shown: OAI-SearchBot, Claude-SearchBot, PerplexityBot and Googlebot. These are the search fetches. Turn one of these away and that engine has nothing of yours to quote.
• The ones that are a preference about model training: GPTBot, ClaudeBot and Google-Extended. Blocking these is a decision about whether your words help train a model. It does not decide whether you turn up in an answer, and if you blocked them on purpose, leave them blocked.
• The ones a person sets off: ChatGPT-User, Claude-User, Perplexity-User. These fetch a page because somebody in a chat asked ChatGPT, Claude or Perplexity to open that link. Blocking one stops that single fetch; it does not decide anything else.

A website can turn any of them away in two places. One is robots.txt, a plain text file at the root of a site listing which robots may read what. The other is the network layer: a security or bot-protection service answering the robot with an error before it ever reaches a page. Plenty of sites are turning these robots away without anyone having decided to; it is often a default somebody left switched on.

When a search fetch fails there is nothing to quote and nothing to link to. A model may still mention your name from what it remembers, but it cannot cite you, and it cannot see a single thing you have changed since. Nothing else in this section helps while that is true.

How we check this, and what that check cannot prove. We ask for your homepage from our own servers, using each robot's name. A security service that verifies a robot by its address can refuse us and let the real one through, so a refusal we see is evidence rather than proof. What robots.txt says is settled either way: it is the instruction they all obey, whoever is asking.

Google's list of pages is not the only list

These tools do not all read the same list of pages. Nobody outside those companies can see exactly which list ChatGPT, Claude or Perplexity is using for a given answer. A site can sit in Google and be missing from another list entirely. If the one being asked is using the list you are missing from, it has no path to you. A Google ranking, or Search Console, does not tell you whether the others have your site in their list.

What you can change on your own site

• The visible copy near the top. That is what an engine has to quote, and it can only use sentences it can find. If the trade and the towns are not in it, those facts do not get used. A homepage that says only "Quality Services" gives it nothing to work with.
• Structured data. A small block of machine-readable code inside the page, written in a shared vocabulary called schema.org, that states the business name, address, phone and hours as data rather than as decoration. It takes the guesswork out of which words on the page are the name. It helps the engines parse the name and address. It does not get a company named in an answer. See the last section on this page.
• The same name, address and phone everywhere. On your own pages, on your Business Profile, on any listing that carries you. When those disagree, every engine and directory reading them has to guess which one is you, and some of them guess wrong.
• A complete Business Profile, where you have one. Hours, address, categories and the basics filled in. Reviews matter here too, as a trust signal and a factor in the map, and we will not tell you they are what makes ChatGPT say your name.
• Being crawlable at all. See above. It is the precondition for the rest.

None of these makes ChatGPT name you. They are the difference between the tools being able to read you, and being invisible when they go looking.

The map and the blue links

The list of links, and the map with a few companies on it. Not the AI summary above them, which is the part above. This is still where most people look first.

Titles and descriptions

Every page carries a title, the headline Google usually prints in the result, and a description, the grey sentence underneath it. Two limits worth knowing. Google rewrites titles when it decides its own wording fits the search better, so yours is a strong suggestion rather than a guarantee. And the description has never been a ranking factor; Google has said so plainly. It is still worth writing, because it is often the last sentence a person reads before deciding whether to click.

A title reading "Services - Services" costs you on both counts at once: it tells Google nothing specific, and it tells a person nothing at all. Plenty of sites carry one; it is usually whatever the theme put there.

One page per job, in one town

Search is matching. A person types what they want and where they are. A homepage that lists several services and no town is a weak match for a search that named one job and one city. A page about that job in that city can be a strong match for it, and it gives the engine a URL it can cite.

Speed, on a phone

Google measures load time on mobile and counts page experience as a small factor in ranking. The larger cost is visitors: a page that takes six seconds has lost a share of them before it has said anything, whether or not it ever moves your position.

Links, and why spam anchors hurt

Links from other websites have been part of how Google ranks pages from the beginning. A link is roughly a vote. Which is exactly why they have been bought and manufactured for just as long, and why Google's spam policies name link schemes as a violation.

The visible words of a link, called the anchor text, are part of what is read. When most of the links pointing at a site use unrelated commercial anchor text, meaning wording that has nothing to do with the business, that pattern was built by somebody else, usually with no idea the owner would ever see it. The realistic outcome is that those links are discounted rather than counted. Google says most sites will not need its disavow tool, and that it is for spammy links that have caused, or are likely to cause, a manual action, so we watch this and tell you what it looks like rather than selling you a cleanup.

Nobody can promise a ranking

Google's own guidance on hiring an SEO says to walk away from anyone who guarantees rankings. Nobody outside Google controls the ranking. It is decided by a system nobody outside the company can see, against other companies who are also working at it, and it changes.

What we can promise is the work: the pages written, the settings corrected, the blocks removed, and a straight account afterward of what changed and what did not.

What we do not sell, and why

Same list as the homepage, because we get asked. Other firms will quote you for these. We will not. Here is why, for each one.

• An llms.txt file. It is a proposed file for telling AI tools about a site. No tool has said it reads one, and we have never seen it change anything. Charging for it would be charging for something that has not worked.
• Structured data as a way into AI answers. Schema is worth having. It states your name, address and hours as data instead of as decoration. It is not what gets a company named in an answer. Anyone selling it as that is selling neat markup as if it were a result.
• A rank position in an AI engine. There is no ranking to hold. ChatGPT writes a paragraph and names whoever it names. Ask again and the list can change. A number in that space would be made up.
• A disavow file. Google says most sites will not need the disavow tool, and that it is for spammy links that have caused, or are likely to cause, a manual action against your site. If your links have junk in them, we will show you what it looks like and whether it comes anywhere near that.
• Credit for a change we cannot prove we caused. Answers and results move on their own, all the time. Unless we hold a baseline and compare, nobody can say a fix caused a change. We report what changed and stop there.
• Bing as a route into ChatGPT. Being listed on Bing is ordinary upkeep and worth doing for Bing. We cannot see which list ChatGPT actually reads, so we will not sell Bing as the way into ChatGPT.

If somebody quotes you for one of these, that does not make them a scam. Ask what result they expect, and how they would prove it happened.