Usage analytics analyze usage events, such as views from the event store. When a user completes an action, such as viewing a page, the event is collected and stored in usage files on each WFE server. This information is pushed to an Event Store where it is stored until it is processed by the Analytics Processing Component. The results are then returned to the Content Processing Component to be included in the
search index. The usage events that are analyzed include:
- Views.
- Recommendations displayed.
- Recommendations clicked.
The following table shows the sub-analyses that act on Usage Analysis results.
The Analytics Processing Component generates an Excel report that shows popularity trends and a listing of most popular items, by view.
Usage counts
The Usage counts analysis analyzes events, such as views. The analysis checks the overall number of times that an item is opened by any means, not just from a search result page. You can set the analysis retention value as any number between one and 14 days (14 days is the default value). The statistical data is aggregated at the following levels:
- Site level
- Site collection level
- Tenant level (SharePoint Online only)
Usage events are kept on disk for the retention value (the default value is 14 days) and are then deleted. Every day, the previous full day of usage count data is analyzed. The counts are added to the search index items to improve relevance. The analysis is also stored in the Analytics reporting database, where this information can be used to display popular site items.
Recommendations
The Recommendations analysis creates recommendations among items based on user activity patterns. To improve relevance, the analysis calculates inter-item relationships and adds the information to the search index items. You can use this information to provide recommendations based on usage by other users. This information is stored in the Analytics reporting database.
Activity ranking
Activity ranking tracks usage events to improve search relevancy. If an item has a high view usage, it will get a higher activity rank score. The analysis identifies trends, so that older items, that have just had more time to amass views, do not automatically get a higher ranking than newer items, that are getting a larger number of recent hits.