Blog

How to improve the efficiency of the Sales Team with Speech Analytics

In every call centre, there are 3 categories of agents:

  1. Agents who are self-driven, motivated and are top performers.
  2. Agents who are at the bottom 10% and don’t perform well.
  3. Agents who are in the middle segment who are doing well, but not great.

Improvement in the efficiency of #3 (the middle segment) can drastically improve the performance of the call centre.

Let’s take an example of a telesales setup where the agents sell Banking and Financial Services products like Loans, Credit Card, Insurance, Demat account, etc. There are so many financial institutions out there. Therefore, the cost of a lead is quite high. Usually, it goes into thousands of Rupees per lead. Imagine spending thousands of Rupees in getting that lead only to see that your telesales agent who wasn’t performing well, couldn’t convert that lead into a paying customer.

This was because they didn’t really understand what the customer wanted and couldn’t offer them the right product. Or, they couldn’t properly explain the product features to the customer. This is not just a lost opportunity, but also a lost cost spent on that lead.

We’ve spent time in the call-centres and we’ve seen the struggle of the floor managers, supervisors and call centre heads who are under tremendous pressure to drive the revenue and conversion targets. Here are a few ways by which Speech Analytics can help in a telesales setup:

  1. Identifying hot leads that were lost because a low performing agent couldn’t help the customer properly. These leads can be passed on to the top performing agents who can again call the prospect and convert them into paying customers.
  2. Analysing every single call of the middle performers to identify the areas of improvement and training gaps.

A Speech Analytics solution can accurately convert every single call recording into text and analyse it across various sales parameters like understanding the customer needs, probing for information, objection handling, tone of voice, creating a sense of urgency, patiently listening, mis-selling, etc. These scores across each call can then be aggregated to create a performance chart for each agent.

Take a look at this chart below where we’ve plotted the performance of an agent Bhupendra (yellow line) against the performance of the top performing agent - Anmol Singh (blue line) and an average performer (red line). This chart has been prepared by analysing every single interaction of Bhupendra, Anmol and others by a Speech Analytics Software. The chart has been plotted on a bunch of sales-related parameters.

Here are a few observations from this chart:

  1. On multiple parameters, Bhupendra (yellow line) falls under the 4 - 6 band, compared to Anmol (blue line), who falls under the 6 - 8 band. For example, “Explaining in Simple Terms” is one such parameter. “Understanding Customer needs” is another such parameter.
  2. Clearly, Bhupendra (yellow line) needs more training in these areas because they are performing quite low compared to the average (red line) and top performers both.
  3. On the parameter named “Comparison with Competitors”, the average (red line) is low and Anmol’s performance (blue line) is also low. This indicates that the entire batch of agents might need training on this parameter.

Further, when Bhupendra is given feedback, he can be shown the highest scoring calls of Anmol, so as to get a sense of how to speak with the customer in such calls. The feedback in this way is more quantified, data driven, accurate, and action-oriented. In this way, it is much easier for Bhupendra to understand and improve the quality of his sales pitch.

In a similar manner, a Speech Analytics solution can help in Collections and Customer Service calls.

  • Collections: Audit each call to check for any rude language or tone by the agent, or any calls where the customer threatens to complain to the regulator.
  • Customer Service: Audit each call to check for adherence to the script and compliance violations to flag fatal calls. For instance, calls where the agent gave incomplete or inaccurate information, or put the call on hold several times.

In Insurance and Broking, there are a few other interesting use-cases:

  • Insurance: Audit each “policy revival” call to check if the agent is offering flexible payment plans to the customers or not.
  • Broking: Audit each dealer call to check if the order placed by the customer on call matches with the order placed by the dealer in the system.

At GreyLabs AI, we’ve built a powerful Speech Analytics Solution that accurately converts each call recording into text and then uses Generative AI powered Large Language Models (LLMs) to extract insights from the calls. We are proud to work with leading Banks and Financial Institutions as our Customers.

Interested in improving the performance of your Call Center? Contact us today for a demo!