How do you get found in ChatGPT, Gemini and Claude?
You get found in ChatGPT, Gemini and Claude by first being easy to find on the web and by answering your customer's exact question directly on your site, with prices, dates and comparisons. Special tricks are not needed. I read the sources myself, because much of what is online about this topic is mainly there to sell a tool.
Example: a page that answers one question, with a price → Custom software, SaaS & AI assistants: what I build → All Datawonder prices on one page →How do you get found in ChatGPT, Gemini and Claude?
You get found in ChatGPT, Gemini and Claude by doing three things well: letting AI crawlers in, putting your text on the page as plain text and answering your customer's exact question directly. Chatbots search the web and choose from what they find, so without a findable page there is nothing to cite. After that, a recent date, a concrete price and an honest comparison help.
What does Google say about appearing in AI Overviews?
Google says no extra requirements or special optimisations are needed to appear in AI Overviews or AI Mode. According to Google you do not need new machine-readable files, AI text files or special structured data. The basis is ordinary search engine optimisation: allow crawling, offer important content as text and make structured data match what is visible on the page.
How does ChatGPT get its search results?
ChatGPT uses the OAI-SearchBot crawler for its search feature. According to OpenAI, sites that block this crawler are not shown in ChatGPT search answers. GPTBot is something else: it collects text for training. So you can switch off training and still be findable. Check your robots.txt and your CDN, because a block can be in there without you noticing. A change can take about 24 hours.
Want to know what can be automated in your business? Describe your bottleneck →
What does research on AI citations show?
A 2026 study with 252,000 trials on six language models shows that relevance to the question and position in the search results carry by far the most weight. After that, a recent date and an explicit price help. Specifications, comparisons, evidence and assertive language give a smaller gain. Text formatting had no consistent effect. It was an experimental set-up with product reviews and not the real services, so treat it as a direction.
Does an llms.txt file help you get found?
I have found no reliable evidence that an llms.txt file gets you into answers more often. Google says you do not need such files. This site does have one, because it takes a quarter of an hour and AI agents find a tidy summary in it. See it as a business card, not as a way to rank higher.
What do you do in practice to get cited?
To get cited, start with these six steps. Check that OAI-SearchBot, PerplexityBot, ClaudeBot and Googlebot are not blocked. Make sure your text is in the HTML and does not only appear after JavaScript. Make one page per customer question with the answer in the first sentence. State amounts, even as a range. Add a real date and only update it when something changes. Compare honestly, including with options that are not yours.
How do you measure whether AI chatbots find you?
You measure whether AI chatbots find you in three ways: through your statistics, Google Search Console and the chatbots themselves. Check your statistics for visits from chatgpt.com, perplexity.ai and gemini.google.com. Track in Google Search Console which questions get impressions. Put your customer questions to the chatbots themselves every month and note who gets named. Count on months rather than weeks, because here too you have to be found before you can be cited.
The short version
Let crawlers in, put your text in the HTML and answer one customer question per page in the first sentence, with price, date and comparison. According to Google, special files and tricks are not needed. Measure monthly and do not expect results within a few weeks.
Want to know what can be automated in your business?
Describe your bottleneck →