GeoSurge measured how AI models decide what to search for. Brands a model already remembered were searched about three times as often as those it did not. For its top five remembered brands, the rate reached 67%. The pattern held across all nine industries tested. Models split user queries into many parallel searches. About 69% of those searches were generic category queries. Of the 31% that named a brand, nearly two-thirds named a top-five remembered brand. Memory-led categories like automotive and finance showed higher brand-specific search rates. Search-led categories like fitness and wellness showed lower rates. GeoSurge recommends diagnosing category type before spending. It also advises funding earned media to build model memory. It suggests measuring recall, not just citations. It notes that training runs, not query-time tweaks, decide most outcomes.
What it unlocks
Diagnosing whether a category is memory-led or search-led before allocating budget. Funding earned media as infrastructure to shape model memory. Measuring brand recall in models rather than just citations in answers.
- 55.7% vs 17.4% search rate for remembered vs forgotten brands
- 67% search rate for top-5 remembered brands
- 69% of searches were generic category queries
- 31% of searches named a brand
Sources