A shift in AI prompt tracking is needed, as measuring the same way as rank tracking is too volatile. When ChatGPT released model 5 in August 2025, almost all AI citation tracking tools showed a drop off because ChatGPT stopped showing as many citation links in the HTML. Third-party tools also only show a small window; one project website has only one to three citations in Copilot according to Ahrefs, but according to Copilot, it actually has over 36,000. One approach is sample design, as outlined by Kevin Indig. The article advocates approaching AI prompt tracking through the dual lenses of volatility and average response tracking. Volatility tracking measures how stable a brand’s presence is within AI model outputs over time. Average response tracking shifts focus to a broader understanding of sentiment, context, and inclusion across a spectrum of related prompts. The measure of success is about pattern recognition over precise placement. Instead of promising a simple upward trajectory, stakeholders must be educated to value risk mitigation, brand sentiment stability, and market share protection within AI models.