According to a Harvard Business Review research, around 81% of customers across industries try to solve issues on their own before reaching out to customer support. This implies that self-serve should be a core-pillar of your customer service strategy, making your knowledge base a top priority for customer success!
However, is your knowledge base empowering your customers to find answers and solve problems independently or are frustrated customers reaching out to your support team for rather benign issues?
Currently, most teams use Google Analytics to track their knowledge base performance but GA is not a good-fit for detailed knowledge base performance tracking. It struggles to measure article effectiveness, identify knowledge gaps, and lacks seamless integration with knowledge base platforms. This results in limited insights into user behavior, search trends, and overall article impact. DevRev’s article analytics capability helps in solving this issue.
While it can be difficult to gauge customer satisfaction with your knowledge base, tracking specific metrics can provide valuable insights into its effectiveness.
Customer-friendly metrics to track knowledge base success
Here are six metrics you can track to understand how useful customers find your knowledge base:
1. Search popularity - what are customers looking for?
This metric shows you the most common search terms people enter in your knowledge base, providing you with insights on what your customer is thinking. Combining this metric with click-through rates provides actionable insights. For example, if a particular term shows up a lot but has a low click-through rate, it might mean you need a new article or have to improve the existing one to address that topic.
How do you calculate?
Sort “search strings” in descending order of count. With DevRev you can also auto-cluster relevant search strings together!
Pro tip: Enabling GenAI search will make your knowledge base more accessible to your customers.
2. Search gaps - are searches leaving customers empty-handed?
This metric identifies searches that don't return any results in your knowledge base. Imagine a customer searching for answers and finding nothing. A high number of these ‘dead-end’ searches means there are gaps in your knowledge base. You'll want to create content to fill those gaps and make sure people find what they're looking for. Plugging these search gaps is likely the lowest-hanging fruit in terms of improving consumer self-serve!
How do you calculate?
It is calculated by displaying all the unique search terms which yielded zero results. You can prioritize by search volume, trends or customer type.
3. Article views- what articles are customers reading?
This metric tracks how often each article is viewed, showing you which ones are most relevant and useful to your customers. High view counts mean certain articles are particularly engaging and effective. By looking at these metrics, you can see which topics are hitting the mark and which articles are performing well. Plus, you can use the most popular articles as templates for future content.
How do you calculate?
This metric is just the count of views for an article.With DevRev, you can further analyze this across splits like unique views, channel, surfaces, signed-in users etc.
4. Reading time - how long are customers engaged with articles?
This metric shows how much time, on average, people spend reading a particular article or FAQs. This metric can be a good gauge of the reader’s attention and interest in your content. If people are spending a good amount of time on an article, it likely means they're finding the content valuable and informative. Low reading times might suggest the content is irrelevant or hard to understand. However, this metric can cut both ways. Unexpectedly long reading times might mean that the user is confused and is probably having to re-read the article/parts of it.
How do you calculate?
This is calculated at article level by dividing total session time of an article by the number of unique article views.
5. Article ratings - do customers find articles helpful?
Imagine a thumbs-up or thumbs-down button for your articles. This metric would show how many people find each article helpful. Articles with a high view count but low ‘helpful’ ratings might need some revision or improvement. These articles usually have some details missing or are outdated. On the other hand, articles with high view counts and great ratings can serve as a template on how to write customer friendly articles. This metric should be looked at in absolute terms, not % split of helpful vs not helpful since users are more inclined to hit not helpful than hit helpful.
How do you calculate?
Calculated at article level, the two output metrics here are sum of likes (thumbs up) and sum of dislikes (thumbs down).
6. Ticket deflection - are customers avoiding support tickets thanks to the knowledge base?
This metric shows how effectively your knowledge base is reducing the number of support tickets created.The idea here is that if the knowledge base is effective, there will be a decreased reliance on tickets since customers are able to solve their issues on their own. A higher ‘ticket deflection’ rate means customers are finding solutions within the knowledge base and are less reliant on contacting support. Hence, this metric can be thought of as a good indicator of the pressure you take off your support team’s shoulder, freeing them up for more complex and nuanced issues.
How do you calculate?
This metric is basically the ratio of the sum of all unique knowledge base sessions where a ticket wasn’t created divided by all unique session for a particular time period, it can be calculated as
Note: this is not an absolute conclusion of KB effectiveness, since there can be cases where the unique users are increasing and queries are not getting addressed but for some reason users are not creating tickets.
Taking action on your knowledge base insights
By tracking these metrics, you can see how users interact with your knowledge base and how well it meets their needs. It provides comprehensive view of what you are doing well and where you can improve.Using these metrics ensure that you:
- Find and fill any gaps in your knowledge base.
- Improve the searchability of your knowledge base by using the right keywords and content structure.
- Revise and update existing articles based on what customers are telling you through their reading habits and feedback.
- Create new content to address frequently searched topics or areas where people seem to be struggling.
Conclusion
Tracking these metrics helps you measure the effectiveness of your knowledge base and ensures that it helps in providing your customers a great experience. By continuously monitoring and refining your knowledge base, you empower customers to solve problems independently and allow your support team to concentrate on more complex issues.
A well performing knowledge base therefore helps in reducing costs, improving customer satisfaction and scaling support by increasing agent productivity.
Book a demo, to see how DevRev can help you deploy a world class knowledge base.
