Bloomreach’s Facet Ranking solution aims to provide the most relevant facet results for end-users in their category browsing and product search experience. Facet ranking takes 3 signals into account:
In general, facets with higher product counts are ranked higher.
Relevance (Dominant Category)
The Bloomreach algorithm takes the top products from the search recall set and analyzes the dominant categories these products belong to. Any facets (attributes) that map to these dominant categories will be ranked higher. This approach largely reduces the noise within facet results because the facets generated by the long tail product result will be ranked lower or potentially hidden.
Facets that have a higher click-through rate (interaction rate) are ranked higher. In the long term, this utilizes the collective wisdom of the crowd to make sure that popular and favored facets are ranked higher.