How to get your business recommended by ChatGPT (what worked for me)
Short answer
AI assistants recommend businesses they can find, understand and trust. Make your site easy for them to read (plain text, clear service pages, structured data), say exactly who you help and with what results, keep the same facts on every profile you have, and make sure search engines index you. Then test it: ask the assistant the questions your clients would ask.
Key takeaways
- Write your services the way clients ask for them, with real results and numbers.
- Let AI crawlers in and give them a plain-text version of your site.
- Keep the same name, services and facts on your site, LinkedIn and every profile.
- Test with real client questions, and ask every new lead how they found you.
What happened
A few days ago a head of brand at a US company filled in the form on my site and messaged me on LinkedIn the same evening. He had asked an AI assistant to find someone for landing pages, funnels and A/B testing, and it recommended me.
So I tested it. I asked ChatGPT, logged out and with web search on, five questions a client might ask. For four of them it put me first, as the “best overall match”:
- “Who are good freelance CRO designers for landing pages, funnels and A/B testing?”
- “I run Meta ads for my ecommerce store but my landing page doesn’t convert. Who can I hire?”
- “I’m a US solar company and our cost per lead is too high. Who should redesign our landing page?” It cited my solar case study by name.
- “Who is a good freelance landing page designer for Facebook ads who can design and code the page?”
For the fifth, a SaaS question, it didn’t mention me at all, which showed me exactly which pages to improve. One caveat: I tested from Serbia, and answers can change with location and over time.
What I changed on my site
- Service pages written the way clients ask. Not “conversion-focused solutions” but “landing page design for ad campaigns” and “CRO for SaaS”, each with who it’s for, what’s included and a real result.
- Real case studies with numbers. Every case says what was wrong, what I changed and what the A/B test showed, like +20% conversion and 15% lower cost per lead for a solar landing page. AI answers quoted those details back.
- A plain-text version of the site. An llms.txt file with a short summary of who I am, my services, case results and answers to common questions, plus a markdown copy of every page. AI tools read plain text more easily than designed pages.
- AI crawlers allowed. My robots.txt welcomes search and AI crawlers like OAI-SearchBot, ChatGPT-User, PerplexityBot and ClaudeBot. Many sites block them by accident through a security setting.
- Structured data. Schema markup that says who I am, what I offer, my certifications and my profiles, so assistants don’t have to guess.
- The same facts everywhere. My LinkedIn, Contra, Dribbble and Behance profiles use the same name, services and results as the site, and all link back to it.
- Indexed by search engines. AI assistants search the web through search engines, so I submitted my sitemap to Google Search Console and Bing Webmaster Tools.
Steps any small business can copy
- Ask the assistant first. Type the three questions your best clients ask before they hire someone like you, with your city or niche. Write down who gets recommended and what the answer says about them.
- Make one page per service and per audience. Say who it’s for, what’s included, the result and how to start. Use your clients’ words.
- Show proof with numbers. Before and after, results, reviews with names, years in business. Vague claims get ignored.
- Answer real questions on the page. A short FAQ with price ranges, timelines and “do you work with…” questions gives assistants something to quote.
- Check your robots.txt and firewall. Make sure AI and search crawlers aren’t blocked.
- Get mentioned elsewhere. Directories, profiles, partner sites and reviews tell assistants that other people vouch for you.
- Ask every lead how they found you. AI visits often show as “direct” in analytics, so asking is the only reliable way to know.
What doesn’t help
- Stuffing pages with keywords or with “best” and “top” claims you can’t back up.
- Hiding your services behind images, sliders or PDFs that crawlers can’t read.
- Different names, prices or services on different profiles.
Then make the visit count
Being recommended only gets the visit. The visitor still has to understand your offer, trust you and take the next step on your site, which is what I work on every day. If your site gets visitors but few enquiries, start with how to get more leads from your website.
Want a site that gets found and converts?
I design and build websites and landing pages that search engines and AI assistants can read, and that turn the visit into a lead or sale, then A/B test them. Website design →
FAQ
How do I get ChatGPT to recommend my business?
Make your site easy to read and understand: clear service pages in your clients’ words, real results with numbers, a short FAQ, structured data, AI crawlers allowed and a plain-text llms.txt summary. Keep the same facts on every profile and make sure search engines index your site.
Does ChatGPT use Google or Bing to find businesses?
AI assistants search the web through search engines, so submit your sitemap to both Google Search Console and Bing Webmaster Tools and check that your pages are indexed.
What is llms.txt?
A plain-text file at the root of your site that summarizes who you are, what you offer and your key pages, written for AI tools. It’s a proposed standard, not a requirement, but it gives assistants a clean summary to read.
How do I know if AI assistants send me visitors?
Visits from AI assistants often show as “direct” in analytics. Ask every new lead how they found you, and test your visibility by asking the assistants the questions your clients would ask.
Sources
I’m George (Djordje) Peric, a CRO designer, CXL certified in Conversion Optimization and Ecommerce Marketing. I design websites, landing pages, funnels and ads, and A/B test them.