Two proposals can report the same brand at 38% and 11% and both be arithmetically correct. XstraStar has published the denominator behind every metric it reports.
SINGAPORE, August 12, 2026 /PRNewswire/ — Spending on AI search visibility now passes through procurement, and buyers are hitting a problem the category has not solved: the numbers in competing proposals are frequently not comparable.
Two vendors can measure the same brand over the same period and report 38% and 11%. Both can be correct. They count different things under the same word.
One counts answers: across a fixed set of questions, how often the brand is named. The other counts sources: what proportion of all cited sources belong to the brand’s domain. The denominators are questions and citations respectively. Reported side by side they explain each other; reported under one label, they mislead.
The mismatch surfaces at acceptance, when a metric named in a commitment carries a different definition from the same-named figure in the report. Because mention rate depends entirely on the question set, a rate quoted without that set can be moved in either direction by editing the questions, with no change to the website.
“Ask what the denominator is, about every percentage in the deck,” said Dean Luo, Chief Technology Officer at XstraStar. “Anyone who has done this work before will answer without hesitating.”
XstraStar has published its measurement definitions as part of a 219-page reference library, in English and Chinese. Twenty-two metric pages state, for each measure, what it counts, what it does not count, and what it cannot be used to prove. The library contains no vendor rankings and no competitor scoring.
The English library is at xstrastar.com and the Chinese library at xingchuda.com. Both are open and require no registration.
About XstraStar
XstraStar is an AI marketing and GEO optimization company working with global technology and software companies on generative engine optimization and measurable organic growth. The company operates its own AI answer monitoring across English and Chinese engines. For more information, visit https://xstrastar.com/
