How do AI-powered search engines like Perplexity rank and cite sources?

Perplexity’s ranking is proprietary but the inputs are observable from the citation patterns. Recency is weighted heavily, so a recently-published authoritative article often outranks an older one on the same topic. Domain authority signals work the way they do across the industry: government, academic, major news, and Wikipedia rank consistently high; mid-tier industry publications rank well for their domain; thin blog content rarely appears. Topical relevance matters: a generalist outlet covering a niche financial topic loses to a specialist outlet covering it well. Link patterns and structured data (tags in a page’s code that tell engines what the content is) round out the signal set. The inline citation layer is the verification mechanism: a user can see exactly which sources Perplexity used and decide whether to trust the synthesis, which is also what makes the engine easier to influence through targeted source-layer work (improving the sources the engines read, not the engines themselves).