Websites & AI search

How to get your business recommended by ChatGPT (and found on Google)

More people ask ChatGPT or Perplexity who to hire. How AI assistants choose who to recommend, and eight steps to make your site one of the answers.

One of my current collaborations started in an unusual way. Someone asked ChatGPT to find a young AI entrepreneur, and it pointed them to my website. A few messages later we were working together.

That is how a growing share of customers now look for a supplier. Instead of scrolling through ten blue links, they ask an AI assistant "who can build X for me?" or "which accountant near me does Y?" and get two or three names. If your business is not one of those names, you are not in the race.

The good news: the work that gets you recommended by AI assistants is mostly the same work that gets you found on Google, done a bit more deliberately.

How AI assistants decide who to recommend

When you ask ChatGPT, Perplexity or Google's AI answers for a recommendation, they draw on two sources:

  1. A live web search. For most "who should I hire" questions, the assistant searches the web, reads a handful of pages, and writes an answer from them. ChatGPT search uses its own crawler and third-party search providers, partly Bing. Perplexity has its own index. Google's AI answers use Google's index.
  2. What the model already knows. Large language models learned from a huge snapshot of the web. If your business was mentioned consistently and in many places, the model may already know who you are.

In both cases, assistants favour businesses that are clearly described, specific, and mentioned by other trusted sources.

1. Say plainly what you do, for whom, and where

The most common problem I see: a website that is beautiful but never states, in plain words, what the business sells. "We create meaningful digital experiences" tells an AI nothing.

Give every main service its own page, with a title and first paragraph that answer: what do you do, for whom, where, and with what result? For example: "AI agents and workflow automation for small businesses in Belgium."

2. Answer the questions your customers actually ask

Assistants look for passages that directly answer the user's question. Write those answers on your site:

  • How much does it cost, or what does the price depend on?
  • How long does it take?
  • How does the process work, step by step?
  • Is it right for my situation?

A good FAQ section on each service page does a lot of this work.

3. Be specific

Specific claims get quoted. Vague ones get skipped. Names of industries, cities, numbers, timelines, credentials and real examples all help an assistant decide that you are a relevant answer.

"We help companies with digital transformation" will never be quoted. A claim like "cut invoice processing time from 16 to 3 hours a month for a 20-person accounting firm" might be, as long as it is true.

4. Make your site easy for machines to read

  • Fast, clean pages that render their content in the HTML, not only after heavy JavaScript.
  • Structured data (schema.org) that tells search engines who you are, what services you offer and where.
  • A sitemap, submitted to search engines.
  • Don't block AI crawlers. Check your robots.txt. Blocking crawlers like OAI-SearchBot or PerplexityBot means those assistants cannot read your pages.
  • An llms.txt file is a newer, proposed standard: a plain-text summary of your site for AI systems. Its impact is not proven yet, but it costs almost nothing to add.

5. Register with Google and Bing

Set up Google Search Console and Bing Webmaster Tools and submit your sitemap to both. Many businesses skip Bing, but several AI assistants rely on Bing's index, so it matters more than its search market share suggests.

6. Get mentioned on other trusted sites

This is the biggest lever for "who should I hire" questions. Assistants trust what other sources say about you more than what you say about yourself:

  • Directories and marketplaces relevant to your field (for agencies: Clutch, Sortlist, Malt; for local businesses: Google Business Profile).
  • "Best X in Y" lists and roundups. Many are written by other businesses in your field. Get on them by having a clear service page, then asking.
  • Press, podcasts and partner websites. A link from a client's or partner's site counts.
  • Reviews. Ask every customer, consistently.

7. Tell the same story everywhere

Your website, LinkedIn, directory profiles and press mentions should describe you the same way: same name, same one-line description, same services, same location. Consistency is how a model becomes confident that all those mentions are about the same business, and what that business does.

8. Publish what you know

Articles that answer real questions in your field show expertise and give assistants something specific to cite. You do not need to publish every week. A handful of genuinely useful guides, kept up to date, does more than dozens of thin posts.

How to measure it

Classic SEO tools only show part of the picture. Add these:

  • Ask the assistants. Once a month, ask ChatGPT, Perplexity, Gemini and Claude the same ten questions your customers would ask, and note whether and how you appear.
  • Check your analytics for visits from chatgpt.com, perplexity.ai and similar sources.
  • Ask new customers how they found you, and include "ChatGPT or another AI assistant" as an option.

The short checklist

  • One clear page per service, saying what, for whom and where
  • FAQs that answer real customer questions
  • Specific numbers, examples and credentials
  • Structured data, sitemap, fast pages, AI crawlers allowed
  • Google Search Console and Bing Webmaster Tools
  • Mentions on directories, lists, partner sites and in reviews
  • The same description everywhere
  • A few genuinely useful articles

This site is set up the same way, if you want to see it in practice. More on how I build websites here.

About the author

Martin Michaux is Head of AI at Herald and was named to Forbes 30 Under 30. Before Herald he co-founded NOX Energy. He writes about what he learns building AI agents in production.

Profile AI workflows More articles
Reply

Book a call with Martin