> ## Documentation Index > Fetch the complete documentation index at: https://docs.frictionai.co/llms.txt > Use this file to discover all available pages before exploring further. # Analysis > Diagnose brand health across visibility, sentiment, purchase intent, providers, prompts, sources, and trends. # Analysis Analysis is the deep-dive view for brand health. Use it to understand why AI Score is moving and which metric, provider, source, or prompt explains the movement. Brand audit analysis page with AI Score, performance trends, metric tabs, and model comparison Brand health analysis is available on Growth, Professional, and Enterprise plans. ## What to inspect Top-level recommendation strength across the selected period. Switch between visibility, sentiment, and purchase intent. See whether movement is recent, sustained, or isolated. Compare performance across available AI providers. Review the prompts and responses that contributed to the score. Inspect cited domains and search keywords tied to brand health. ## Core metrics Measures whether the brand appears in relevant answers, how prominent it is, and how consistently it appears across providers and prompt types. Measures the tone and framing of brand mentions. Weak sentiment often points to outdated sources, mixed reviews, or competitor-led narratives. Measures whether AI guides users toward choosing or buying from the brand in high-intent contexts. ## Workflow Start with 30 days. Use 7 days for recent changes and longer windows for stable trend interpretation. Open the metric tab that changed most or is weakest against expectations. A provider split often reveals whether the issue is model-specific, search-specific, or broad. Use prompt evidence to see the answer text and citations behind the score. Use CSV export for source, keyword, or stakeholder reporting workflows. ## Common diagnostics Check whether competitors gained share of voice, owned pages disappeared from sources, or prompt coverage shifted. Look for negative or cautious language in responses and identify the sources that support it. Compare the answer against competitors. AI may mention you but choose another brand as the recommendation. If one provider behaves differently, inspect that provider's citations, retrieval behavior, and answer style before making a global conclusion. ## Related pages See whether AI recognizes your brand before relying on live search. Compare brand health trends against the promoted deep-dive competitor. Turn underperforming prompt evidence into recommended fixes. Audit website pages for AI-discoverability issues.