In short
ChatGPT and other AI assistants recommend the suppliers that many consistent, easy-to-quote sources on the web talk about: beyond your own website, company databases, review sites, industry media, LinkedIn and list articles. You get on the list by measuring where you stand now, finding out which sources you are missing from, and filling those gaps in a targeted way. Nobody can promise a guaranteed position, but the movement can be measured.
Have you ever asked ChatGPT who it recommends in your own field? If not, now is a good time. A buyer looking for a new supplier does exactly that: they type in their question and get a short list of company names. If your name is not on it, you miss out on that request for proposal without ever knowing it.
In this article, we look at how AI puts this list together, how you can find out where you stand now, and what you can do to make it think of you.
How do buyers use AI to find suppliers?
Buyers ask AI the same thing they would have typed into Google before, just in a longer and more precise form, and instead of links they get a ready-made shortlist.
A typical question sounds like this: "Which Hungarian companies take on industrial automation projects for a mid-sized factory?" or "Which ERP implementation partner should a manufacturing company choose?". The answer typically lists a few companies with a short justification, often with source links. Besides ChatGPT, the same happens in Gemini, Claude, Perplexity and Google AI Overviews, which we covered in a separate article.
This list is not the decision yet. But it decides who gets invited to submit a proposal at all.
Why does it matter whether you are on the AI shortlist?
Because in B2B, the winning supplier is almost always on the first shortlist already, and a company that is left off it rarely gets back into the race later.
Close to 4,000 B2B buyers responded to the survey behind the 6sense 2025 Buyer Experience Report. According to it, in 95% of cases the winning vendor was already on the shortlist the buyer started with on day one of the buying process. A year earlier, this figure was 85%.
Important: the 6sense survey is not about AI. It is our own conclusion that if buyers increasingly use AI assistants to put together this first list, then appearing in AI answers directly influences who ends up on it.
No report shows you this loss. You do not lose the deal at the proposal stage or in the negotiation, but before you even know about it.
What does AI base a supplier recommendation on?
AI repeats what the web consistently says about you: your own website and the external sources that mention, compare or review you.
In B2B, external sources matter at least as much as your own site:
- Company databases and business information sites: what you do, where, and how big your company is.
- Review sites: where clients write about you and where the players in your category are compared.
- Industry media: trade articles, interviews and case studies that name you.
- LinkedIn: your company page and the professional presence of your team.
- List articles: "best providers" style roundups, which AI answers often reference.
That is why one more blog post changes little if your competitor is on the sites AI uses as sources and you are not. We call this difference the source gap. Your own site really helps when AI can easily read from it what you do, for whom and where: with short, clear answers, accurate company details and structured data.
How can you find out who AI recommends in your industry right now?
By asking the same buyer questions on several AI platforms, several times, and counting how often and in which position each company appears in the answers.
Trying out a question or two is a good first signal, but it is not a measurement: AI answers can differ from run to run and from user to user. That is why we ask 30 typical buyer questions for an industry on 5 AI platforms (ChatGPT, Google AI answers, Gemini, Perplexity, Claude), three times per question. From the answers, we look at three numbers:
- AI share: what portion of all company mentions goes to you compared with your competitors.
- Appearance rate: in what percentage of answers you appear at all.
- Recommendation position: where, on average, AI mentions you in its list.
If you want to see where you stand now, the free AI visibility assessment is made for exactly that: after a short conversation, we prepare your industry's AI ranking and show you where you rank in it.
Curious who AI recommends in your industry right now?
After a short conversation, we prepare your industry's AI ranking free of charge: who AI recommends for typical buyer questions, and where you rank compared with your competitors.
Get my industry's AI ranking →How do you get on the list?
By removing the reasons AI does not think of you right now: technical barriers, content that is hard to interpret, and missing external sources.
The order matters. This is also the process we follow in our AI visibility service:
- Baseline measurement: where you stand compared with your competitors, and what AI says about you that is wrong.
- Source gap analysis: which sites carry weight in your category, and which ones you are missing from.
- Technical foundations: AI crawler access, indexing, sitemap, structured data.
- Consistent company details: the same name, activity and contact details everywhere, from your Google Business Profile to LinkedIn.
- Answer blocks and buyer-focused content: short, quotable answers on your key pages to the questions your customers actually ask.
- External presence: well-maintained company profiles, and a presence on the external sites AI uses as sources.
- Follow-up measurement: the same questions on the same platforms, so that any movement comes from the work, not from a change in methodology.
Nobody can responsibly promise you a guaranteed first place here, because AI answers are not controlled by any single player. What can be promised: the change is measurable, and you can see in numbers which way your position is moving.
How long does it take to see results?
The effect of technical fixes can be checked within weeks, while the effect of content and reputation work typically shows up in AI answers within 2-6 months.
That is why AI visibility is not a one-time setup, but a cycle of measuring, fixing and measuring again. At the end of the first month, a technical re-check makes the most sense; lasting change shows up in later measurements on the same set of questions. How many inquiries all this brings in is worth measuring separately. We describe how, step by step, in this article.
Frequently asked questions
Does ChatGPT recommend the same companies to everyone?
Not necessarily. Answers can differ from run to run, from user to user and depending on the exact wording of the question, so it is not worth drawing conclusions from a single test. You get a reliable picture by asking several questions repeatedly.
Is it enough to optimize my website?
Rarely. In B2B, AI answers are shaped at least as much by external sources, such as company databases, review sites, industry media, LinkedIn and list articles, as by your own website.
Can it be guaranteed that AI recommends you first?
No. Nobody can responsibly guarantee AI answers. What you can measure is how your AI share, appearance rate and recommendation position change compared with your competitors.
How do I find out where I stand now?
In the free AI visibility assessment, after a short conversation we prepare your industry's AI ranking, which shows who AI recommends right now and where you rank.