Are you tracking traffic coming from AI, and what is the most important metric for measuring AI's impact on your website?

Yes, we actively track AI visibility and AI influenced discovery patterns across ChatGPT, Gemini, Perplexity, Claude, Grok, and Google AI Overviews. One of the biggest challenges is that traditional analytics still struggle to fully capture AI influenced discovery journeys. A user may first encounter a business through an AI generated recommendation, then later return through branded search or direct traffic, masking the original influence source. Because of that, we measure more than referral traffic alone. In our current AI visibility tracking framework, we measure: - 40% Seen by AI — visibility across 30+ tracked prompts, per platform - 6% recommendation visibility where AI systems actively recommend or cite the business - 1% "Chosen by AI" visibility where the business becomes the top recommended option We also use a three tier visibility model: - Memory visibility: whether AI models already recognize the brand from training data - Live visibility: whether the brand appears during real time AI assisted searches - Forensic visibility: whether the brand surfaces during actual browser based buyer journeys The most important metric is not raw AI referral traffic alone. It is whether AI systems confidently retrieve, recommend, and reinforce the business during meaningful buyer decision moments.