How answers are collected

aeotime asks every active prompt on every engine in your plan and stores the full answer, the sources it cited and the searches it ran when the engine exposes them. Seven of the eight engines search the web live; DeepSeek answers from what the model learned in training, which shows how your brand is represented in model knowledge.

Engines

Engine	How it's collected
ChatGPT	OpenAI API with web search on
Perplexity	Perplexity API with web search
Gemini	Google Gemini API with Google Search grounding
Claude	Anthropic API with web search on
Google AI Overviews	Google search results data
Google AI Mode	Google search results data
Grok	xAI API with search on
DeepSeek	DeepSeek API — no web search, so answers come from the model's own knowledge and have no citations

Market, language and persona

Each prompt is asked for a market (country) and in the project's language. The engine is told where the user is located and asked to prefer sources relevant to that country. If a prompt has a persona, it's added as context before the question.

Schedule

• Weekly — every active prompt runs once a week on your plan's schedule.
• Run now — start an extra run from the project overview at any time.
• Paused projects and paused prompts are skipped.

What's stored for each answer

• The full answer text, with your brand and competitors highlighted
• Brands mentioned, in order, with the sentiment toward you
• Cited URLs and domains
• The search queries the engine ran (fan-out), when available

Open Answers to read them, filter by prompt, engine, market, persona, mention or sentiment, and export CSV.

Why results differ from the apps

Answers from APIs can differ slightly from what a signed-in user sees in the consumer app, which personalises answers and changes frequently. aeotime's method keeps every run consistent so trends are meaningful. When you check a result by hand, use a logged-out or private session in the same country for the closest match.