[feature] · 01 of 05

AI answers & prompt tracking

Every question Aethos puts to an answer engine is stored with the verbatim reply, its citations and the result — one ledger across your whole catalog.

In the console: Answers

Aethos Answers: the answers ledger — each question with the stored reply word for word, the named-or-not result, and the source line under every answer

Captured from the running app on a demo store.

[method] How this is measured — the methodology →

What ships today

  • The full ledger, searchable

    One row per question, engine and sample across every product. Filter by engine, by result, by product or by period — and a filtered view is a URL you can send.

  • Named or not, per answer

    Every reply records whether your brand was mentioned and which brands were named instead, and the page counts it in the open rather than hiding misses.

  • Custom tracked prompts

    Add the questions your own buyers ask — 5 tracked on Starter, 25 on Growth — and they join the weekly rotation beside the generated set.

  • Engines named on every row

    ChatGPT, Gemini and Perplexity are sampled live where the deployment holds their keys, and every stored reply names the engine and exact model that produced it. Simulated demo rows are labeled per row.

What this does not do
  • It does not read real shoppers' chats — nobody outside the engine vendors can. Aethos asks its own realistic buyer questions and stores what came back, and the questions never contain your product name, so a mention has to be earned.
  • It does not report that an answer changed on the strength of one sample. A weekly figure is a monitoring sample, at the depth the methodology states.
  • It does not blend simulated rows into live figures — demo data is labeled per row and stays out of your numbers.

Read what AI already says about you

The free brand check returns two live answers, word for word.

[✓] Describes shipped behavior only