Which sources matter most?

Why there is no fixed list

The engines do not share one source list, and none of them keeps its own list still. ChatGPT, Gemini and Perplexity each build an answer from the sources they trust for that question on that day, and the mix shifts as new coverage appears, a thread gets active, or a page loses whatever made it worth citing. A source that shaped your summary in one engine can be absent from another, and gone from both a month later.

Across most categories the sources fall into a handful of kinds: news outlets, Reddit, YouTube, review sites, Wikipedia and the company's own pages. Which ones lead for your questions is what the measurement is for. Reach is a poor guide; the engines cite communities and creators as readily as the press, and often skip the trade outlet you worked hardest to land.

What the report gives you

Every citation in every answer is logged, so the sources come back as a table rather than a guess:

• Citation domains and cited pages, ranked, with each one's share of citations for your questions
• A tag on each source: yours, a competitor's, or third-party
• The change since the last period, with movers flagged as new, grew, shrank, or dropped

Your newsroom and press pages are tracked as owned sources alongside the outlets and the communities, so you can see whether the engines are reading you or only reading about you.

How the list becomes a plan

The briefing turns the ranking into placement targets: the outlets the engines actually read in your category, the threads worth a reply, and the pages your newsroom is missing. That is a target list ranked by what gets cited, not by circulation.