Once upon a time, measuring customer service success was straightforward. Hire some agents, train them, and use standard self service KPIs to measure how well they perform.
Modern contact centers are made up of a complex combination of humans and virtual assistants – using both natural and artificial intelligence – operating over multiple channels and using a wide range of tools to solve customers’ issues. With many of these channels and tools enabling self-service, traditional KPIs are no longer sufficient.
The Challenge of Measuring Self Service
When a virtual agent fills a customer’s enquiry, collects all relevant details and passes it to a human agent for final approval, how should average handling time (AHT) be measured?
When a customer searches through FAQs, follows a video tutorial, asks a question on a user forum and still does not have a satisfactory answer to their question, resulting in a call to customer service, is that resolution considered a first contact resolution (FCR)?
When more enquiries are handled by AI-assisted agents – for example with best-next-step recommendations – are traditionally measured agent performance metrics still accurate?
As self-service becomes more widely adopted in the customer service arena, gone are the days of two-minute performance-boosting calls asking to update an address.
Calls asking for basic operating instructions are also a thing of the past. And long gone are those password reset requests. These simple enquiries can now be handled via self-service channels, leaving agents to field the more complex calls that are currently beyond the capabilities of machines.
Complex calls take longer, of course. They sometimes require research or approval from a supervisor. No matter how capable the agent, or how efficient the process, these complex cases will obviously have a negative effect on AHT, and sometimes even on FCR, if a follow-up conversation is required.
Customer care center metrics in the era of self-service clearly require a different approach. New, more relevant self service KPI metrics that measure the success of business self service analytics along with agent performance are now required.
5 Metrics to Measure The Impact of Self Service
There are a number of different categories that can be used to calculate self service scores.
Call Deflection Rate
Call deflection occurs when a customer’s enquiry is routed to an alternative service channel. The goal of call deflection is both to ensure that customers receive the answers they are seeking in the most efficient manner, and to reduce the number of calls routed to human agents. Instead, enquiries may be deflected using different support methods such as:
- live chat
- community forums
- knowledge center databases.
While this metric refers to “calls,” it also encompasses any contact requiring agent attention, including chat sessions and emails.
Measuring the call deflection rate is complicated. One method, according to DB Kay & Associates, is to estimate both the percentage of users who are successful with self-service and the percentage of users who would have contacted a live agent. The product of those two percentages represents the deflection rate.
Measuring call deflection with absolute precision is almost impossible as the self service metric essentially measures what didn’t happen. It is more useful to monitor the metric over time, as deflection should increase as the self-service channels improve and customers become more comfortable with the technology.
Total Cost Per Contact
To measure the total cost per contact, enterprises must go beyond the cost of each call, and measure the cost of each channel – including email, live chat and virtual agents – as well as the combined cost per contact. Enterprises must consider staffing costs, call center management systems expenses and the cost of self-service tools. Experts estimate that the current average total cost per contact across all industries is approximately $15.
In today’s omnichannel contact center, it is exceedingly difficult to measure each of these channels separately, as customers generally experience more than one channel during a support episode.
A customer may use a number of different channels during a call, for example:
- FAQs and live chat
- searching the knowledge bank while on hold for an agent
- sending an email after watching a video tutorial.
The cost of the initial contact would be the cost of the two channels combined, and the total cost per contact would be the total cost of all available channels.
This self service KPI metric provides enterprises with the actual cost of support when multiple channels are used to provide a customer resolution. It also indicates whether adding self-service options has decreased the overall cost of support.
Fulfillment speed represents the length of time it takes to conclude a self-service transaction from initial request to completion. The speed of fulfillment should be faster than the time it would take the customer to contact a human agent. If, after implementing a self-service channel, fulfillment times have increased or even stayed the same, the implementation cannot be considered successful.
For example, providing payment details via a chatbot will always be faster than having a customer call in to provide payment details to a live agent. Does the same hold true if a customer wants instructions on using an electronic device? If a customer wants to upgrade his level of service? For this self service KPI to be green, it should.
No matter how excited the business may be about implementing self-service channels, if customers are not satisfied by their usability or efficiency, then the self-service channel cannot be considered a success. Customer satisfaction must be tracked for each self-service channel – including surveys, direct feedback and Net Promoter Score (NPS) – to learn which channels are most successful as well as which ones need improvement. A temporary drop in satisfaction scores is to be expected when rolling out new channels or restricting access to familiar ones. However, if done correctly, customers will appreciate the ability to use self-service to get the answers they need when they need them.
Self Service Success Rate
The self service success rate can be determined by tracking how many customer enquiries are handled by self-service channels without being escalated to a human agent.
The success rate can be linked to a number of self service metrics, such as:
- The percentage of times a “how to order” FAQ leads to an order rather than a customer-initiated chat session.
- The percentage of times a video tutorial explaining how to activate a device eliminates a customer call to the contact center.
- The percentage of times a knowledge bank search leads to a useful article.
The self service KPI is typically measured by asking the user to rate the article or mark it as “useful” or “this solved my problem.”
Self Service KPIs: The New Way of Measuring Customer Success
Self service has come a long way. Businesses should therefore aim at evolving the way they assess self service performance on customer experience, to continually optimize for enhanced customer loyalty and maximum lifetime value.