How AI assistants answer questions about Expo 2027 Belgrade, and how a company becomes the cited source

From two places: what the model learned in training, and what it finds on the web when it searches at the moment of the question. Training data has a cut-off date and is not linked to sources. Search is where citations come from. ChatGPT, Claude and Perplexity each run a search crawler that reads public pages, and Google's AI Overviews and AI Mode draw on pages in the Google Search index. For an event that changes week by week, such as a participant count or a ticket date, the search side is what matters.

That makes the job familiar. To be quoted in an answer about the Expo, a page first has to be found, read and understood by a crawler, and then be the clearest source for the question asked. Google says it plainly for its own AI features: there are no additional requirements to appear in AI Overviews or AI Mode beyond being indexed and eligible to show with a snippet. The work is good search practice, done carefully. Our SEO and GEO service is built around it.

The AI companies document separate crawlers for separate jobs, so you can allow search while making a different choice about training. OpenAI says sites that opt out of OAI-SearchBot will not be shown in ChatGPT search answers, and that GPTBot is the crawler whose content may be used for training. Anthropic warns that blocking Claude-SearchBot may reduce a site's visibility in user search results. Perplexity says PerplexityBot is for search results and not for training. Google-Extended controls Gemini training and grounding and, Google says, does not affect inclusion or ranking in Google Search.

If you want to be cited, do not block the search crawlers in the second column. Whether to allow the training crawlers is a separate business decision. Check your robots.txt, and also any firewall or bot protection in front of your site, because a security rule can block a crawler that robots.txt allows. Our own site allows every crawler and names several AI crawlers explicitly.

Specific answers to the questions people actually ask, in text a crawler can read. An assistant asked which companies at the Expo offer a certain product, or who in Belgrade can help with something, looks for a page that says so clearly, near the top, in a sentence that stands on its own. Vague pages ("innovative solutions for a connected world") give it nothing to quote. A page that says what you do, for whom, where and since when does.

• Text in the HTML, not only in images, video or scripts that run after the page loads.

• A direct answer under each question-form heading, in two or three sentences that make sense alone.

• Facts with a date and a source: "140 international participants as of 30 June 2026, according to the organisers", not "over 100 countries".

• Your company name, address, contacts and description the same on your site, your profiles and every directory.

• A page per topic, so one page is clearly the best answer to one question.

• A visible "updated" date, changed only when the content really changes.

Possibly, and it costs little, but do not count on it. llms.txt is a proposal from September 2024 for a plain-text file at the root of a site that summarises the site and links its key pages for AI tools. Support varies: several AI companies publish llms.txt files for their own developer documentation, but the proposal page names no AI search product that commits to reading one. Google says that you do not need AI text files or new markup to appear in its AI features. We publish one on this site because it is cheap to generate from the same content as the pages; we treat it as a convenience, not a ranking factor.

Structured data describes a page in a standard format: this is an organisation, this is an article, this is its author, this is its date. It helps search engines connect your pages to one entity and understand what each page is. Google asks that structured data match the visible text, so it adds clarity to what is on the page rather than replacing it. Useful types for an Expo company are Organization (with the same name and contacts as everywhere else), Article for guides and news, and FAQPage where a page really answers questions.

With a written list of the facts you publish and where each comes from. Before the Expo, check them monthly; from April, every two weeks; and whenever the organisers publish news. Change the number, the date next to it and the page's "updated" date together. An assistant that finds two different counts on your site may quote neither. Remove what has expired, such as a talk that already happened, instead of leaving it to be quoted in August.

This is how the guides on this site work: each Expo fact sits in a box with its source, each article shows the date its facts were last checked, and the same facts appear in our Expo 2027 Belgrade for companies guide. The full digital plan is in digital marketing for Expo exhibitors.

Our SEO and GEO service checks that crawlers can read your pages, sets up structured data and llms.txt, and writes pages that answer real questions with dated, sourced facts. We cannot promise a citation; nobody honestly can. See how it fits the rest of your preparation on our Expo 2027 page, or send us your site on the quote page.