Understand search ranking
You can use the search ranking feature to influence the relevancy score of content items based on specific attributes. Ranking enables you to fine-tune search results and boost certain documents, ensuring that the most relevant and valuable content surfaces for site visitors.
For example, imagine a visitor is searching for articles about artificial intelligence. They receive a list of 100 results that match their query based on textual relevance. Within this list, you want articles with higher review ratings to appear higher in the search results. You use search ranking to create this functionality. It works by elevating certain documents, in this case, the highest-rated articles about artificial intelligence, to a higher position within the search results.
Ranking is a mechanism for boosting the relevance and visibility of specific documents without changing the order of the entire result set. It is different from sorting, and it's important to understand this distinction. When you configure an attribute for search ranking, it doesn't mean the search results are ordered by that attribute. In the preceding example, the articles about artificial intelligence with the highest review ratings do not necessarily appear at the very top of the search results, and they are not ordered by descending review ratings. Instead, ranking contributes to the relevancy score of a document, which subsequently influences its position in the search results.
To make the most of search ranking, it's important to follow these best practices:
-
Use only the attributes that are truly essential for ranking. If you find yourself with a long list of attributes for ranking, it might mean that your approach needs refinement. Generally, 0-5 attributes are probably enough, but this number can vary based on your specific requirements.
-
When you assign weights to attributes, we recommend that you set weights for all ranking attributes to 1. If, after this initial configuration, you find that further fine-tuning is necessary, increase or decrease weights by increments of 0.1. Small weight increments provide a more balanced experience by ensuring that weights for search ranking don't drown out other factors used to calculate relevancy scores.