# How Is Your Business Performing in AI Discovery?

> Canonical source: [How Is Your Business Performing in AI Discovery?](https://www.answer.cloud/pages/how-is-your-business-actually-performing-inside-llms-like-chatgpt-hosted/)

Combine crawler requests, human referrals, answer tests, and on-site chat evidence without treating one signal as complete visibility.

There is no single website metric that reveals everything said about your business inside AI services. Use several complementary sources of evidence.

## Four useful signals

- **Crawler and user-fetch requests:** server or edge logs show which URLs were requested. They do not reveal every search query or prove a citation.
- **Human AI referrals:** browser analytics can show some visitors arriving from identifiable AI sources. Missing referrers limit coverage.
- **Repeatable answer tests:** record the provider, model or product, settings, prompt, date, citations, and outcome for a defined question set.
- **Your website chat:** observed questions can reveal what visitors need and where your content needs clarification.

Many crawler requests do not execute your browser analytics script. For an existing Cloudflare website, the [Worker guide](https://www.answer.cloud/docs/cloudflare-worker-agent-analytics-hosted/) explains request-level collection.

Use [answer.cloud](https://www.answer.cloud/) to publish a reliable source and observe the signals available to your site. Keep measurement definitions explicit with the [metrics guide](https://www.answer.cloud/pages/llm-visibility-metrics-the-kpi-playbook-for-seo-and-marketing-leaders-hosted/).