Overview
SearchAssist captures and presents analytics of the Search Assistant utilization and performance in both the Search and Result phases. Analytics are the metrics related to the usage of the Search Assistant and are built into informative dashboards that you can readily deploy. These dashboards provide insights to measure the overall and particular performance of the Search Assistant.
Based on the insights drawn from the analytics, business users can take informed decisions to alter the index and search configurations and/or improve some sections of the content. Insights guide you to define experiments and determine the most suitable configurations to improve the quality of search experience and/or result relevance.
These metrics can be further categorized as follows:
- Analytics are included in:
- Dashboard giving an overview of the Search Assistant performance based on
- total users
- total searches
- result rate, clicks, and
- average click position Refer to Accessing Dashboard
- User Engagement feature that display the user engagement data such as
- User Trends
- User Volume, etc. Refer to Getting User Engagement Metrics.
- Dashboard giving an overview of the Search Assistant performance based on
- Insights giving details related to search queries and results, and providing the following pointers:
- Search Insights provide a quick overview of various metrics such as
- top search queries with results and
- queries without results Refer to Getting Insights from Search inputs
- Results Insights provide valuable information on metrics related to the results giving you an understanding of the user expectations such as
- Appearances
- Number of Clicks
- Click-through Rate, and
- Average Position. Refer to Getting Insights from Results
- Search Insights provide a quick overview of various metrics such as
- Experiments enabling you in assessing the impact of certain aspects of the index and search configurations. You can create experiments with multiple index and search configurations to evaluate which one is more effective. Refer to Experimenting Variants