It’s 2024, and the tide has finally turned on AI.
61% of customers would actively prefer the prompt response of an AI solution over waiting for assistance from a human customer support representative. But there’s one prerequisite to every AI solution that works:
Great knowledge management.
Internal and external knowledge management
Almost every customer support team has two fundamental methods of knowledge management:
Internal knowledge management includes the processes and systems used to create, organize, and share knowledge to support internal operations and service delivery. That might mean documenting processes, troubleshooting guides, and best practices your agents can refer to when replying to customers. This knowledge often lives in a knowledge base but could also live in your ticketing system or other collaboration tools. The audience you’re talking to is your team.
External knowledge management is the process of creating help materials to enhance customer experience. Your customers are the audience here, and the goal is to empower them to find solutions independently without contacting customer support. It includes the processes you use to create and update content in your help center and how your customers see and engage with it.
Unsurprisingly, the metrics you’ll need to measure depend on what you want your knowledge management team to focus on.
There is a lot of crossover between internal and external knowledge management, so you can adapt most of these metrics to work for both.
You could target:
Customer engagement and satisfaction.
How effective your content is.
Search metrics.
Internal contributions and usage.
The best knowledge management teams consistently monitor a range of these while focusing explicitly on one or two areas at a time.
Customer engagement and satisfaction metrics to measure knowledge management
91% of customers say they would use a knowledge base if it met their needs.
There are two pieces to that puzzle:
Customers need a comprehensive knowledge base containing the information they want.
It should “meet their needs.”
The best way to measure that is with customer engagement and satisfaction metrics:
Article views. What are the total monthly article views on your knowledge base? Compare that to the size of your active user base and the number of tickets you get to get a sense of whether article views are high or low.
Interactions. Another way to measure engagement is to look only at interactions, like searching, clicking, or rating articles. Since a help center is typically there to provide a solution, a low interaction rate doesn’t indicate much. It could be that your agents or customers find the information they’re looking for and immediately close the page.
Helpfulness score. Almost all knowledge base software has a built-in rating feature that lets your customers tell you how helpful your articles are. The proportion of people who rate an article is typically not high, but you’ll find that your most important articles garner enough scores to show how they’re performing.
Customer Effort Score (CES). This one’s only for external help centers. An effort score asks your customers to rate how much effort they invested to find a solution. CES is one of the best qualitative metrics for self-service because it gets your customers’ perceptions. Some companies use Net Promoter Score (NPS) for this instead.
These are often the foundational metrics of your knowledge base. Since many of the KPIs below influence how these perform, most practical tips to improve these metrics come down to investigation. You need to identify the specific cause of the issue.
For example, if views are low, look into how customers get to your help center. Or, if your CES is too high, look into average pages per session or the tickets created after someone has viewed an article.
Metrics to measure how effective your help content is
The next question to answer is: Does your help center actually solve the questions your customers (or agents) have?
Contact rate. How often do your customers contact you? A low or decreasing contact rate is often the result of great self-service or proactive support.
Self-service score. This is the ratio of user sessions to the number of users creating tickets. A score of 20 means your support center gets 20 times as many sessions as you have users contacting you.
Helpfulness score. This metric, native to Help Center Analytics, uses the article votes in Zendesk Guide to calculate the helpfulness (the percentage of positive votes out of total votes) and allows you to filter it by any period of time and any article, section, or category.
Ticket deflection rate. It can be hard to measure if your help center isn’t well-integrated into your contact form (and any other contact points). Ticket deflection is the percentage of tickets successfully resolved through self-service options.
Cost-effectiveness. You can translate ticket deflection or contact rate into cost-effectiveness by calculating your cost per ticket and figuring out how many tickets you’re saving with your help center.
Time to competency measures the duration it takes for new support agents to handle support tickets proficiently. TTC is only relevant for an internal knowledge base. An external counterpart might be something like time to value if help articles are a major component of your customer onboarding.
There’s one important caveat here.
