Monitoring AI engine responses to specific user queries (prompts) over time to track citation patterns, share of voice, and brand mentions per query.
Prompt-level tracking is the foundational measurement layer for AEO programs. Instead of tracking only 'how often does our brand show up in AI?' (aggregate), prompt-level tracking measures 'how often does our brand show up for THIS specific query?' Tracking is typically done by running each prompt against each engine multiple times per day (5x, 25x, or 50x sampling) and logging citations, sentiment, and competitor mentions. Tools like Profound, Peec AI, Scrunch, and Lantern all do prompt-level tracking as their core feature.
Aggregate metrics hide where AEO is winning and where it's losing. Prompt-level tracking surfaces specific opportunities: 'we're cited for prompt X but invisible for prompt Y; let's optimize content for Y.' This is the difference between knowing you have an AEO problem and knowing where to fix it.
Lantern tracks 150 prompts per customer on the Growth tier. Each prompt runs 25 times per day across 4 AI engines = 15,000 API calls per customer per day. The data is aggregated weekly to show share of voice trends per prompt and identify prompts where the customer is gaining or losing visibility.
The terms in this glossary aren't theoretical — they're what Lantern's product calculates and reports every month for B2B SaaS teams. See yours in 7 days. 14-day free trial.
Join Waitlist