How Businesses Get Recommended by ChatGPT and AI Search

More people now ask an AI assistant instead of typing a search: “Who builds Shopify stores near Kitchener?”, “What should I look for in a website care plan?” The assistants answer with a short summary, sometimes naming businesses and linking to sources. Getting your business into those answers is often called generative engine optimisation (GEO) or AI search optimisation.

Let’s be clear up front: nobody can guarantee how or when an AI assistant mentions your business. Anyone who promises placement is guessing. What you can do is make your business easy for these systems to access, understand, trust and quote. That’s the same foundation good SEO is built on, with a few additions.

How AI assistants find information

Broadly, assistants draw on two sources:

• What they learned during training: large amounts of text gathered from the web before a cut-off date.
• What they retrieve live: many assistants search the web or their own index when answering, then summarise what they find and cite sources.

Either way, your website, your business listings and what others say about you are the raw material. If that material is hard to reach, unclear or inconsistent, you’re less likely to appear accurately, or at all.

The four things AI needs from your site

1. Access: can AI crawlers reach your pages?

• Check your robots.txt file. Some sites block AI crawlers without realising it, often through a security plugin or a default setting.
• Different crawlers do different jobs. OpenAI, for example, uses GPTBot for training and OAI-SearchBot for search; Anthropic, Perplexity and others have their own. Google-Extended controls whether your content is used for Google’s AI models; it doesn’t affect your appearance in Google Search. Decide deliberately what you allow.
• Make sure your important content is visible in the page’s HTML, not only after heavy JavaScript runs.

2. Understanding: is it clear what you do, where and for whom?

• One clear page per service, with plain headings.
• Say where you’re based and which areas you serve, in plain sentences.
• Use structured data (Organization, Service, LocalBusiness where accurate) so machines can read your key facts.

3. Trust: are your facts consistent everywhere?

• The same business name, address, phone and description on your site, your Google Business Profile and directories.
• Real authorship and real proof: named people, real case studies, published prices or policies.
• What other reputable sites say about you: reviews, local press, industry associations.

4. Citability: do you answer questions clearly?

• Answer common customer questions directly, near the top of the relevant page, in a sentence or two that could be quoted on its own.
• Back answers with specifics: prices, timeframes, what’s included.
• Keep facts current. An assistant quoting an old price or a discontinued service does you no favours.

What about llms.txt?

llms.txt is a proposed convention: a plain-text file at the root of your site that summarises its key pages for language models. It’s cheap to add and does no harm, but it’s not a standard that all assistants have committed to reading, so treat it as a small extra rather than the main event. (Our own site publishes one.)

What about Google’s AI features?

Google’s own guidance is that the same SEO best practices apply to its AI features as to ordinary search, with no special markup required. If your site is crawlable, helpful and accurate, you’ve done most of what’s needed. Our SEO services start from exactly that foundation.

What doesn’t work

• Stuffing pages with questions and keywords written for machines rather than people.
• Fake reviews or invented credentials. Assistants cross-reference sources, and inconsistencies undermine trust.
• Mass-produced AI content that adds nothing new. If your page says what a hundred others say, there’s no reason to cite you.
• Tools that promise to “submit” you to AI assistants. There’s no such submission process for general answers.

How to check where you stand

1. Run our free AI search check. It looks at AI crawler access in robots.txt, machine-readable summaries, structure and structured data on a page.
2. Ask a few assistants the questions your customers ask, including your town. Note who’s mentioned and which sources are cited.
3. Check that your business facts are consistent across your site and main listings.

A practical starting plan

1. Fix access: review robots.txt and make sure key content renders without scripts.
2. Clarify: one page per service, clear location statements, accurate structured data.
3. Make facts consistent across your site and listings.
4. Write answer-first content for your top customer questions.
5. Earn genuine mentions from reputable local and industry sites.
6. Re-check every few months; assistants and their sources change quickly.

Technical companies have extra reasons to get this right; see our guide for startups outgrowing their template site. For local businesses, much of this overlaps with local SEO; our local SEO checklist covers the listings and reviews side.

What “answer-first” content looks like

Here’s the difference in practice. A typical service page might open:

“Welcome to our company. We are passionate about delivering excellence and have been serving our community with dedication…”

An answer-first version opens:

“We install and service residential heat pumps in Cambridge, Kitchener and Guelph. Most installations take one to two days, and we offer free in-home assessments.”

The second version tells a person, a search engine and an AI assistant exactly what you do, where and what to expect, in sentences that make sense when quoted on their own. Apply the same idea to FAQs: ask the question the way customers ask it, answer in the first sentence, then add detail.

Frequently asked questions

Is AI search replacing Google?

Not replacing, but changing how people search. Many people use both. The good news is that the foundations overlap: a site that’s accessible, clear, accurate and trustworthy does well in both.

Should we block AI crawlers?

It depends on your goals. Blocking training crawlers keeps your content out of future model training, but blocking search-focused crawlers can stop assistants citing you in live answers. Decide crawler by crawler rather than blocking everything.

How do we know if it’s working?

Check periodically: ask assistants your customers’ questions, note whether you’re mentioned and which sources are cited, and watch for referral traffic from AI tools in your analytics.

Get help

Our AI search and GEO service starts with a $99 AI Search Audit that checks access, structure and consistency and tells you what to fix first. We improve the signals you control and report honestly on what changes. No guarantees, because nobody can honestly give them.

Written by the CHRISPS studio in Cambridge, Ontario. AI crawler names and Google guidance are as published by the respective companies at the time of writing; they change often.