Use an AI mention monitoring tool (Profound, Otterly.ai, Peec AI, AthenaHQ) or manually run a weekly test set of 30 to 50 adviser-relevant prompts across ChatGPT, Claude and Perplexity. Log every citation in a database so you can see trends over time.
Tools to use - **Profound** — enterprise AI-mention tracker. - **Otterly.ai** — focused on brand mentions in LLMs. - **Peec AI** — multi-engine tracking with dashboards. - **AthenaHQ** — agency-focused reporting.
DIY approach 1. Build a prompt set (30 to 50 typical queries). 2. Run weekly across ChatGPT, Claude, Perplexity and Gemini. 3. Record citations, snippets and source URLs in Supabase or a spreadsheet. 4. Track citation rate (% of prompts that cite you).
KPIs to track - Citation rate per engine. - Share of voice vs competitors. - Trend over 4 to 12 weeks. - Which pages get cited most.
Typical realistic target 10 to 25% citation rate across a branded prompt set is achievable within 90 days of a serious GEO programme.