You can optimize for each of these to the detriment of your customer experience. For example, you could have an extremely high ticket deflection rate and save a lot of money by simply making it harder for your customers to contact you–which defeats the purpose of your knowledge base.
That would often coincide with high CES and low helpfulness scores. That’s why combining these metrics with engagement and satisfaction is essential to get a full picture.
You can improve these metrics by:
Collecting feedback in articles to understand where people get stuck or confused.
Evaluating the top ticket drivers in your product and tackling these in a systematic way to reduce contacts.
Finding smart ways to integrate your help articles where customers experience issues or have questions.
Making it easy for your agents to flag missing information (internally and externally).
💡 Pro Tip: Don't just wing it when it comes to measuring your help center's success. Analytics tools are your friends! Use a general tool like Google Analytics or a specialized help center analytics solution — there are upsides and downsides to both, which you can explore in this blog post. Whichever you choose, you'll gain the insights to make data-driven improvements.
Search metrics to improve knowledge base performance
40% of customers contact a call center after they've tried to self-serve. Self-service solutions often fall short–largely because customers can’t find the information they want.
That’s where search metrics come in.
Click-through rate (CTR). CTR measures the percentage of people who click through to an article after searching. Sometimes you can improve a low CTR by playing around with the titles and making them more specific.
Number of searches with no results. There are two potential explanations for many searches with no results. Either your help center isn’t as comprehensive as it needs to be, your customers (or agents) are using different terminology in their searches, and you need to adapt your content.
Average pages per session. A high number of pages per session suggests that the person had to search extensively for the correct information. Looking at user flows and your navigation structure can be a starting point for improving these.
Average sessions per user. If many of your users are returning users, that suggests people are finding value in your help content over time.
Search analytics are your best friend. They can make a massive difference in how easy it is for your customers to find information. It’s also a good idea to ensure your help center is optimized for SEO, in case most of those searches happen outside your help center.
Increasing internal contributions and usage of your knowledge base
The final set of metrics you can measure for your knowledge management team are based on internal usage.
Process adherence. This measures the extent to which your agents follow predefined workflows or processes, especially when creating and updating help articles.
Rate of new articles. What’s the proportion of new articles published per month? A goal around this is great if you need a larger body of help center content and are working to build that.
Number of article updates. This one can also be measured on a monthly or quarterly basis. How often do you review and update articles to reflect changes in your product?
Linked articles per ticket. Measuring the average number of linked articles per ticket can indicate how well your support team knows your articles. It also ensures that you’re slowly educating your customers about the presence of your knowledge base.
Average age of an article. You can calculate this based on when articles were last updated. Outdated help content erodes trust over time, so it’s important to review it regularly and ensure it’s still relevant.
Internal contributions are paramount if you want to implement Knowledge-Centered Service. But they’re also good to measure if you have a wide range of knowledge and experience across your team and need to promote better knowledge-sharing.
They’ll often have knock-on effects on the effectiveness and customer engagement metrics. Ensuring that your support team is fully involved in reviewing and updating your help articles will dramatically improve their quality.
These are some of the most straightforward metrics to improve consistently and consciously.
Conduct a help center audit regularly.
Implement a system that automatically flags articles that haven’t been updated in a year.
Include linking a help article in your quality guidelines for ticket handling.
Develop an internal review process to ensure your agents get regular feedback on their process adherence.
Level up your knowledge management
Great knowledge management can transform your customer experience. When your internal knowledge base is on point, your agents will provide better answers faster. And an impressive external knowledge base will lead to happier customers while saving your business money.
Metrics can provide all the direction you need to make that happen. While this is a lengthy list, you only need to regularly work with 3-4 metrics to see results.
At Swifteq, we’ve developed various tools to manage their help center. If you’re looking for ways to take your knowledge base to the next level, start a 14-day free trial today.
Written by Nouran Smogluk Nouran is a passionate people manager who believes that work should be a place where people grow, develop, and thrive. She writes for Supported Content and also blogs about a variety of topics, including remote work, leadership, and creating great customer experiences. |