Speech Analytics in Call Centers: How Does It Work? What is it for?

Speech Analytics has ceased to be a differential element of certain call centers and over time has become a fundamental tool required by several companies, but how can Speech analytics in call centers help in practice?

Before getting into topics, we must understand the general process of Speech Analytics:

  1. Call transcription
  2. Automatic analysis
  3. Human analysis


Call transcription

Like the others, this is an extremely important process for Speech Analytics, so important that if it is not done correctly it will undoubtedly have a great impact on subsequent Analytics processes and can even take everything down the drain.

The call transcription process, in a nutshell, consists of transforming phone calls into text.

The problem with this process is that not all telephone recordings are of satisfactory quality to the point of being perfect for transcription. Calls in many cases will have:

  • Echo
  • Buzzing due to loss of client signal
  • Even cases where the customer is talking to someone else (and not the agent)
  • Low volume of one or both parties
  • Bad signal-to-noise ratio
  • Noise in the environment
  • Bad pronunciation and diction
  • Regional accents and words
  • Others


All the above cases will undoubtedly affect the results of the Speech analytics, which is what we will talk about in the next topic.

It’s also important to note that there are several types of transcription providers, from giants like Google, Amazon, and Microsoft, to niche providers for the call center universe specifically, and there can also be huge variations in quality and price…

Automatic analysis

After the transcription, we start the automatic analysis part.

With phone calls transformed into text, the possibilities for automatic analysis and classification are enormous and can vary depending on the needs of the company using the tool.

The ideal and what we recommend is that you configure your Speech analytics tool to:

  1. Identify the points that bring money to your business.
  2. Identify the points that make your company lose money.

As I said before, this varies from company to company, but in general, Speech Analytics can detect several types of results in a single phone contact, such as

  • Complaint signals
  • Suspicion of fraud
  • Classification of reasons for contact
  • Most pronounced words
  • Quiet time vs. talk time
  • Silence blocks in the middle of the call
  • Reasons for Silence Blocks
  • Automatic filling of a quality assurance scorecard
  • Voice speed
  • Dissatisfaction
  • Possible mood swings
  • Correct greeting
  • Confirmation of data
  • More assertive sales scripts
  • Sales audit
  • Others

These are the possibilities that CYF Analytics today can detect and integrated with human analysis, you can have a huge return on top of these results.

Human analysis

There are several types of areas in call centers that, when integrated with automated analysis, can have their work much improved and at the same time less manual.

We have, for example, CYF Quality, which is integrated with CYF Analytics, which allows the quality assurance area to focus its efforts on more complex analyzes, leaving aside the completion of elements that do not require human analysis to be identified.

Other areas that can greatly benefit are:

  • Business intelligence: Information on complaint signals, scripts that close sales the most, and arguments that resolve cases on first contact, can be a key point in rethinking all internal scripts, improving business results.


  • HR / Training: in addition to identifying the agents who fail the most with automatic analysis, the training area can delve into, for example, the doubts that most generate a silence block (period in which the assistant asks the client for a time to search your system), thus focusing the training on these more complex topics for the agents.


  • Operation: Speech analytics can easily help operations managers to better achieve their objectives, identifying which agents are not following the processes or which processes have better results.


  • Quality: we have already mentioned above that the quality department can benefit from the identification of dissatisfaction.


It is important to note that the same tool can be shared with all the areas mentioned above, facilitating its adaptation to the budget of each department.


Speech Analytics is a very powerful tool that should be used by the most strategic areas of any company that wants to improve its results.

CYF Analytics can capture everything that we have mentioned in this content, it is a system that can be configured to search for what you want, making the possibilities practically endless and helping a lot to understand the main needs in your call center.

Leave a Reply

Your email address will not be published. Required fields are marked *