When AI helps people explore products, human memory and an assistant's shortlist may differ. Kantar's 2026 outlook asks brand builders to consider both routes to choice.
The demo compares unaided human recall with an illustrative AI ranking for the same question. Real analysis should measure people and collect answers for defined prompts and models separately.
AI rankings vary by model, prompt and time. One answer is not a stable measure of brand preference.
When to use
Use when brand tracking looks healthy but the brand is absent from AI shortlists.