Tracking AI Mentions

Start with the questions buyers actually ask before choosing, run each one more than once on every platform that matters, and record who was named and which sources were shown. Do that before buying a tool. A single visibility score cannot be trusted until you have seen the variance underneath it: in our tests the same query produced different visible sources across platforms and across repeated runs, and tools define the score differently.

The articles below cover how to choose the query set, how many runs a claim needs, what an AI visibility score does and does not measure, and why a citation count says nothing about who owns the cited source.

Set up the measurement

1. Use Voice of Customer to Choose Which AI Search Queries Matter

   Use Voice of Customer as the starting point for AI search tracking, then use search data, product priorities, market language, and platform signals to test and expand the query set.

2. How Many Times Should You Run the Same AI Search Query?

   There is no defensible universal repeat count. Three, five, ten, or 100 runs do not become reliable simply because the number sounds substantial. The sample should match the decision you need to make.

Read the numbers

1. AI Visibility Measurement: Your Score Went Up. What Actually Improved?

   An AI visibility score is the share of tracked AI answers in which a brand appears, and Profound, Semrush, and Otterly each compute it differently. The number alone cannot tell you whether the company was cited, recommended, or described accurately.

2. ChatGPT vs Perplexity Citations: Same Query, Different Sources

   The same provider-selection query produced largely different visible source sets in the tested ChatGPT and Perplexity experiences. See why AI visibility should be measured by platform, mode, query, and repeated run — not collapsed into one score.

3. AI Citation Counts Hide the Commercial Interests Behind the Sources

   A manual review of 62 cited domains shows why companies need to examine publisher ownership and commercial relationships — not just count AI citations.