Speech Analytics and Quality assurance: How to Combine Them?

One of the main concerns of any company is providing good customer service. After all, it is through customer service that companies are able to retain their customers and ensure a good return on investment.

To ensure quality of service, companies usually invest in various solutions, such as quality assurance and speech analytics.

In this content, we’ll explain in detail about both and how to combine them to take your customers’ experience to another level.

What are the objectives of Speech Analytics and quality assurance in call centers?

The purpose of speech analytics is to automatically analyze conversations to generate insights about the service provided, in order to improve the quality of service.

Quality assurance, on the other hand, is a manual process that evaluates the service provided to ensure that it complies with established standards, also generating insights for the company.

Although the processes differ, speech analytics and quality assurance can be combined to optimize the service provided by the call center.

What processes done in quality assurance can Speech Analytics optimize?

There are several processes that can be done in quality assurance, from simple filling out forms to analyzing complex data. However, some of these processes might not be as effective as expected, and that’s where a time optimization tool like Speech Analytics comes in.

Let’s go step by step:

Selection of contacts

In common quality assurance:
The selection of which contact to evaluate is done by sample, and the market usually evaluates 0.5 to 1% of contacts.

In practice, many call centers evaluate each attendant at least 4 times a month, with the goal of one evaluation per week.

With Speech Analytics:

Quality assurance can be efficient for manual processes, but it certainly fails for several extreme cases of conflicts and problems in customer contacts. That’s why Speech Analytics delegates specific contacts to the assurance team based on criteria such as:

  • Negative feelings
  • Offensive word detection
  • Missing script follow-up
  • Process errors
  • Undesirable behaviors

Among others, depending on your settings…

Form filling

In common quality assurance:

Completing forms is a routine task in the life of quality monitors, and there are several forms in which items of:

  • Script tracking
  • Product knowledge
  • Negotiation
  • Argumentation
  • Posture, ethics and language.

To evaluate all these aspects, it is essential to calmly analyze the contact and interactions between agents and customers, whether by text or audio.

With Speech Analytics:

The fact is that if your call center has well-defined service processes, a large part of filling out the quality form can be automated, especially points that are simple to detect, such as:

  • Data confirmation
  • Sales audit
  • Argumentation

Between others…

The important thing is that the quality monitor invests his time in complex analyzes that cannot be reproduced by a machine, devoting human time to what matters most.

Feedback application

In common quality assurance:

Feedback is applied in the vast majority of evaluations, even those in which the attendants behaved well in their services.

With Speech Analytics:
The feedback process is much more analytical, considering that we will be talking about contacts that were problematic, this renews the efficiency of the operation, focusing on the main pain points of the service team.

Report analysis

In common quality assurance:

We will analyze typical reports from the area, such as:

  • Quality score
  • Offenders report
  • Productivity of monitors and supervisors

With Speech Analytics:

You will gain views that evaluating manually would be almost impossible to build, such as :

  • Reasons for silence on calls
  • Customer sentiment (and what is driving negative sentiment)
  • Most problematic contact reasons

Among others, depending on your needs…

How to deploy speech analytics software?

1. Set up your account

To get started, you’ll need to set up an account with a speech analytics software provider. Look for a provider that offers a free trial, so you can try out the service before committing to a plan, just like we do with CYF Analytics.

2. Select your audio files

After setting up your account, you’ll need to select the audio files you want to analyze. Generally, we offer several ways to upload audio files, including drag and drop, integration with popular cloud storage services, and APIs.

This will depend a lot on how it is negotiated.

3. Choose your settings

After selecting your audio files, you will need to choose your analysis settings. Like, for example :

  • Keyword search
  • Reasons for contact
  • Blocks of silence

Among several others. Choose the settings that best suit your needs.

4. Run the analysis

After choosing your settings, you will need to run the analysis.
We generally offer a variety of ways to run the analysis, including real-time, D-1 (one business day), and cloud-based. Choose the option that best suits your needs.

5. See the results

After the analysis is complete, you will be able to preview the results.

Having visions like:

  • Customer sentiment
  • Response speed
  • Average silence time
  • Blocks of silence
  • Reasons for contact
  • Offensive words
  • Comparison between teams
  • Entity Recognition
  • Script tracking

And much more.


Using Speech Analytics takes effort to make the tool perfect for your business rules. Unlike a quality assurance software, Speech Analytics works almost like a dedicated project for your company’s processes and needs.

Therefore, when deciding which tool to choose, consider the advice you will get about your needs and your Speech Analytics settings.

Talk to us and see how to combine CYF Analytics and Quality to take your operation to another level.


Leave a Reply

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