Service

AI search (GEO)

Whether an AI assistant can show your site when a customer asks which business to call depends partly on your robots.txt: OpenAI, Anthropic and Perplexity each say blocking their search crawler can keep a site out of their search answers or make it harder to find. This work opens your site to the right crawlers, states clearly who you are, and then checks, the same way each time, whether you get named.

Who it is for

This is for owners who want to know, from evidence, whether AI assistants can find and describe their business. It suits:

What is included

Search crawlers and training crawlers are different bots

Each AI company sends more than one crawler. OpenAI says OAI-SearchBot finds pages for ChatGPT search, and sites that opt out of it are not shown in those answers, while GPTBot collects pages that may be used to train its models (OpenAI). Anthropic splits the work three ways: Claude-SearchBot improves Claude's search results, Claude-User fetches a page when someone asks Claude a question, and ClaudeBot gathers training data (Anthropic). Perplexity says PerplexityBot surfaces and links sites in its search results and is not used to crawl content for AI foundation models (Perplexity).

So keeping your pages out of AI training and appearing in AI search answers are separate decisions, and whether to allow training is yours to make. What we look for is the block nobody chose, such as a plugin's line that shuts out the search crawlers too. The free scan reports a block on OAI-SearchBot, Claude-SearchBot, Claude-User or PerplexityBot as a problem (how that check works). A block on GPTBot or ClaudeBot shows only as a note, because it is your choice and does not keep you out of those search answers (the training note).

Google's AI features run on ordinary search

Google says there are no additional requirements and no special optimizations for appearing in AI Overviews or AI Mode: a page qualifies once it is indexed and can show with a snippet in regular search (Google Search Central). A page marked noindex is out of both (our noindex check).

You may have read that every site needs an llms.txt file. It is a proposal for a file that helps AI tools use a website (the proposal), so we treat it as optional (our llms.txt check).

Saying who you are the same way everywhere

Before an assistant names a business, it has to work out which business it is looking at, and a phone number that differs between pages makes that harder. We line up your site, your Business Profile and your listings so they agree.

Structured data is the part of your site written for machines. For a local business, Google requires the name and the address, and recommends the phone number, the web address, the map coordinates and the opening hours (Google Search Central). We fill in those fields and link your site to your profiles elsewhere (structured data check, business identity check).

How we measure it

Nobody outside these companies can see how often an assistant names you, so we ask. Each round, the same list of questions your customers might type goes to several AI assistants, and the report records whether you were named and states its method. Answers vary from one asking to the next, so one round proves little; the same questions over months show the direction.

What it costs

Published plans that include this service: AI Search add-on, Growth and Authority. Every price is published, and monthly plans run month to month.

AI Search add-on

$500/mo

For the Local plan: AI-answer sampling, entity and schema work, and citation seeding.

Request to start

Authority

$3,750/mo

Everything in Growth, plus eight articles a month, a full AI-search program, a quarterly data study with local press outreach, and conversion fixes.

Or request to start

All prices, add-ons and the terms that apply to every plan are on the prices page.

How the work gets done

  1. You grant access to your website and to Search Console from your own accounts.
  2. We record the starting point: a scan, what your robots.txt allows each AI crawler, your structured data, and a first round of answers.
  3. If a rule blocks a search crawler, we show you the line and ask whether it was deliberate before changing it.
  4. Structured data changes wait on a staging copy of your site until you approve them.
  5. Each round asks the same questions again, and the report sets the answers beside the first round.
  6. When you stop, the work stays in your accounts, and a handover file names every account and the access to take back.

Access we ask for

You grant each role from your own account, and you can remove it from the same screen. We never ask for a password. The access guide shows every step.

What you approve

What we do not promise

Questions

Should I block AI crawlers from my site?

Block GPTBot and ClaudeBot if you do not want OpenAI and Anthropic collecting your pages for training. OpenAI says each of its settings is independent, so you can keep GPTBot out and still let OAI-SearchBot in (OpenAI). Leave OAI-SearchBot, Claude-SearchBot, Claude-User and PerplexityBot alone if you want to appear in those assistants' search answers.

Do I need an llms.txt file?

No. It is still a proposal, so we treat it as optional. If you want one anyway, we add it without claiming it moves anything.

Can you get us recommended by ChatGPT?

No, and nobody outside OpenAI can. We can make sure OAI-SearchBot may read your site and that your site and listings agree about who you are, then check each round whether ChatGPT names you.

How will I know whether AI search work is doing anything?

From the report. It sets each round's answers beside the earlier rounds and shows exactly what was asked.

How is this different from regular SEO?

Most of it overlaps: Google says its AI features need nothing beyond what regular search needs (Google Search Central). What this adds is attention to the other assistants' crawlers and a measure based on asking them questions.

Should I take the AI Search add-on or move up to Growth?

Take the add-on if you are on Local and the scan finds your site in good technical shape: it adds AI-answer sampling, entity and schema work and citation seeding. Choose Growth if the scan finds technical problems or you need new pages, because Growth includes technical fixes and four articles a month along with the visibility report.

Related

Services

Industries

Related reading

Guides

Related terms

Sources