Call center monitoring is one of the most effective ways call centers can improve their customer service; it gives managers the opportunity to watch (or rather listen) to their agents in action.
By monitoring what agents say during interactions, managers can easily identify problems within their teams, maintain quality standards and compliance, improve the customer experience, and improve overall call center performance. However, listening in on a few calls here and there isn’t going to do any good for your agents, your customers, or your call center at large.
Here are three ways you can improve your call center monitoring practices:
1. Analyze 100% of calls.
Let’s say call center managers were able to manually review 5 random calls per agent each month for quality assurance and training purposes. Depending on the call volume that means only a few percent of calls are actually used for coaching and training, which is nowhere near a statistically significant sample size.
How can a manager effectively monitor their call center when they don’t have trust-worthy data to pull actionable insights from?
Call quality monitoring is all about changing agent behavior and managers need to understand how agents act day-in and day-out, not just basing their entire review and training plan on 1 or 2 calls. Instead of manually reviewing a small sample of calls, managers can improve call center monitoring with a speech analytics system where 100% of calls are automatically recorded AND calls are monitored in real-time.
Now you are getting a 360 degree view of your call center and can trust your action items are based on sound data, not just casual observations.
Make sure your call center monitoring goals accurately represent your customers’ expectations and aren’t just about what numbers you wish to see from your agents. If you’re not monitoring for and measuring what customers’ value then you’re just wasting your time and tasking your agents to hit certain goals that could actually undermine the customer experience.
For instance, Average Handling Time (AHT) is a commonly reviewed metric because the faster an agent wraps up an interaction the faster they can move onto the next call, thereby shortening call wait times. But just because an agent gets through their calls quickly that doesn’t mean they are getting through them effectively. An agent primarily obsessed with speed might not fully resolve the issue, prompting the customer to call back or try another service channel. Vice versa, an agent might spend 10 hours on the line with one customer but that doesn’t automatically mean they are a terrible employee even though they failed to meet your call center’s AHT goals.
It’s important to remember that you can’t evaluate your agents in a vacuum.
3. Involve agents when developing the scorecards/evaluation forms used during the monitoring process.
When your whole team understands what you’re looking for during the monitoring process you have a much better chance of adjusting behavior as people are aware of how they are being evaluated.
For instance, define what constitutes a “quality” interaction and the subcategories that are being scored.
- How important is script adherence?
- How important is handling time?
- Will agents be judged on how well they handle escalating calls with frustrated customers?
Providing examples that demonstrate the excellent, average and or poor quality of the behaviors being measured gives your agents a solid base to work from and outlines expectations of performance. And since you are now analyzing 100% of calls they know that each interaction has to be the best it can. They can’t just phone it in (pardon the pun) most days and put on their best behavior when they know they are being monitored.
Having a call center monitoring procedure in place is half the battle, but if you want to get real actionable insights and make real performance changes it’s imperative that call centers focus on improving their monitoring practices as much as they can. The more trustworthy data you can collect the more positive impact you can have.