Skip to main content
Yes. Prefer captures every URL that ChatGPT, Gemini, Google AI Overviews, Google AI Mode and Perplexity cite in answers to your tracked prompts, with the engine, prompt and position. It tags each source (owned, review, forum, listicle and more) and shows which pages earn citations for you and for each competitor. Last updated 3 October 2026 Yes. Prefer records every URL an AI engine links to when it answers one of your tracked prompts, then rolls those URLs up so you can see where each engine goes for information. You get both halves of the picture: which of your own pages earn citations, and which outside pages earn them for your competitors.

​
What Prefer stores for each citation

Prefer stores 4 facts for every citation: the engine, the prompt, the position in the answer and the destination URL. Duplicates are removed per run, prompt, engine and URL, so a long answer that links the same page twice still counts once. The Sources and Citations docs define a citation as any URL an engine links to as part of its answer for a tracked prompt. Those raw rows feed one headline number, Citation Share: citations to your domain divided by all citations in scope. Like every Prefer metric, it recomputes when you filter by engine, topic, persona, region, tag or date range, and it shows a change vs the previous run.

​
How sources are classified

Prefer tags each cited URL with 1 of 9 source types, so you can filter by intent instead of raw domain. The docs add a practical note: in software categories, listicles and reviews often drive the highest share of citations, so getting included or corrected there usually comes before writing new pages.

​
Where you read citation data in the app

Citation data shows up in 3 places in Prefer, from the whole account down to one prompt.
  • Monitor, Overview, Top cited content: your owned URLs ranked by citation share, so you know which pages earn the most references.
  • Monitor, Competitors, Sources: which domains cite which brands, split per engine.
  • Prompt detail, Top Domains and Top Citations: the source list for one prompt, with links out.

​
Find the pages you should own

To turn citation data into work, follow the 4-step routine from the docs:
  1. Open Competitors, then the Sources tab.
  2. Filter to a single competitor.
  3. Sort by citation share. The top rows are pages engines cite for that rival and not for you.
  4. Create an Action from that row in the Action Center: an outreach task, a guest piece or an owned page that can compete for the same slot.
Check that a gap holds on more than 1 engine before you act on it, since the engines cite different sources. Our track AI citations workflow and the competitive analysis module cover this end to end. To see which sources the engines cite in your category today, run a free AI visibility audit.

​
Frequently asked questions

No. Prefer keeps the two apart. A mention is your brand named in the answer text; a citation is a URL the engine links to as a source. You can be named without being cited and cited without being named, so Prefer reports Visibility Score and Citation Share as separate numbers.
Yes. Prefer’s Competitors console has a Sources tab that maps which domains cite which brands, split per engine. Filter it to one competitor and sort by citation share: the top rows are sources that cite that rival and not you.
No. Prefer deduplicates citations per run, prompt, engine and URL. A page cited twice in the same ChatGPT answer counts once for that run, which keeps Citation Share from being inflated by one long answer.

​
Also asked as

  • Can Prefer show the sources ChatGPT cites for my category?
  • Does Prefer track AI citations by URL?
  • How do I see which of my pages get cited by AI in Prefer?

​
Sources