# Search Institute: Full Content Independent research on how search and AI systems find, rank, retrieve, and recommend information. ## What this site is Search Institute is an independent research publication about search engines, information retrieval, large language models, recommendation systems, and the evidence those systems use. Sources: ## Methodology summary Reports prioritize primary evidence, preserve experimental boundaries, separate recall from retrieval and recommendation, disclose uncertainty, and link consequential factual claims to inspectable sources. Sources: ## Authorship Named writers receive named bylines. Institutional work uses the Search Institute Team collective byline. Andrew Ansley has a dedicated author profile and article archive. Sources: , --- ## Search Institute Canonical URL: https://searchinstitute.co/ Independent research on how search and AI systems find, rank, retrieve, and recommend information. --- ## About Canonical URL: https://searchinstitute.co/about What Search Institute studies, how it publishes, and how authorship works. --- ## Andrew Ansley Canonical URL: https://searchinstitute.co/authors/andrew-ansley Author profile and archive for Andrew Ansley. --- ## Contact Canonical URL: https://searchinstitute.co/contact Editorial, correction, source-material, and general contact route. --- ## Funding Canonical URL: https://searchinstitute.co/funding Current self-funding status and editorial-independence policy. --- ## Methodology Canonical URL: https://searchinstitute.co/methodology The standards governing source selection, experimental boundaries, uncertainty, and review. --- ## Mission & Standards Canonical URL: https://searchinstitute.co/mission The Institute's mission and six publication standards. --- ## Research Canonical URL: https://searchinstitute.co/research The complete archive of Search Institute reports. --- ## Do AI Models Favor Brands They Already Know? Reading the Brand-Familiarity Experiments Canonical URL: https://searchinstitute.co/research/ai-models-brand-familiarity-experiments Published: 2026-08-21T00:00:00.000Z Authors: Andrew Ansley Controlled and observational evidence suggests familiar brands receive a real but conditional advantage that better product evidence, retrieval, context, and wording can overturn. Familiarity behaves like a prior in AI recommendations: it shapes consideration sets and breaks ties, but competes with product specifications, retrieval architecture, prompt wording, and context. Product parameters explained 82.4 percent of ranking variance in the broadest controlled experiment reviewed; brand identity explained 1.2 percent. --- ## AI Recommendations Canonical URL: https://searchinstitute.co/topics/ai-recommendations How AI systems choose, rank, and explain products, brands, and sources. --- ## Brand Visibility Canonical URL: https://searchinstitute.co/topics/brand-visibility What makes organizations legible, retrievable, cited, and recommended across search and AI. --- ## Information Retrieval Canonical URL: https://searchinstitute.co/topics/information-retrieval The retrieval, ranking, and evidence-selection layers behind generated answers. --- ## Search Systems Canonical URL: https://searchinstitute.co/topics/search-systems How search engines and emerging answer systems understand and surface information.