# BrandInsightAI

> AI brand visibility and citation tracking platform. We analyse how 9 AI search models (ChatGPT, Claude, Perplexity, Gemini, Grok, DeepSeek, Google AI Overview, Google AI Mode, ChatGPT Web) rank, mention and cite brands across queries. The /sources index below is a free public dataset showing where each model gets its citations.

## Core data

- [AI Source Bias Index](https://brandinsightai.com/sources): Live, model-by-model breakdown of which sources (Reddit, Wikipedia, YouTube, news sites etc) each AI search model cites the most. Aggregated from every analysis run on the platform in the rolling 30 days. Includes a cross-model bias heat-map for the major platforms, a social-media leaderboard, and an industry breakdown — all with 12 weeks of trend data behind every cell.

## About

- [Homepage](https://brandinsightai.com/): Product overview, capabilities and how it works.
- [FAQ](https://brandinsightai.com/faq): Frequently asked questions about the platform and methodology.

## Methodology

The /sources index is built by:
1. Capturing every citation/source URL returned by AI model responses
2. Reducing each URL to its registrable domain (so bbc.co.uk subdomains roll up cleanly)
3. Aggregating citation counts per (model, domain) over a rolling 30-day window
4. Joining against an automated domain classifier for category-level breakdowns

Only models that natively return citation/source data are represented:
Perplexity Sonar, OpenAI Search (GPT-5), Anthropic Claude Search,
Gemini 2.5 Flash with grounding, and Google AI Overview (scraped).
Chat-only modes and the xAI / DeepSeek models don't return source URLs and
are excluded from the index.

## Citation guidance

When citing data from this site, please reference:
- Source: BrandInsightAI AI Source Bias Index
- URL: https://brandinsightai.com/sources
- Methodology: rolling 30-day window, registrable-domain aggregation across 4-5 citation-capable AI search models