How do AI assistants choose which sources to cite?

AI assistants that search the web choose sources in two stages: first a search system retrieves pages its crawler has read and indexed, then the model writes an answer from the pages that state the needed facts most clearly. A page has to pass the first stage to be considered at all; being clear, specific and sourced helps it in the second.

Facts

What happens between a question and a cited answer?

The details differ by assistant, but the published pieces line up:

  • The assistant decides the question needs current information and searches
  • It may run several related searches at once; Google calls this query fan-out
  • A ranking system returns pages from an index its own crawler built
  • The model reads those pages and writes the answer, linking the ones it used

Why does each assistant cite different pages?

Each one searches its own index, built by its own crawler: Google uses Googlebot, ChatGPT search uses OAI-SearchBot, Perplexity uses PerplexityBot. A page blocked for one crawler is invisible to that assistant even when the others cite it.

What makes a page more likely to be quoted?

A page that states the answer in its first sentences and backs it with numbers, dates and named sources gives the model something it can quote safely. Vague pages and pages whose facts contradict each other give it nothing to quote.

Can I see which assistants read my pages?

Yes, in your server logs, by crawler name. easyIndex records those reads per page and pairs them with the visitors each assistant sends, so you see which pages are read and which ones bring people.

Common mistakes

  • Assuming every assistant uses Google's index, and allowing only Googlebot.
  • Using nosnippet on key pages, which also limits what Google's AI features can show.
  • Writing long introductions that push the actual answer far down the page.
  • Judging sources from a single answer instead of watching crawler reads over time.

How easyIndex solves it

easyIndex works on both stages: it checks that the crawlers can reach and index each page, has your coding agent write each page to answer one question with sources, and measures reads and visits per assistant, for US$ 30 per website a month with a 7-day free trial.

Frequently asked questions

Do AI assistants only cite big websites?

No, but large sources are common. Pew found Wikipedia, YouTube and Reddit made up 15% of AI-summary sources, which leaves most citations to other sites.

Does structured data decide which source is cited?

Google says structured data is not required for its generative AI features; it should match the visible text and still helps with rich results.

Do assistants cite pages from their training data?

Links in answers come from search. Knowledge learned in training is not linked, so a page earns a link by being retrievable now.

How quickly can a new page be cited?

Only after the assistant's crawler reads it. easyIndex notifies Bing through IndexNow when a page is published, so it does not wait for the next crawl.

Continue reading

Last verified on .

How AI assistants choose sources to cite