Improve Agent Performance and Customer Satisfaction with Analytics

Amtelco Updated Logo

By Kevin Beale

One of the biggest challenges we face in the call center industry is hiring, training, evaluating, and retaining agents. When a call center is fortunate enough to find a new candidate, it’s critical to train and help them become productive as quickly as possible. Advanced analytic tools can reduce training time, streamline evaluations, and consistently increase customer satisfaction.

Agents that are not properly trained or lack the knowledge and confidence to complete calls result in poor caller satisfaction and loss of revenue, both for the call center and for the clients the call center serves. Those agents typically will not last if they continue to struggle.

Training an agent has traditionally started with sitting with the agent as they take calls, providing feedback and direction on how to improve. As the agent progresses, the process shifts to the recording of their calls and listening to those recordings. With more agents working remotely, that traditional onboarding and training process becomes much more difficult.

Systems that utilize automated scripting with dynamic prompting and automated dispatching, such as Amtelco’s Genesis Intelligent Series with Web Agent, can reduce the training time. Web Agent guides an agent through each step of a call, providing a user friendly, intuitive interface. This can be a big benefit but is not enough on its own.

As agents become more experienced, they can process calls independently. This is when the next phase of training and evaluation kicks in. This phase typically involves a manager listening to a small number of each agent’s calls per shift. Managers listening to the call recordings typically have an evaluation form they fill out as they listen to each call. This process can take several minutes per call. Due to the large volume of calls an agent handles each day, only a small percentage of the calls can be evaluated using this method.

Imagine having a tool that automates this process. This can have a tremendous impact on agent productivity. Agents receive more immediate feedback, enabling them to improve their performance incrementally rather than waiting for an evaluation to take place, or worse yet, waiting for a customer complaint.

Automated evaluations can have a tremendous impact on management productivity, saving valuable time by eliminating the need to listen to call recordings and manually fill out evaluation forms. Many call centers have established a target of evaluating three to five calls per agent shift. This small number is chosen because it is all the management team can handle. Compare that to the benefits of evaluating 100 percent of each agent’s calls.

Active Insights

How can we fully automate listening to calls and evaluating the outcome of the calls? We can use software such as Amtelco’s Active Insights to help us. Active Insights automatically transcribes the call recordings into text and feeds that text into a machine learning and artificial intelligence model to produce an automated evaluation of the call.

Active Insights with artificial intelligence can take common situations that occur every day in a call center and produce a summary of those situations and can raise alarms for situations that require attention and follow up.

Example 1

Here is a sample transcription of a call where a caller became angry because they had called in recently and had to repeat the same information.

Transcription Snippet:

Agent: Thank you for calling ABC Clinic. How may I help you?

Caller: Are you that same operator I just had ten minutes ago? They couldn’t figure out how to get me to my doctor. I need to talk to my doctor. No excuses.

Agent: Well, I’m not sure sir. Uh Let me see if I can help you with that.

Caller: Ok, I sure hope so because my foot really hurts still.

Agent: Uh Can I have your name please?

Caller: Don’t you have my info? I just gave it to you ten minutes ago. Can you pull it up? I don’t wanna have to go through this all again.

Agent: Um I’m sorry sir, but I don’t.

Caller: Ok. Uh fine . . .

As we can see, this call did not go well from the start. The caller was angry that they had to give their information again after calling in a short while ago. Active Insights produced this outcome for this call:

Summary: The caller contacts the call center and expresses frustration that they had spoken to someone ten minutes ago who couldn’t help them reach their doctor. The agent tries to help but the caller is rude and uncooperative. The caller provides their name and phone number but speaks over the agent when they tried to confirm the number. The caller becomes even more frustrated and asks to speak to a supervisor, but the agent says the supervisor is currently at lunch. The caller abruptly ends the call after requesting that the agent contact their doctor and have the doctor call them back. The overall caller sentiment is angry and frustrated.

Active Insights also flagged this call as having an Angry caller. This Angry caller alert can immediately be sent to a manager, alerting them to the problem. This is a valuable tool that can alleviate and address issues before they become a problem.

Caller Angry: The caller became angry because they had spoken to another operator previously who was unable to help them, and they were frustrated that they had to go through the process again.

Example 2

In another example, a caller may become frustrated because the agent lacked the proper knowledge or training to complete the call to the caller’s satisfaction. This can be a valuable teaching tool for agent performance and improvement:

Transcription Snippet

Agent: Thank you for calling ABC Clinic. How may I help you?

Caller: Hi, I’m looking to get in touch with my doctor.

Agent: Oh, well, please, uh, bear with me. I’m new. I’m not sure how to find the doctor for you.

Caller: That’s ok. I have time. So take your time.

Agent: Uh, well, I’m not sure if I’m doing the right things here. But, uh, can I have your name?

Caller: Chris

Agent: And your phone number?

Caller: xxx-xxx-xxxx. (Redaction of PHI is an option.)

Agent: Well, I’m not sure if this is the right thing to do here, but, uh, is there a certain doctor you’d like to speak with?

Caller: Yeah. So that’s ok. So usually when I call and they ask me, uh, who my primary doctor is, and I usually tell them Doctor Smith. So probably Doctor Smith.

Agent: Yeah, I’m not sure how to find Doctor Smith. Um, it might be best if you call the office in the morning . . .

Summary: The caller wanted to get in touch with their doctor and provided their name and phone number to the agent. The agent was new and unsure how to help the caller but offered to take a message for the doctor. The caller asked the agent to tell the doctor that they hurt their foot, but also mentioned they would try calling the office directly. Overall, the caller seems unsatisfied with the conversation, and unsure if the agent was able to help them. Caller sentiment: Negative


The above calls are common situations that occur every day in a call center. Most of these calls would go unnoticed using the traditional method of listening to a small percentage of calls each day. The unsatisfied or angry caller would reach the call center client and would express their frustration directly with the client.

This would degrade the client’s confidence in the call center and could cause the client to leave the service. In addition, without feedback on these calls, the agent would not improve and would continue to struggle, resulting in the probable loss of that agent.

In addition to the above caller sentiment analysis, Active Insights can evaluate agents’ use of grammar, positive attitude, proper manners, and accurate answer phrase.

Automated call analysis can have a positive impact on the day-to-day call center operations and on the bottom line. Agents will quickly come on board, gain confidence, and enjoy their job more because they receive immediate feedback on their performance.

Callers will be satisfied because they are speaking with confident and positive agents. Call center clients will be happy and will remain with the service because their callers are happy.

Kevin Beale is vice president for software, research, and development for Amtelco. Reach out to Amtelco to explore these options with them. They can help you take advantage of advances in technology and improve your call center.

%d bloggers like this: