Skip to main content
Search personalization uses AI technology to re-rank search results at query time based on the user profile you provide. It works alongside full-text search and hybrid search to deliver results tailored to each user.

Why use search personalization?

Not everyone searches the same way. Personalizing search results allows you to adapt relevance to each user’s preferences, behavior, or intent. For example, in an e-commerce site, someone who often shops for sportswear might see sneakers and activewear ranked higher when searching for “shoes”. A user interested in luxury fashion might see designer heels or leather boots first instead.

How does search personalization work?

  1. First generate a user profile: "The user prefers genres like Documentary, Music, Drama"
  2. When the user performs a search, you submit their profile together with their search request
  3. Meilisearch retrieves documents based on the user’s query as usual
  4. Finally, the re-ranking model reorders results based on the user profile you provided in the first step

Use cases

  • E-commerce: Surface products aligned with a shopper’s purchase history, brand preferences, or browsing behavior. A customer who frequently buys running gear sees running shoes before formal shoes when searching for “shoes”.
  • Content platforms: Rank articles, videos, or podcasts based on the topics a user engages with most. A reader interested in machine learning sees ML-related content higher in results for broad queries like “tutorial”. Combine with analytics to measure impact.
  • Marketplace search: Tailor listings to a buyer’s location, budget range, or past interactions so the most relevant offers appear first.

Next steps

Getting started

Enable personalization and send your first personalized search

Generate user context

Build user profiles from behavior data

Personalize e-commerce search

Step-by-step guide for personalizing product search results