Can I pay to be recommended by ChatGPT, Gemini or Perplexity?

Not for organic answers. OpenAI and Perplexity both describe their search crawlers as the way sites can appear in answers, and neither documents a way to buy an organic recommendation. OpenAI documents ads separately, with its own crawler that checks landing pages submitted as ads.

How long do changes take to show up?

A robots.txt change takes about 24 hours to apply for OpenAI and for Perplexity. For Google, recrawling a page after a change can take days to months, so request a recrawl in Search Console for important pages.

Do I need a different strategy for each assistant?

The foundation is the same: crawlable pages, specific content, consistent facts and third-party mentions. What differs is which crawler each assistant uses and which sources it cites, so check each assistant separately.

Is this different from SEO?

Mostly it builds on it. Google says SEO best practices remain relevant for its AI features. What is new is measuring whether you are named in the answer, and looking at the third-party sources an answer cites.

How to get recommended by ChatGPT, Gemini and Perplexity

You can’t submit your business to ChatGPT, Gemini or Perplexity. Each one builds an answer from web pages and listings it can reach, then names the businesses those sources describe most clearly for the question. So the work is to make your business easy to reach, easy to understand and easy to name, and then to check whether it worked.

This guide covers the five steps, what is not required, and a 30-day plan. We cite the platforms’ own documentation wherever it exists, and we say so when something is our reading rather than a documented fact.

Do ChatGPT, Gemini and Perplexity work the same way?

No, but they share one need: they can only name a business whose pages and listings they can read. Each platform documents its own crawler or eligibility rule.

Assistant	What lets your pages be used	Where it is documented
ChatGPT search	OAI-SearchBot allowed in robots.txt. GPTBot is a separate crawler for model training.	OpenAI crawler overview
Perplexity	PerplexityBot allowed in robots.txt. Perplexity-User fetches a page when a person asks.	Perplexity crawler guide
Google AI Overviews and AI Mode	The page is indexed and eligible to show with a snippet. Google says there are “no additional requirements”.	Google Search Central
Gemini	Google documents Google-Extended as the robots.txt control for training and grounding in its other AI systems.	Google Search Central

Each bot’s setting is independent. OpenAI notes you can allow OAI-SearchBot to appear in search while disallowing GPTBot to opt out of training. OpenAI also says ChatGPT-User, which runs when a person triggers a fetch, is not used for crawling and robots.txt rules may not apply to it.

Step 1: let the right bots in

Open yourdomain.com/robots.txt and check that nothing blocks the search crawlers. A minimal, explicit version looks like this:

User-agent: OAI-SearchBot
Allow: /

User-agent: PerplexityBot
Allow: /

User-agent: Googlebot
Allow: /

Then check the layer most owners forget: your host, CDN or firewall. OpenAI and Perplexity both say to allow requests from their published IP ranges as well as to allow the bot in robots.txt. A security rule that blocks unknown bots can keep you out of answers while your robots.txt looks fine.

Allow about 24 hours for OpenAI and Perplexity to pick up a change. Whether to allow GPTBot (training) or Google-Extended is a business decision. They do not control whether ChatGPT search or Google Search can show you, but allowing them is the widest-reach choice.

Step 2: publish one page for each service and area you really serve

Buyers don’t ask “who is the best dentist”. They ask “which clinic is best for dental implants in Hackney”. An assistant answering that question looks for pages about that service in that place.

Google says AI Overviews and AI Mode may run several related searches across subtopics (it calls this query fan-out) and show a wider set of supporting links than classic search. A page that answers one specific sub-question clearly is a candidate for that. A single “Services” page that mentions ten treatments in one paragraph each is not.

A page that earns its place says:

who the service is for and who it is not for;
what you do, in plain words, and how long it takes;
a real price or a “from” price, in the currency your buyers use;
where you are and the areas you serve;
proof that is specific: qualifications, years, a named case, a real review;
how to book, and what happens next.

Only create a page where you can say something true and different. Pages that swap the place name into the same text risk being treated as doorway pages. We cover the line in location pages that don’t look like doorway spam.

Step 3: keep your facts identical everywhere

Your name, address, phone number, opening hours and services should read the same on your site, your Google Business Profile and every directory. Google’s AI features guidance tells owners to keep Business Profile information current.

Our reading of why this matters: when two sources disagree, a system has to guess which is right, and an answer that has to guess is less likely to name you with confidence. This is reasoning, not a documented ranking rule, but it costs little to fix. Pick one canonical version of each fact and copy it exactly.

Step 4: get named on the pages assistants cite

Assistants often cite third-party pages: local guides, directories, review sites, industry lists. If you ask an assistant a buyer question and open its sources, you will usually see which pages it relies on. If your competitor is on those pages and you are not, that is a concrete, fixable gap.

Ask for a listing or a mention in the way a person would, with real facts and a real reason the page’s readers would want to know about you. Don’t buy or fabricate mentions.

Step 5: measure visibility as a percentage, not a rank

In a study by SparkToro and Gumshoe.ai, ChatGPT and Google’s AI returned the same list of brands in fewer than 1% of repeated runs of the same prompt, and the same list in the same order in fewer than 0.1%. The study was run in late 2025, is not peer reviewed, and was partly sponsored by an AI tracking company, which the authors disclosed.

What held steady was how often a brand appeared across many runs. That is why a single screenshot proves little and why we measure visibility: the share of runs in which you are named. Here is how to run that check yourself.

What you don’t need

Google says you don’t need to create machine-readable files, AI text files or special markup to appear in AI Overviews or AI Mode, and that there is no special schema.org structured data for them. Structured data is still worth having where it matches the page, and we cover it in schema markup for local service businesses.

We publish an llms.txt file on rankweb.app because it costs almost nothing, but Google’s guidance doesn’t call for it and we don’t treat it as a lever.

Why it is worth doing

BrightLocal’s Local Consumer Review Survey 2026, which surveyed 1,002 US adults, reports that use of ChatGPT and other generative AI tools for local recommendations rose from 6% in 2025 to 45%, making it the third most popular source. The same survey found that 40% of consumers trust AI platforms to provide business recommendations and 32% do not. This is one survey of US consumers, so check how your own buyers search.

A 30-day plan

Week 1: check robots.txt and your firewall for OAI-SearchBot, PerplexityBot and Googlebot. Run the 30-minute visibility check and save the results.
Week 2: fix your Google Business Profile and the top directory listings so every fact matches your site.
Week 3: publish pages for the two service-and-area pairs where you are never named.
Week 4: get listed on two of the third-party pages your competitors appear on. Re-run the check and compare.

If you would rather not run the check by hand every month, RankWeb does it for you and drafts the fixes for you to approve.