Published: November 05, 2012 | Comments (2)
Call centers are awash in data that can be used to measure call center agents’ performance. We know how long an average call takes, how quickly the phone gets answered, and the percentage of time agents are logged in and ready to take calls.
A lot of this data is used to set individual performance targets. Unfortunately, some of these metrics can actually hurt customer service if they divert agents’ attention away from the desired performance. Three examples of call center metrics that can hurt service are average handle time, call monitoring scores, and average time to close support tickets.
Average Handle Time (AHT)
In theory, shorter calls allow each agent to handle more calls per hour which leads to reduced staffing costs. As a result, many call centers expect their agents to adhere to an AHT standard. Unfortunately, focusing on AHT can cause agents to rush through phone calls without solving the customer’s problem or they can be too eager to transfer calls to another department.
A 2008 study by the SQM Group found that the average call center solves just 68 percent of customers’ problems on the first call. This means that 32 percent of customers must call more than once to resolve their problem. Multiple calls per problem can increase call volume, require additional staffing, and ultimately reduce customer satisfaction.
Solution: Eliminate AHT as a performance standard for individual agents and use it only as a scheduling and forecasting tool. In its place, emphasize first call resolution (FCR) and call control. Agents should be given training, coaching, and resources to efficiently move each call to a swift resolution, which will increase customer satisfaction and decrease your call center’s overall call volume.
Call Monitoring Scores
Call center agents are often evaluated based on how well they adhere to a prescribed set of standards on each call. These standards typically include a scripted greeting, a standardized call flow, and required upsell offers.
Unfortunately, agents must often choose between adhering to these standards or providing great service. In some cases, the call quality standards are so rigid that an agent can’t adapt their approach to meet their customers’ needs. In other cases, the required scripting sounds unnatural and impersonal. I’ve personally listened to many calls where the agent nailed 100 percent of the standards while sounding like a bored robot.
Solution: There are two steps to improving call quality. First, implement guidelines that focus on broad actions rather than scripted responses. For example, you might ask reps to develop rapport with their customer, identify the customer’s needs, provide an acceptable solution, and express appreciation or empathy as appropriate. This gives agents the ability to be flexible while still focusing on the desired results.
Step two is holding regular team meetings where agents listen to recorded calls, evaluate the call based on the broad guidelines, and discuss their responses as a group. This process, known as calibration, becomes a simple way to share best practices and establish mutually agreed-upon expectations for customer service. It’s also an opportunity to discuss more subjective topics such as tone, rapport, and empathy.
Ticket Closing Speed
Agents who work in a technical support call center often open a “trouble ticket” to track each new support issue. While it seems reasonable to evaluate agents based upon their average speed to close a ticket, this can actually lead to two problems.
The first problem occurs when agents close tickets before verifying the problem has been solved. In one call center, agents doing this would simply open a new ticket if the customer called back to complain that they were still experiencing the problem. This practice reduced their average time to close a ticket, even though the average time to resolve a problem was much longer.
The second problem is agents may prioritize tickets that are easy to close. They realize that a high volume of tickets closed very quickly will offset a few challenging tickets than can take a long time to resolve, thereby lowering their overall average. Of course, this also means that the customers with the most challenging problems have to wait even longer to get a resolution.
Solution: Evaluate agents based upon a post-transaction customer satisfaction survey rather than the time required to close a ticket. This focuses agents on customer satisfaction and reduces their incentive to close a ticket before the issue is fully resolved.
Call centers provide a sea of data that can be used to manage more effectively, but it is essential that the metrics used to evaluate agent performance are clearly aligned with the call center’s goals.