> ## Documentation Index
> Fetch the complete documentation index at: https://chatobserver.com/docs/llms.txt
> Use this file to discover all available pages before exploring further.

# How do prompts and AI answer allowances work?

> Understand the difference between stored prompts and collected answers, and manage your monitoring allowance.

A prompt is a saved question and its monitoring configuration. An answer is a collected response from a selected platform and region. Running one prompt across several combinations can collect several answers, so prompt count and answer usage measure different things.

## Plan your monitoring set

The main choices affecting recurring monitoring are the number of active prompts, selected platforms, selected regions, and run frequency. Review them together when estimating capacity.

Use the [pricing allowance estimator](https://chatobserver.com/pricing) to compare monitoring scenarios, then check the current plan and usage in your account. Your billing period and purchased capacity determine the allowance available to your organisation.

## When monitoring is paused by an allowance limit

Review the usage state and billing options before requesting more runs. A saved prompt can remain in the workspace even when collection cannot proceed. Changing its wording or repeatedly clicking **Run now** does not resolve an allowance hold.

You can reduce future demand by pausing lower-priority prompts or changing their [schedules](/docs/getting-started/schedule-prompts). Review plan and capacity changes in **Billing** with the billing owner.

## Keep units clear in reports

Use answer counts to explain the evidence behind a visibility percentage. Use prompt counts to explain the questions covered by your research. Neither count measures visits to your website; those belong to [AI referral analytics](/docs/data/ai-referrals).