Knowledge Base Analytics: What Should Be Measured?
Imagine this scenario. You and your team put in the sweat, from research to implementation, and you have rolled out a shiny new knowledge base to your users. Better yet, it’s being used! Co-workers, customers, and other stakeholders have begun to regularly rely on the knowledge base, extolling its virtues and benefits to others while leveraging the information managed within it to their advantage and your organization’s overall success.
Nice work! Enjoy this moment. But know that your journey has just begun. Enter knowledge base analytics.
We often think we know exactly what our users need but may find ourselves surprised when we examine the resources they’re using. That’s why making sure that your stakeholders have access to the information that drives their success requires more than launching a knowledge base. It also requires a recognition that users’ needs will change over time and that a useful knowledge base must keep pace. Using knowledge base analytics effectively can help you understand whether your knowledge base is fulfilling your users’ needs as intended. — including regular monitoring and management of the articles in your knowledge base, studying current user behavior and anticipating changes, and routinely updating pages to ensure that users have access to current and accurate information.
Knowledge Base Analytics: Search
Statistics around user searches are a great place to start since many knowledge base interactions will leverage that mechanism. Here are some key metrics you should be tracking and considerations for each:
Total number of searches. Are users primarily relying on search or are they browsing and clicking on links to navigate through the knowledge base? Using this metric in conjunction with total site visits will help determine how heavily users rely on search and can contribute to a deeper understanding of usage trends.
Search terms. What are your users searching for? Maintaining an awareness of the top most used articles can help you prioritize articles for updating and revision. Plus, you can spot opportunities for adding new content.
Unsuccessful searches. Are your users able to find what they need? Keeping track of searches that return zero results can help determine information gaps in your knowledge base and could trigger the creation of a new article.
Top searches over time. Are there any trending searches? Stability in this area may indicate that more content on a topic should be added, or that a concerted effort should be made to keep that content current.
Dive into your content next by interpreting metrics about article or item views:
Total number of views. Is an article being found or used? Items with high view rates may be an indication of overall popularity or usefulness.
Duration of views. Are visitors reading the article? How long a visitor stays on a page can reveal its level of usefulness or visitor engagement.
Top articles over time. Which articles are stakeholders using most? High levels of traffic for articles illustrates the areas of your business or service that need to be explained the most.
View source. Where did the view come from? Whether users get there from a topic listing, search, internal or external website link, email link, etc., knowing the traffic sources for an item can help you make tactical decisions about how to promote the knowledge base.
Ratings, page-level comments, and other user-driven feedback. Does your knowledge base allow readers to rate pages? Do you ask if an article was helpful — and if not, what could be improved? Can knowledge base visitors post comments directly to articles? If you are collecting input from visitors via the knowledge base, make sure you’re reviewing this input and acting upon it.
In addition to quantitative analysis of your knowledge base, we recommend periodic check-ins with your users. Actual conversations may be the key to resolving the ambiguities that come with usage analytics and allow you to interpret the data more accurately. Using the “total number of views” metric as an example, does an article with very few or no views at all indicate an obsolete article that should be archived, a stale article that needs to be updated with current terminology, or is it just something everyone knows?
While determining a proper course of action may take more research and conversations with users, analytics help identify potential issues and formulate the questions you’ll need to ask.
In any case, don’t forget to review your data and draw insights from how people are truly interacting with your knowledge base.