Skip to main content

Search Profiles

Search Profiles are configuration objects that control how search queries are executed. They provide a powerful way to customize search relevance, filtering, and ranking without requiring code changes, enabling merchandisers and business users to optimize the search experience.

How Search Profiles Work

A Search Profile defines:

  • Which fields to search: Select which product fields are included in search queries
  • Field weights: Assign relative importance to each field for relevance ranking
  • Default filters: Apply business rules that filter results for all searches
  • Boost rules: Promote or demote products based on specific conditions
  • Match type: Control how scores from multiple fields are combined

When a search request is made, the system uses the specified profile (or the default profile) to configure the search behavior.

Search Profile Components

ComponentDescription
slugUnique identifier used to reference the profile in search requests
fieldsList of fields to search with optional weights for relevance ranking
filtersDefault filter expressions applied to all searches using this profile
boostsRules for promoting or demoting products based on conditions
text_match_typeHow to combine scores when multiple fields match (max_score, max_weight, sum_score)

Field Weights

Field weights control how important each field is when ranking search results. Higher weights mean matches in that field contribute more to the relevance score.

Example: A profile with name (weight 10) and description (weight 5) will rank products with search term matches in the name higher than those with matches only in the description.

Filters

Profile filters enforce business rules that apply to all searches:

  • Ensure only live/published products appear in results
  • Restrict results to specific categories or channels
  • Hide products based on visibility settings

Multiple filters are combined with OR logic, then ANDed with any filters in the search request.

Boost Rules

Boost rules allow you to promote or demote products in search results:

  • Positive weights: Push matching products to the top of results
  • Negative weights: Push matching products to the bottom (burying)
  • Zero weight: No effect on ranking

Rules are evaluated in weight order, and the first matching rule determines the product's boost.

Text Match Types

TypeDescription
max_scoreUses the highest score from any single matching field. Prioritizes match quality.
max_weightUses the score from the highest-weighted matching field. Prioritizes field importance.
sum_scoreSums weighted scores from all matching fields. Rewards multi-field matches.

Default Profile

Each store has a default search profile that is used when no profile is specified in the search request. You can set any profile as the default using the set default endpoint.

Use Cases

  • A/B testing: Create multiple profiles with different configurations to test search relevance
  • Channel optimization: Different profiles for web, mobile, and B2B channels
  • Seasonal merchandising: Profiles that boost seasonal or promotional products
  • Customer segmentation: Different ranking strategies for different customer tiers
Ask External AI