Engines and collection
Last updated
An engine is an AI product that answers questions, such as ChatGPT or Google AI Overviews. Aivting collects the answers the engines give to your prompts and stores them. It does not read your own chats and does not control what any engine says.
How to: No answers yet · Filter your dashboard
Which engines are collected
Solo and Agency Lite collect ChatGPT, Google AI Overviews and Google AI Mode. Agency Pro adds Gemini and Perplexity.
In the engine filter, each engine is in one of three states:
- On. Answers are collected and counted.
- Off, not collected. The engine is listed, but Aivting does not collect it, so it has no data.
- Not offered. Aivting does not offer this engine.
Each engine is a separate product, so its numbers are reported on their own. Filter by engine before you compare.
When collection happens
A run starts when an owner or editor clicks Run now, and the run is processed while its run page is open. Prompts run when you start a run. Automatic daily runs are on our roadmap. Creating a project does not start a run. The chart has one point per day, the UTC date of each run; Aivting is not real time.
Automatic collection
Each project has an Automatic collection setting under Settings. Owners and editors can switch it on or off for the project. When it is on, Aivting starts runs for the project without anyone pressing Run now, and Settings shows the next scheduled run in UTC.
- Not switched on yet. Automatic collection is not running yet. Until it is, the setting is saved and Settings says so; nothing runs on its own, and the next scheduled time is not shown.
- What is skipped. The demo project, projects with the setting off and projects with no active prompts.
- Same limits as Run now. A scheduled run goes through the same spend checks as a run you start yourself.
- Collection paused. If a scheduled run cannot start, Settings shows "Collection paused." with the reason, for example "Collection is switched off right now." or "An engine's spend limit is reached." Items that could not be collected show as blocked in the run and are never counted as "not mentioned".
- Turning it off. Answers are then collected only when someone presses Run now.
Answers per prompt per run
A run collects a set number of answers per prompt from each collected engine. More answers per prompt give a steadier rate, because one answer says little. See Sample size (n).
Answer statuses
- ok. The engine returned an answer. Only these answers count toward n.
- no answer. The engine returned nothing usable for this prompt. Counted separately, never as "not mentioned".
- failed. The collection did not complete. Counted separately.
- blocked. The item was not collected because a spend limit was reached. Excluded from n.
What Aivting does not do
Aivting collects the answers ChatGPT and other engines give to your prompts. It does not guarantee that any engine will name or cite a brand.
Related
Runs and answers · Filters and date ranges · Limits
ChatGPT is a trademark of OpenAI. Google and AI Overviews are trademarks of Google LLC. Aivting is an independent product and is not affiliated with or endorsed by OpenAI or Google.