AE-200 · Lesson 11 of 12 · Module 4: Measure & Scale · 2:09
Measure what wins
Measure the whole chain: are you mentioned, are you cited, is the answer accurate, do AI crawlers read your pages (server logs), and do people arrive from AI answers (referrals in analytics) and convert? Rank prompts by what their pages actually earned. A prompt can win without being cited, and a cited prompt can bring no one. Visits beat mentions.

What you will learn
- Track mentions, citations and accuracy across assistants
- Measure AI crawler reads and AI referral visits
- Rank prompts by the visits they bring (winning prompts)
The one thing to remember
Visits beat mentions.
Common mistake: Reporting only screenshots of mentions
They look good but do not show what the business earned.
The fix: Keep a scorecard per prompt: reads, mentions, visits, sign-ups.
Key terms
- AI referral traffic
- Visits that arrive from an AI assistant, seen as referrals in your analytics.
- Winning prompt
- In Sophyx, a prompt whose published pages actually earned AI-crawler reads, indexing and visits.
Part of Answer Engine Specialist (AE-200)
How AI answers are made, and how to get a brand named in them. The core course, and the place to start.
12 lessons, practice quizzes, a final exam and the Certified Answer Engine Specialist certificate. Regular price USD 89. Your first certification is free. Free with any paid Sophyx plan.
Start the course free