Perplexity is now the fifth engine Shruwd measures, beside Google AI Overviews, Google AI Mode, ChatGPT and Gemini. Every plan gets it, the free one included, at the same price.
Where its answers come from
Its answers come from Perplexity's Sonar API. It is the one engine Shruwd reads through an API rather than the app or the results page people use, and an API can answer differently from its app, so read Perplexity's numbers as what the Sonar API says about you. See where the answers come from.
What changes
- Each measurement asks your prompts through Perplexity too. It comes after Gemini in the
engine menu on the overview, prompts, competitors and sources pages, and the API and the
MCP server take
engine=perplexity. - Perplexity's numbers start now, with no backfill. Until it has ten answers, its metrics show as insufficient data, never as zero.
- Perplexity has listed ten sources with every answer so far. Findings read them, and the pages they point to, as they do for the other engines.
Rechecks of earlier fixes
A fix marked applied before Perplexity has answered your questions for 30 days is judged without it, on the engines that answered through the whole baseline, and the finding says Perplexity was left out, as it does for Gemini. See rechecks.