GTM AI Recommendation Index

Field notes

The index turns model recommendations into category-level standings by asking six systems repeated buying questions and aggregating first-choice shares. Its visible dataset contains 438 prompts, 2,628 answers, and 17,395 judge labels across 3,019 products and 2,580 vendors. A category is considered contested when its leading product reaches 40% share or less. The page also flags that its calibration repeat has not yet run, so it does not present changes as trends.