Automated Interaction Selection and Evaluation for Quality Management

In this guide, we'll learn how to automate the selection and evaluation of customer interactions for quality management. This process removes the need for manual selection by managers and allows for flexible policies to define which conversations are reviewed and how they are distributed. We will also cover how to set up evaluation forms, assign scores, and generate reports based on the results.

Let's get started

First, we are focusing on automated selection of interactions. This will eliminate the need for managers to manually search the system for interactions to evaluate with agents. This can be fully automated by setting up a policy for each department. Then, specify which conversations you want to select. I select conversations that are between three and eight minutes long. This ensures they are not too short or too long for the evaluation process. Next, specify who should receive the interactions—either the manager or a team member, such as a quality manager. Then, indicate how many interactions you want to assign. In this case, I selected five per agent per month. These are sent to me as the quality manager and appear in my inbox on the right-hand side.

Each agent sends me five evaluations, with five interactions per month.

1

You can also choose weeks or days, but at the end of the selected time period, those five per agent will arrive in my inbox.

2

I can see in the inbox who the agent is and then go directly to that interaction.

3

We are fully flexible in setting this up. As the quality manager, I can review the interaction as usual, with the form displayed on the right side.

4

That's the second major part of this process. We use the GBI CSC form this year to identify and record the questions that matter most to us. First question: Did we establish who the customer is? Yes, no, or not available. If the agent had no opportunity to do so, we do not need to score it. You can say, "Not applicable" or "Not available." If you select "yes," it will assign a specific number of points to each question.

5

We can set this up individually. For example, we might decide that listening skills are more important than customer recognition. A "yes" on question two scores five points, while this response scores only one point. Strategic focus is crucial. We can set this up flexibly to concentrate on the most important points, while still including other key elements for effective customer interaction. Once this is done and the form is set up, I can go through it, answer the questions, and submit it. Submitting it will send it to the agent. They need to accept the result. After that, it will move into reporting.

6

I chose Mike Graham's team this June. They had seven evaluations completed by Mike, with an average score of 94%. All the points where we answer yes or no will be summarized. The system will calculate a score out of 100%, showing what the agents did correctly or incorrectly. Let's look at one that Mike and Jill filled out together.

7

In this case, "Did we establish who the customer is?" If not, that's zero points. For listening skills, if yes, that's three points. You can see there are already different weights in the current GBI form. For example, the last question asks, "Did the advisor demonstrate best practice and correct use of C4C?" If yes, it's one point; if no, it's zero. The system then calculates the total available points. Lead generation was not available here, so the agent could not create a lead or discuss advisory services. This will not be counted, and it will not affect the overall result.

The system calculates the available points and scores them. It then translates these scores into KPIs that we can agree on with the agents.

8

That's essentially the main part—those two sections. First, the agent interaction selection is fully automated. Second, we can review evaluation forms to score whether everything was completed during the interaction.