Skip to content

Using Speech Analytics to Improve Your Contact Center

By Donna Fluss
Speech Analytics

View this document on the publisher’s website.

Speech analytics solutions gather insights into the performance of contact centers and other operating areas within an enterprise. These solutions analyze recorded conversations, which means that they “know” what is happening in the contact center and can often identify what needs to be done to resolve the issues. Speech analytics provides the following types of information to contact center managers:

  1. Agents: Which agents are doing a good/excellent job and who needs additional training in specific areas
  2. Regulatory compliance: Which agents are following or failing to comply with internal and external regulations
  3. Policies and procedures: Which operating policies are annoying customers and causing frustration
  4. Complaints: The underlying reasons why customers and prospects are unhappy with a company, and what to do to fix the problems
  5. At-risk customers: Which customers are at risk of attriting, and what it will take to retain them
  6. Competitive challenges: Which competitors have new offerings that are attracting the attention of your customers
  7. New products: Which new products and services your customers want, or how they want them delivered


It’s surprisingly easy to gather basic findings from speech analytics. The process is to create a search using key words/phrases and then run the application to see what it captures. Within a day or two of implementing a speech analytics solution, managers often collect so much information that they are overwhelmed. For this reason, DMG recommends that organizations start small by creating a list of 2 to 5 issues that they want to address. Since recordings are captured within the contact center, many companies start by concentrating on improving or optimizing that department and agent performance. This is an excellent approach, as it gives speech analytics analysts an opportunity to learn to use the solution prior to rolling it out throughout the enterprise. Here are some typical findings when speech analytics is used in a contact center:

  1. The reasons people call – the top reasons why people call your contact center, and what you can do about it
  2. Which agents are doing a good job and who needs coaching – these findings can be striking, as they may differ from the results of the established quality assurance process
  3. The specific types of coaching needed by agents – the areas of opportunity that will most contribute to improving customer satisfaction while improving productivity
  4. Where training programs need to be improved – which topics are not being covered well enough to prepare agents to handle calls
  5. Why agents put customers on hold or transfer them – policies and procedures, system issues, training, etc. that have to be addressed with the department to reduce the number of times and duration that customers are place on hold


Some companies have started to tie speech analytics findings to their gamification environments, while others share team findings with all supervisors to encourage healthy competition. While there are some countries where regulations prohibit findings from being shared broadly, the point is that it’s important to let the staff know why and how the speech analytics solution is being used and to reward agents, supervisors and managers for using the data to improve the member experience and reduce operating costs.

The real fun of speech analytics is that you don’t always know what is going to be found; use this to your advantage, and reward behaviors and performance that promote your company and its bottom line. By starting small and applying findings in the contact center, you’ll earn the respect of managers inside and outside of the department, which presents a great opportunity to broaden the use of speech analytics to other operating areas. Once you show how impactful speech analytics can be in the contact center, other managers will want to use the solution for their own departments.