Prompt history and answers

Select a prompt in the Prompts table to open its detail panel. It shows everything Rankry knows about that one question.

At the top

Its own period picker, so you can look back further on one prompt without changing the whole page.
Tiles: scans that named you, your position when named and the engines that named you.

Who wins this prompt

The brands the answers name for this question, ranked. If a brand keeps appearing and you don’t track it, Add to competitors adds it from here.

Your place in each engine’s answer

A grid of scans by engine: where you were named, and where you weren’t. A sparkline shows the trend, and the history list below has each scan with its result.

The full answer

Open a scan to read each engine’s answer as it came back:

the answer itself;
the brands it named, in order;
the sources it cited;
the sentiment of what it says about you.

Google AI Overviews and AI Mode answers are shown the way Google laid them out. For local businesses that includes the map and the list of places.

Meta Muse and Copilot answers say where they came from and when: Muse official API response and Azure GPT + Bing API response.

Full answer text is kept for recent scans. For older ones, you see the ranked brands and sources Rankry pulled from the answer.

When an answer fails

Sometimes an engine fails to answer. The panel lists those answers, and the prompt’s Status column in the table shows a retry icon. A failed answer uses no report requests; retrying one uses one when it arrives.

How to use it

Read why a rival wins. The answer usually says it: a feature, a price, a review, a source. That’s what to match or beat.
Look at the sources. If the same review site or forum thread shows up in several answers, that’s a page to get onto. See Sources.
Watch the grid after you ship a change. A new page or listing shows up first in the engines that search the web hardest.