Listen: Associative Embeddedness
A measure of how deeply a brand is woven into a model's memory, from how often and how early the model recalls it across many runs.
Transcript
How deeply is your brand woven into the memory of an artificial intelligence? It is not just a matter of whether a language model knows you exist, but how central your brand is to its entire web of associations.
To measure this, researchers have developed a metric called associative embeddedness. It is the powerhouse behind the AI Brand Authority Index, which recently ranked nearly three million brands based on their standing within Google's Gemini model.
The metric is built by analyzing recall behavior. After running hundreds of thousands of tests asking the model to name brands at random, researchers tracked two things: how often a brand was mentioned, and how early it appeared in the list. By combining these factors, they created a weight for each brand. A brand that is always recalled first scores near a perfect one, while a brand mentioned only once, near the bottom of a list, scores close to zero.
This data is then run through a specialized ranking algorithm to map out a graph of directed associations. The final result quantifies brand authority in a way traditional rankings simply cannot. Alongside frequency and share of voice, associative embeddedness is becoming a vital new measure of how visible a brand truly is in the age of AI.