Knowledge analytics shows how employees and AI use your knowledge base, how well knowledge supports self-service, and where your content may need attention.
Use it to understand search demand, see how AI uses knowledge, identify knowledge gaps, and find content that employees find helpful or difficult to use.
Knowledge analytics focuses on knowledge performance and effectiveness. It is different from the Knowledge dashboard in Knowledge Management, which focuses on knowledge structure, ownership, and inventory.
Filter the data
Use the filters at the top of the dashboard to focus on the data you want to investigate.
You can filter by:
- Time period
- Role
- Channel
- Available custom user properties
Start with a broad time period to understand overall performance, then narrow the period or apply additional filters to investigate a particular audience or channel.
Searches
Shows the total number of times the knowledge base was searched during the selected period, including searches made directly by employees and searches triggered automatically by the AI assistant on their behalf.
Use Searches to understand how much demand there is for knowledge.
A high number of searches tells you that employees and AI are looking for information. It doesn't tell you whether they found a useful answer or whether the content resolved their question.
Cited by AI
Shows the number of times the AI assistant drew from a knowledge article to answer an employee's question.
A search doesn't always lead to a reference. Cited by AI shows how often the knowledge base is actively being used to answer questions, rather than simply being searched.
Use this alongside Searches to understand how knowledge is being used by AI.
Content deflection rate
Shows the percentage of employee queries resolved using only the knowledge base, without other tools or HR involvement.
This is a stricter measure than the self-service deflection rate shown on the Global tab because it specifically measures queries resolved using knowledge content alone.
Use Content deflection rate to understand how effectively your knowledge base supports self-service.
Article helpfulness
Shows the percentage of thumbs-up ratings submitted by employees on knowledge articles, out of all explicit ratings received.
Use Article helpfulness to understand how useful employees find the content itself. It measures feedback on knowledge articles, not whether employees found the AI assistant helpful overall.
Knowledge performance by category
Shows how each knowledge category is performing across views, AI references, deflections, and employee satisfaction.
Use this view to identify which areas of your knowledge base are working well and which may need attention.
For example, a category with high AI references but low deflection may be an area worth investigating. Review the related metrics rather than using one measure on its own.
Most searched topics
Shows the topics employees and AI are searching for most frequently during the selected period, together with how effectively those searches are leading to a resolution.
Use Most searched topics to understand where demand is concentrated.
A high search volume with a low deflection rate can be a signal that content needs improvement.
Knowledge gaps
Shows topics employees are actively searching for that the knowledge base is failing to resolve.
A knowledge gap may occur because content is missing, incomplete, or doesn't match how employees phrase their searches.
Use Knowledge gaps to identify high-priority opportunities to improve your knowledge base.
What employees are saying
Shows topics automatically derived from the feedback employees leave on knowledge articles.
The topics use sentiment analysis of written comments to show how positively each theme was rated and how many article views it relates to.
Use What employees are saying to understand what employees are thinking about your knowledge content and identify themes in their feedback.
Article feedback
Review the helpfulness, feedback, and views associated with individual knowledge topics.
Use this information alongside What employees are saying to identify content that employees find useful and content that may need to be reviewed or rewritten.
How employees use knowledge
Shows how employees move through each stage of engaging with a knowledge article, from the initial search through to reading the article and leaving feedback.
Each stage shows the percentage of employees who searched who made it that far, so you can see where engagement drops off.
Use this view to identify where employees stop engaging with knowledge content.
For example, a large drop between searches and article views may indicate an issue with search results or relevance. A drop later in the funnel may indicate that employees are finding the content but aren't continuing to read or provide feedback.
Use analytics to improve your knowledge base
Use the dashboard to identify where to focus your content work.
- High searches + high deflection: Employees are looking for the information and the knowledge base is resolving their queries effectively.
- High searches + low deflection: Employees need the information, but the knowledge base may not be resolving their queries. Investigate the topic for potential content gaps.
- High AI citations + low deflection: AI is using the knowledge content, but the content may not be sufficient to resolve the employee's question.
- Low helpfulness: Review the article and employee feedback to understand what may need to change.
- Knowledge gaps: Prioritize topics where employees are searching, but the knowledge base isn't resolving their queries.
- Engagement drop-off: Investigate where employees stop interacting with knowledge content.
Don't use a single metric to determine whether content is performing well. Compare search demand, AI usage, deflection, feedback, and engagement to understand the bigger picture.