How we measure citation share, method v1.0
The full method behind every number we publish. Query construction, engine coverage, run counts, what we store, how we parse it, and the four places this method is weakest.
Research
Findings, method notes, and measurements. Everything here states when the data was collected, which engines were queried, and under which method version. Where a number is weak, that is written next to the number rather than in a footnote.
The full method behind every number we publish. Query construction, engine coverage, run counts, what we store, how we parse it, and the four places this method is weakest.
A US commercial real estate company found us through ChatGPT. Three weeks later a Zurich proptech company found us through Google. Here is why we are not calling that evidence, and what we are doing to find out.
A firm selling AI citation whose own site is not readable has no argument left. Every technical decision behind martenfield.com, including the ones that cost us something.
Start here
The audit measures 50 buyer-intent queries across five engines at three runs each, maps the sources the models draw from, and returns a ranked 90-day roadmap. Two weeks, $3,500.