Call listening: How (and why) to change your contact center’s approach

“Do you like calling into a call center?”

I’ve asked my customers this question every chance I’ve gotten for the past eight years (and whenever someone asks me what I do). The answer has always been “no.” Personally, I have been calling into call centers since the early 1980s. I hate doing so, and for the most part, I haven’t noticed the experience getting any better.

So why is this experience not getting better when most call centers have people responsible for call listening (i.e., listening to calls for quality assurance and agent training purposes)? 

The issue is call centers simply haven’t been listening enough. Even when they do listen to calls, it isn’t effective. The QA managers or team supervisors can only listen to a small percentage of an agent’s calls each month.  People also only have time to listen for and track a limited number of things on each call. These small samples don’t paint an accurate picture of an agent’s overall performance or the customer experience. 

I’m not saying contact centers should stop listening to their calls; I’m saying call centers should listen more using technology! After working in the contact center industry for more than thirty years, I know the importance of understanding what’s going on in your customer conversations. When you understand what’s happening in your customer conversations at scale, you can do a better job coaching your agents and make better decisions that improve your customer experience.

There are many parts of the call listening process you can streamline or automate so managers spend less time reviewing calls and get better insights than they could through manual listening. To do this, you need conversation intelligence technology, which automatically transcribes spoken words and tracks what is being discussed in all of the customer conversations.  It’s essentially “listening” to more calls while freeing up managers and supervisors to spend more time on high-value activities.  

What is call listening?

There are two primary call listening approaches in contact centers:

• Call recording listening: This is when supervisors or QA managers listen to a call recording after the call ends. The goal is to understand what happened on the call and what the agent could do better next time. Contact center supervisors or QA managers often complete evaluation forms or QA scorecards while listening to call recordings so they can reference these forms when coaching their agents.
• Live call listening: This is when supervisors listen in on an agent’s call while it’s taking place. The supervisor can listen silently, “whisper” (speak to the agent without the customer hearing), or “barge in” (join the conversation so both the agent and customer can hear). 

Conversation intelligence can improve both live call listening and call recording listening, and I’m going to be talking about both.

How conversation intelligence improves call recording listening

Conversation intelligence software loads call recordings and converts spoken words to text transcripts. After generating a transcript, this technology uses machine learning (a type of AI) to analyze the meaning of specific words and phrases. In Tethr, we automatically add category labels to different points in the transcript. This makes it easy for managers to see when specific topics or behaviors occurred. 

So how does this improve call listening?

1. You can save time by reading the transcript

First, you can’t overlook the value of automatic transcription. You can read twice as fast as you can listen, which means you can cut your call review time by 50%. With Tethr, you can also complete your QA forms (we call these “Evaluations”) in the platform while reviewing transcripts. You don’t have to toggle between different tabs or pause recordings and listen to the same section three or four times to make sure it meets your evaluation criteria.

2. You can pinpoint what you need to listen to

There are situations where you may still want to listen to call recordings. For example, maybe you want to hear how customers talk about a specific product or how agents handle a specific issue. But let’s say you have a 45-minute call recording and the customer and agent only discuss the subject you’re interested in for three minutes. It’s not a good use of your time to listen to the full call to answer one question.

Conversation intelligence software allows you to find specific call moments quickly by automatically applying labels to transcripts based on the phrases that appear. (In Tethr, we call these “categories.”) For instance, a company that sells pool equipment might set up a category for pool pumps so that a label appears at every point in their transcripts when a customer or agent mentions “pool pumps” or related terms. From there, a manager could jump to all the relevant sections and listen to them to answer their questions.

Tethr also makes it easy to browse transcripts and see all the categories that appeared in each one. This gives you a high-level overview of what happened on the call before you even open the transcript–and makes it easy to filter your results to see only transcripts with specific categories.

Pro tip: The audio in telephony systems is “stereo,” with each participant's voice recorded in separate channels. When a company stores the recordings, they can maintain the recording as stereo or they can combine and compress the channels into one “mono” channel which reduces the quality of the audio. I recommend using stereo instead of mono for your audio recordings whenever possible. Transcription accuracy is much better with stereo audio files because it’s easier for machines to identify what’s being said when there are multiple audio channels.

How conversation intelligence improves live call listening

Some conversation intelligence platforms (including Tethr) offer real-time analytics in addition to post-call analytics. For real-time analytics to work, the platform has to load and transcribe conversations as they’re taking place and apply category labels to relevant phrases as they appear. 

Contact centers can set up automatic triggers for specific categories. These triggers can either send a screen pop to the agent with a reminder or tip for navigating the conversation or send an alert to a supervisor to let them know they should check on the conversation.

By automating live call listening, conversation intelligence can improve the experience and productivity of both managers and agents.

1. Agents get guidance without being derailed

In traditional call listening, a manager might use a “whisper” approach to advise an agent without the customer hearing. While this might be beneficial for some agents, especially inexperienced ones, it can also be distracting. The agent ends up with both the customer and their manager in their ear, potentially talking at the same time. Additionally, the manager’s advice might come too late or be wrong–often, managers are trying to monitor multiple conversations at once, and human error is always possible.

A conversation intelligence platform can “listen” to all ongoing conversations and send agents relevant prompts at the right time. Seeing these prompts is a bit like seeing a billboard while driving down the highway. In both cases, you can quickly process the content without losing focus on your primary task (at least I hope so).

The agent benefits by getting non-intrusive prompts that help them perform well on their evaluations, and the customer benefits by getting efficient help resolving their issue.

2. Managers assist where they’re most needed

There’s nothing wrong with a manager sitting with an inexperienced agent and listening to a call to help them–but this shouldn’t be the norm. When this happens, the contact center is paying for two people to handle one call. And it’s not just the cost that’s an issue. Managers may be responsible for 10 to 15 agents (or more, depending on the staffing ratio). There’s no way they can listen to the calls of all their agents at the same time.

Conversation intelligence technology relieves the pressure on managers by sending automated advice and reminders to agents in real time. And it still allows managers to stay in the loop and direct their attention where it’s most needed. For example, in Tethr Live (our real-time analytics platform), managers can see a dashboard with a list of their agents’ transcripts. Managers can configure notifications so that when there’s a category hit for a serious issue, such as a legal threat, they get an immediate alert and can take action. 

Final thoughts

It’s time to change the way we think about call listening in the contact center. When I first started managing contact centers in the late 1980s, manually listening to calls was the norm because we didn’t have the technology to make that process more efficient. Now, 35 years later, even though the technology to automatically listen to calls at scale is widely available, too many contact centers are still relying on manual call listening.

When you use conversation intelligence to reduce the time managers spend manually listening to calls, you’re not reducing your call listening. You’re actually expanding your call listening to all your customer conversations–it’s just a different concept of listening. Automating your call listening process–while still keeping humans in the loop–lets you analyze your calls at scale, uncover better coaching insights, and improve your overall customer experience. I look forward to the day when I ask someone “Do you like calling a call center?” and get a “yes.”