Metrisque today launched a measurement instrument that tells companies where AI models place their brand, products, and services, and gives the same answer every time it is queried in an LLM. In a pre-registered study, 92% of AI product recommendations landed where the instrument predicted before models were asked.
Ask an AI assistant the same shopping question twice and it usually names different products making AI visibility almost impossible to measure. LLMs change the reading on every run, so a company can never tell whether something they changed actually worked, or whether the model simply answered differently that day. Metrisque measures a different thing. Instead of counting how often a brand appears in AI answers, it measures how closely a company's own words match what buyers are asking. The reading is consistent, so a change can be made, measured, and proven.
"Every AI already has a belief about your company and has filed your products on a shelf before anyone asks it anything, and it has always been invisible," said Lian Pham, Co-founder, Metrisque.
Proving it out: In a pre-registered study published with a permanent citation, roughly eleven hundred real recommendations from two leading AI models were tested against predictions made before the models were asked anything. Ninety-two percent landed where the prediction said they would. "We also published the prediction we got wrong. A company that only shows you its wins hasn't shown you anything," said Pham.
Metrisque reads six things a company cannot see any other way. Brand Recall shows what AI models already believe about a company with nothing open in front of them, whether they know it exists, and whether what they recall is true. Category Fit shows how AI models categorize a company and its individual products. One beauty brand's makeup set was filed as a bug collecting kit because words on page "gotta catch them all" sounded like catching creatures. Buyer Match shows whether company's words match questions buyers ask AI assistants across twenty ways to ask same thing.
AI Recommendations shows who gets named when buyer asks, and where company sits against competitors including whether company is in consideration list at all. Competitor's Citations shows which websites each AI model actually reads when answering buyer's question - of 46 sources cited on one question only one was cited by all three, 39 cited by single model. Question Finder shows which buyer question is worth fighting for whether open, contested, or already owned.
"Every channel that mattered got placed and had a measurement layer before it got a budget. Search and social also had one. AI answers didn't have one given the speed at which things moved and the money is already moving," said Pham.
About Metrisque
Metrisque launched the first way to measure AI visibility that gives the same answer every time. Its instrument measures Brand Recall, Category Fit, Buyer Match, AI Recommendations, Competitor Citations, and Question Finder to tell companies where AI models place their brand, products, and services.
