Improve your self-service with the right metrics
Once you’ve taken the self-service plunge and started rolling it out across your business, the logical next question is: What happens now that I’ve jumped in?
Take this quiz to identify the metrics that best capture the impact of your self-service efforts. Think of these metrics as ways to identify opportunities for tweaking, testing, and learning as you implement self-service across your enterprise.
Improve your self-service efforts with metrics
What type of business do you have and how would you implement self-service?
For business-to-business companies, other companies are your customers. Software-as-a-solution (SaaS) is a B2B product, for example. Self-service helps these customers find useful information about these products or services, or allows them to troubleshoot on their own.
Business-to-consumer companies use self-service to provide on-the-spot answers for customers and agents about all sorts of inquiries, such as returns and exchanges, shipping, and refunds.
Business-to-enterprise means your employees are your self-service customers. Teams such as HR create self-service content to answer common employee questions, like payroll queries or help with passwords.
Improve your self-service efforts with metrics
What are your priorities?
That is: In which aspect of the business would you like to see the most improvement as a result of your self-service efforts?
Building community and relationships
Engagement with your self-service content, such as upvotes, downvotes, and comments within your community platforms, are pulse points about your customer base. Levels of self-service engagement tell you whether your knowledge base content is useful—or if it could use some work.
For a growing company, one-to-one support simply doesn’t scale. Self-service support allows customers to self-assist on simple questions, and frees your agents for complex issues that do require a personal touch.
Streamlining internal operations
A knowledge base can make agents more efficient, providing at-their-fingertips information as they assist your growing customer base. Furthermore, this info can be used by everyone on your support team, as well as by multiple functions across your business.
Here are your results!
Scroll down to see the four metrics that matter most at this point in your business.
Ticket deflection is tricky. How do you measure something that hasn’t happened? To measure it, divide the total number of unique users that interacted with help content on a topic by the total number of unique users who opened tickets on that same topic. If you have a popular help center page and notice a lower number of tickets on that same subject, congratulations: You’re likely deflecting tickets!
This is the number of views of each page in your help center. Aim to keep this number high and your bounce rate low, which indicates people are finding and spending time with your content.
Tickets created after search
This metric helps you track what customers search for and what actions they take after conducting those searches. A large number of tickets created after search suggests that the content provided wasn’t enough to help the customer solve the problem on their own. This could also indicate a product issue as customers might be having difficulty even with sufficient self-service content.
This is the amount of time it takes for agents to resolve issues. Often times, the better your knowledge base, the faster your agents can resolve tickets, because existing content allows them to more quickly and efficiently answer frequently asked questions. Resolution time can be broken down into time to resolution, which measures the time it takes for a support issue to be solved, and first-contact resolution (FCR), which measures the percentage of support issues that were resolved in a single interaction.
Handle time is the time that an agent spends working on a single support interaction. Digging into the details about the amount of effort that went into a resolution is a whole other level of data, a step beyond measuring the amount of time it took.
Abandoned shopping carts
This is one of the more painful metrics. An abandoned shopping cart means a customer browsed your site, placed items in the shopping cart, but exited without completing the purchase. Often times, customers do this because they can’t find what they need, such as information about shipping or returns—which should be made available on the spot via self-service.
Unique users refers to the number of people who clicked on your self-service articles. If your number of unique visitors is low or flat, it could mean you aren’t doing enough to drive traffic to your knowledge base. It’s also possible that you aren’t providing the right content. Encourage advocates to proactively share knowledge base content as they manage tickets, so that customers know they can reference this resource going forward.
Inbound ticket volume
This is the number of tickets created by customers. For many, inbound ticket volume is the gold standard for determining self-service health. A high volume of inbound tickets can lead to strained support staff and long wait times for customers, which means nobody wins. Find out which simple issues can be solved with self-service, then make it easy for customers to find that information. Inbound ticket volume should get smaller as a result.
One-touch tickets are the number of tickets solved via a single interaction. While a high number of one-touch tickets could be an indicator of efficient support agents, it might also mean that too many customers are submitting tickets that could easily have been resolved without direct help from an agent—or with easy access to a help center article. Keep track of one-touch tickets, specifically tracking the kinds of issues they revolve around, then create knowledge base articles that directly address those topics.
Ticket escalation rate
Not all tickets can be resolved by the first agent who answered the call or opened the support request. Tickets requiring deep knowledge are often escalated to agents with more specialization (which is fine; that’s what they’re there for), but too many ticket escalations can indicate either an ongoing, larger issue, or maybe front line agents aren’t being properly trained, or don’t have adequate (or any) help center content to reference.
Searches with no clicks or results
“Searches with no results” refers to searches for which no relevant help content was provided. “Searches with no clicks” refers to the searches that didn’t result in any clicks within the help center, indicating the content provided wasn’t useful to the customer. Both of these metrics indicate that your help content needs to be revised, either with more relevant article titles or with new articles that leverage search terms customers are using.
This is the percentage of single-page sessions within your knowledge base. A high bounce rate may indicate that your content didn’t answer their question, which means customers may have sought answers elsewhere or switched channels.
These are tickets that customers either didn’t finish or didn’t submit to an agent. This should generally be a low number; a higher one could indicate a poor user experience, where the customer finds it difficult to either describe their problem or submit a ticket.
Tickets created from comments
These are tickets created from comments made on your help center articles. Customers add comments when they need more information or clarification. A high number of tickets submitted this way may indicate that the article isn’t sufficient, but it’s actually a great heads up to enrich your knowledge base with more detail and unexpected use cases.
Though channel switching isn’t necessarily a bad thing when you’ve built a truly omnichannel experience, it could be noteworthy from a self-service perspective. If customers are finding but leaving your help center and seeking 1:1 assistance, something could be amiss in your self-service channel.